V4 Process Technology 2

Volume 4
Process Technology II
1. Kiln Systems................................................................................................................... 2
2. Internal Kiln Fittings .................................................................................................. 172
3. Refractory Linings ..................................................................................................... 204
4. Firing Systems ........................................................................................................... 244
5. Burners and Flames................................................................................................... 286
6. Combustion Engineering .......................................................................................... 327
7. Alternative Fuels ........................................................................................................ 366
8. Circulation Phenomena ............................................................................................. 508
9. Kiln Control................................................................................................................. 566
10. Heat Balance............................................................................................................... 607
11. Main Fans.................................................................................................................... 671
Cement Manufacturing Course Version 2005 Volume 4 - Page 1

Chapter 1
Kiln Systems
1. Kiln Systems - Overview ............................................................................................... 3
2. Rotary Kilns .................................................................................................................. 36
3. Suspension Preheater ................................................................................................. 48
4. Precalcining Systems .................................................................................................. 69
5. Clinker Coolers........................................................................................................... 110

Kiln Systems - Overview
Urs Gasser
PT 99/14501/E
1. PROCESS REQUIREMENTS FOR KILN SYSTEMS ......................................................... 6

2. PROCESS TYPES .............................................................................................................. 8
2.1 General ................................................................................................................. 8
3. WET PROCESS ................................................................................................................ 12
3.1 General ............................................................................................................... 12
3.2 Long Wet Process Kilns ...................................................................................... 13
3.3 Wet Process Kilns with Slurry Preheaters........................................................... 14
4. SEMI WET PROCESS ...................................................................................................... 16
4.1 General ............................................................................................................... 16
4.2 Semi Wet Process Long Kilns............................................................................. 16
4.3 Semi Wet Grate Preheater Kilns ......................................................................... 16
4.4 Semi-Wet Suspension Pre-heater Kiln................................................................ 18
5. SEMI DRY PROCESS....................................................................................................... 19
5.1 Semi-Dry Process Long Kilns ............................................................................. 20
5.2 Semi-Dry Process Grate Pre-heater Kilns .......................................................... 20
6. DRY PROCESS ................................................................................................................ 23
6.1 Long Dry Kilns..................................................................................................... 23
6.2 Raw Meal Suspension Preheater Kilns............................................................... 23
6.2.1 General............................................................................................................ 23
6.2.2 One and two Stage Cyclone Pre-heater Kilns ................................................. 24
6.2.3 Four Stage Cyclone Pre-heater Kilns .............................................................. 25
6.2.4 Precalciner Kilns.............................................................................................. 27

SUMMARY
Todays kiln systems for burning cement clinker of major importance use a rotary kiln.
Exceptions are vertical shaft kilns still used in certain geographical areas. With very rare
exceptions, new plants use the dry process. However, there are still important markets
where older wet process plants are predominant (USA, Russia).
A first classification of the process can be made based on the water content of the kiln feed:
< 1% water dry-process

10 ... 12% water semi-dry-process
17 ... 21% water semi-wet-process
25 ... 40% water wet-process
Dry-Process
Precalciner kiln with 4 to 6 cyclone stages (contemporary technology):
Separate tertiary air duct
50 - 60% fuel to the precalciner
Large capacities possible > 10000 t/d
Up to 4000 t/d in 1 string
Heat consumption < 3000 kJ/kg possible (6 stages)
Sensitive to circulation phenomena (-> kiln gas bypass!)
4-stage cyclone pre-heater kiln (standard technology 1970 to 1980):
Cyclone stages (co-current flow) for raw meal preheating
Large application world wide
Capacities of up to 4500 t/d technically possible
Heat consumption: 3150 to 3350 kJ/kg cli
2-stage cyclone pre-heater kiln:
Less sensitive to circulation phenomena than 4-stage pre-heater
Higher heat consumption than pre-heater with more stages
Shaft pre-heater kiln:
Counter current heat exchange between hot gas and raw meal
Practical efficiency inferior to cyclone pre-heater
Long-dry-kiln:
Rather simple equipment
High dust emission from kiln tube
Without heat exchange internals: high heat consumption of up to 5100 kJ/kg cli
With chains and/or crosses: 4200 kJ/kg cli achievable

Semi-dry and semi wet process
Grate pre-heater kiln (LEPOL, ACL):
Raw meal must be suitable to be nodulised with water (semi-dry)
3450 kJ/kg cli (no waste heat available for primary raw material drying)
Long rotary kiln and suspension preheater:
Filter cakes fed or slurry injection into vertical dryer; rather rare cases
Wet-process
Long wet kiln:
Fed with raw meal slurry of approx. 32 - 42% water content
Internal heat transfer improved by chains
High heat consumption of 5300 to 6300 kJ/kg cli due to evaporation of water
Heat consumption reduced by slurry thinners for a slurry with 25 - 30% H2O
Slurry preheaters can reduce kiln size and improve heat exchange

1. PROCESS REQUIREMENTS FOR KILN SYSTEMS
The kiln system has to be designed to cope with the requirements of the chemical process
during which the kiln feed material is converted into cement clinker.
This process as a whole is endothermic and takes place at maximum material temperatures
of 1450C. Receiving its thermal energy from hot gases of up to 2000C generated by
combusting fuels, it is also referred to as pyroprocess.
Type of reaction and temperature development are compiled in sequence of reactions
occurring in a rotary kiln (table 1) and graphically as the quasi-qualitative variation of
minerals with temperature (figure 1).
The chemical process taking place in the kiln system where raw meal (input) is converted to
cement clinker (output) can be subdivided into the following five steps:
1. Drying
2. Preheating
3. Calcining
4. Sintering
5. Cooling
Process and equipment has been developed and improved with the aim at performing these
steps forever improved economy, which means
High availability
Low heat consumption
Low power consumption
Higher unit capacity
Stable kiln operation
Good, uniform clinker quality
Table 1 Sequence of Reactions occurring in a Rotary Kiln
Temperature Type of reaction

range (C)
Heating Up
20 - 100 Evaporation of free H2O
100 - 300 Loss of physically absorbed water
400 - 900 Removal of structural H2O (H2O and OH groups) from clay minerals
> 500 Structural changes in silicate minerals
600 - 900 Dissociation of carbonates CO2 driven out)
> 800 Formation of belite, intermediate products, aluminate and ferrite
> 1250 Formation of liquid phase (aluminate and ferrite melt)
approx. 1450 Completion of reaction and re-crystallisation of alite and belite
Cooling
1300 - 1240 Crystallisation of liquid phase into mainly aluminate and ferrite

2. PROCESS TYPES
2.1 General
The criterion normally used to distinguish the process types is the moisture of the kiln feed
material. Four basically different process types for clinker burning can be defined:
Process Type Feed Material Cons. Feed Moisture Feed System

Dry process Raw meal Dry < 1% H2O Mechanic, pneumatic
Semi dry process Nodules Moist 10 ... 12% Mechanic, pneumatic
H2O
Semi wet Filter cake, Moist 17 ... 21% Mechanic, pneumatic
process nodules H2O
Wet process Slurry Liquid 25 ... 40% Hydraulic
H2O
Table 1 gives a general survey of the various rotary kiln systems in operation for industrial
clinker production. Shaft kilns, which are still used in China or experimental systems such as
sintering grates or fluidised beds, are not considered in the scheme.
We can distinguish two main groups of kiln systems:
a) Long kilns with or without internal heat exchanging installation
b) Short or medium kilns with external preheaters
(e.g. suspension preheaters, grates or external slurry preheaters)
The heat consumption of burning depends strongly on the water content of the kiln feed
This can be illustrated by the typical specific heat consumption: The fuel consumption of wet
kilns is nearly twice as high as for modern dry process suspension pre-heater kilns.
The comparison of the heat economy within each process group (dry or wet) shows clearly:
The more intensive the heat-exchange for drying and preheating, the lower the heat
consumption.
Other than based on the feed moisture, kiln systems can be grouped in different ways:

Process Type wet >25% H2O in feed Slurry
semi wet 17 - 21% H2O in feed nodules from slurry
semi dry 10 - 12% H2O in feed nodules from meal
dry < 1% H2O in feed raw meal
Production Mode batch+cont. < 200 t/d shaft kilns
continuous 300 t/d 10000 t/d rotary kilns
Heat Consumption 2900 kJ/kg cli state of the art system
( 700 kcal/kg cli)
> 6000 kJ/kg cli long wet or dry kilns,

(> 1400 kcal/kg cli)
not optimum operation
Power Consumption 20 to 65 kWh/t cli kiln feed to clinker cooler
OVERVIEW OF KILN AND PROCESS TYPES

When the concept for a new plant is developed, not only the present situation but also the
possible future developments of all relevant factors must be taken into account.
The following main parameters must be considered when selecting the kiln system:
Raw material:
moisture content
grindability
homogeneity of deposit
number of components for raw mix
chemical composition (sulphur, chlorides, alkalis, organic compounds etc)
filtration properties of slurry (for semi-wet process only)
Plant installation and operating costs
Requirements for clinker quality (e.g. low alkali clinker)
Aspects of environmental protection (emission of dust, SOx, NOx, etc)
Technical standard of the country
Long wet (and dry) rotary kilns are the oldest and most simple type of installation to produce
cement clinker. The pyroprocess takes place in a long rotating tube, which has usually
internal equipment to improve heat transfer, and, in wet kilns, to reduce dust loss. Unit
capacities of up to 2000 t/d are typically achieved, higher outputs are possible, however,
they require kilns of gigantic dimensions.
Today, economy requires plants for 3000 to 10000 t/d. Therefore new plants are almost
always based on the dry process with preheater, pre-calciner and reciprocating grate cooler.
The semi wet process for a new plant could be preferred in special cases, e.g. where raw
material with a high natural moisture must be used (e.g. quarry below water level).
The three following graphs illustrate the development of the significance of the various
processes within the Holcim group, which can be considered representative of the global
situation.

Kiln Systems in the Holcim Group I
70
60
50
40
wet
semi wet
30
semi dry
dry
20
10
0
1981 1983 1985 1987 1989 1991 1993 1995 1997

Kiln Systems in the Holcim Group III
70 1800
1600
60
1400
50
1200
40 1000 Total Group
avg kiln cap

30 800
600
20
400
10
200
0 0
1981 1983 1985 1987 1989 1991 1993 1995 1997
3. WET PROCESS
3.1 General
The wet process was the most important process for clinker burning in the past and almost
all plants were wet. Heterogeneous quarries and corrective addition were no problem;
stirring of the liquid slurry in the slurry tanks provides very good batch-wise blending.
Grinding was done in slurry mills, which consume 30%, less energy than dry ball mills, but at
higher lining wear rates.
The disadvantage of the wet process is the high heat consumption. Compared to e.g. a
suspension preheater kiln, the difference is more than 2000 kJ/kg clinker or 60 to 70%!
Today, with efficient dry homogenising technology available, the wet process is no longer
applied for new plants. Investments as well as operating costs of a wet system are higher
than for dry systems of the same output. Technical development allows using more efficient
kiln systems even where wet plants would have been built in earlier times.
Another reason for preferring the wet process in the past was the production of low alkali
cement (alkali content < 0,6%) and the fact that difficult circulation problems are easier to
control in wet kilns. Today secondary firing or efficient bypass installations with precalciner
are possibilities to keep these problems under control also in modern kiln systems.
Because of the lower specific gas volume and the shorter rotary part, rotary kiln dimensions
as well as gas handling, dedusting and fuel preparation can be designed accordingly
smaller. Although new wet kilns are no longer considered for new plants, they still play an
important role in the US as well as in many countries of Eastern Europe and Central Asia.

3.2 Long Wet Process Kilns
Long wet kilns have been the most commonly used burning reactors for a very long time, but
because of the high water content of the feed, their heat consumption is up to twice as high
as for modern dry systems.
The milled and homogenised raw material is a slurry with a water content of typically 32 to
42% and is pumped to the kiln inlet.
In the first zone heat transfer for the evaporation of water is always increased by means of
chain systems (extended surface, higher relative velocities, increase of turbulence). The
chain systems should also reduce the dust losses and clean the kiln shell. These internal
heat exchanger installations require very special know-how, based to a large degree on
experience (see separate paper chain systems).
In order to decrease fuel consumption the water content should be kept as low as possible.
The limit is normally the pumpability of the slurry. It is basically possible to further reduce the
slurry moisture by using slurry thinners. This technology has been successfully applied and
will provide an economical advantage if adequate quantities are available at low cost, e.g. as
industrial by-product.
Example: Beauport (Canada): 28% feed moisture

Wet kilns are relatively insensitive to circulation problems because the critical temperature
ranges are in the rotary part of the kiln (see also circulation phenomena).
Low alkali clinker can be produced from high alkali raw material simply by selectively wasting
of dust: The highest enriched kiln dust (e.g. from the last precipitator compartment) is
removed from the process (i.e. dumped onto a dust pile) as necessary. The rest of the dust
can be reintroduced to the kiln by dust scoops or insufflation into the burning zone.
Today, discarding dust creates increasing problems because of restrictive permitting of dust
piles.
Note:
Kiln dust cannot just be blended to the slurry because it would react and thicken the slurry.
Typical technical data for long wet kilns with chains:
Heat consumption q 5000 ... 6300 kJ/kg cli (1200 ... 1500 kcal/kg
cli)
Kiln exit gas temperature 150 ... 250C
System pressure drop 0,5 ... 1,0 kPa
Dust emission in % of clinker 5 ... 100%
production
Probably the largest wet process kiln in the world is installed at Holnams Clarksville plant
(Michigan USA). This kiln has a diameter of 7,6 m and a length of 232 m with a daily
capacity of about 3600 t.
3.3 Wet Process Kilns with Slurry Preheaters

External Slurry Preheaters
In order to improve the heat exchange between gas and slurry and to reduce the kiln size,
external slurry preheaters have been developed by MIAG (Kalzinator) and Krupp
(Konzentrator). Both of them are revolving drums with special internal packing. These drums
have about the same diameter as the kiln, its length being slightly smaller than the diameter.
The capacity of these machines is limited to 800 -1000 t/d and frequently operating problems
arise. Very often, external preheaters were large sources of false air.
Internal Slurry Preheaters
F.L. Smidth designed a slurry pre-heater system integrated into the kiln compartment, which
should avoid the disadvantage of external slurry preheaters. In practice, this construction
turned out to be very sensitive to clogging.
A better system developed by Fives Cail Babcock is installed in the three kilns at Obourg.
Lifting buckets and chain curtains produce a slurry curtain that keeps back a high amount of
dust and improves heat exchange.

Wet process kiln
dust : 0.15 kg/kg
0.9 Nm3/kg, 180C
35 % H2O
5500kJ/kg 20 C
20mbar 50mbar 200 C

- 10 mbar
2 %O2
clinker : 1kg cli, 1400C gas : 900C 3.0 Nm3/kg
secondary air : 1.4Nm3/kg, 670C meal : 400C
1 kg cli
installed : 3.0 Nm3/kg, 20C
80C
operation: 2.3
Capacity : 100 - 3600 t/d Kiln feed : slurry, 28 - 43 % H2O

Diameter : 2.5 - 7.5 m Kiln circumferential speed : 10 - 20 m/min
Length : 40 - 230 m Kiln slope : 2.5 - 4 %
Length/Diameter : 30 - 38 Total chain weight : 10 - 15 % rel. to capacity
Kiln load (inside lining) : 0.4 - 0.8 t/m3d Dust losses : 0.05 - 0.25 kg/kg cli
BZ load (inside lining) : 40 - 100 t/m2d Spec.heat consumption : 5000 - 7000 kJ/kg cli
CC-99072.dsf / Kma 17.2.99

4. SEMI WET PROCESS
4.1 General
A process is considered semi-wet if the kiln feed is produced from wet slurry. A mechanical
water extraction process reduces the water content of the kiln feed to 17 to 21%. A number
of filter presses operating batch-wise are commonly used, but also continuous filter band
presses or similar equipment would be possible.
4.2 Semi Wet Process Long Kilns

Principally, long kilns with heat exchanger crosses can be fed with slurry, filter cakes or dry
meal. Feeding filter cakes is a straightforward and simple solution and is used by Italcementi
in some cases.
4.3 Semi Wet Grate Preheater Kilns

Most of the semi-wet systems use a grate preheater kiln fed with filter cakes.
A grate preheater system includes a short rotary kiln (similar to a four stage preheater kiln)
where only calcining and sintering take place. For drying, preheating and partial calcining, a
travelling grate is installed in front of the kiln, where heat of the kiln exhaust gases is used.
For the semi-wet grate kiln, the slurry must be prepared in a special way so it can be fed to a
travelling grate:
The pumpable slurry as starting material is fed to filter presses where the moisture content is
reduced to approx. 20% applying a filtration pressure of 15 to 20 bar. In a special type of
extruder (Siebkneter), the filter cakes are converted into cylindrical nodules (diameter 15 ...
20 mm, length 30 ... 50 mm) and then fed to the preheater-grate. The economy of this way of
preparation depends strongly on the filtration properties of the slurry.
Operating and performance data are similar to the semi-dry grate preheater systems
described under 5.2.
Characteristic data of a semi-wet grate pre-heater system:
Feed Nodules made from Slurry Filter Cake
Moisture Content of the Feed 10 ... 12%
Heat consumption q 3770 kJ/kg cli ( 900 kcal/kg cli)
Exit gas temperature after grate 100 ... 120C
System pressure drop 2,6 kPa
Example of a semi-wet LEPOL kiln:
ABs kiln 10 at the Lgerdorf plant (Germany)
Maximum kiln capacity: 3600 t/d
Kiln dimensions: 6.0/5.6 m x 90 m
Grate dimensions: 5.6 x 61.7 m
Secondary firing with Fullers earth (special)
(Shut down; replaced by semi wet precalciner kiln in 1996)

Production of Cement by the Semi-Wet Process
Feed Bins
Water
Slurry Mill
Chalk Quarry Clay Quarry

Corrective Materials
Electrostatic
Filter Press
Coal Precipitator
Coal
Mill
Hot Gas
Air
Generator
Compressor
Rotary Kiln
Filter
Mineral Slurry Basin
Clinker Cooler Gypsum Components
Bulk Dispatch Packing Machine Bag Palletization
Clinker
Storage
Cement Cement
Silo Silo
Figure 4
Cement Mill
CC-99056.dsf Kma 17.2.99

4.4 Semi-Wet Suspension Pre-heater Kiln
The suspension preheater kiln is normally fed with dry meal (details see separate paper).
However, there are some rare cases where suspension preheater kilns are fed with nodules
prepared from slurry. These nodules should not be too strong because they must be cracked
by thermal shock or abrasion before being fed to the kiln system via top stage of the pre-
heater.
A two-stage pre-heater kiln operated with semi-wet nodules was e.g. the Liesberg plant.
There, the nodules were cracked in a vertical dryer before being fed to the preheater.
The first modern kiln system using this principle has been built in the late 1980s by FLS in
Aalborg Cements RORDAL plant. It is a three stage two string kiln system with precalciner
for a capacity of 4000 t/d. The high operating cost of the filter presses has been avoided by
directly injecting the slurry into a drier-crusher followed by a vertical drier. The semi-wet
process was selected because the raw material (chalk) is mined under water and has very
high natural moisture.
From the Holcim group:
Example of a semi-wet pre-heater/pre-calciner kiln:
ABs kiln 11 at the Lgerdorf plant (Germany)
Maximum kiln capacity: 4500 t/d at 3900 kJ/kg
Kiln dimensions: 4.8 x 65 m; 2 supports, gearless friction drive
Preheater: 3 stages, 2 strings
Utilisation of various alternative fuels in both firings
Supplied by Polysius; start-up: 1996
Filter cakes produced in already existing filter-presses

Semi wet process kiln (with grate preheater "LEPOL")
100 C
- 8 mbar
12 %O2
3.5 Nm3/kg
dust: 0.01kg/kg
1.3 Nm3/kg
120C
Filter cakes
20% H2O
2.1 kg/kg
300C
5%O2
1.7Nm3kg
3800 kJ/kg
1000C
clinker : 1kg cli, 1400C dust

dust
secondary air: 1.0 Nm3/kg, 810C 0.1 kg/kg
0.02kg/kg
1 kg cli
Capacity : 300 - 3000 t/d Kiln feed : filter cakes, 18 - 23 % H2O
80C
Diameter : 3 - 5.6 m Kiln speed : 1 - 2 rpm
Length : 45 - 90 m Kiln slope : 3 - 3.5 %
Length/Diameter : 11 - 16 Spec.heat consumption : 3600 - 4500 kJ/kg cli
Kiln load (inside lining) : 1.5 - 2.2 t/m3d
BZ load (inside lining) : 80 - 150 t/m2d
Grate load : 8 - 10 t/m2d
CC-99071.dsf / Kma 17.2.99
5. SEMI DRY PROCESS

The semi-dry process is characterised by the fact that kiln feed nodules are made from dry
raw meal. Water is added in order to produce nodules with 10 - 12% moisture.

5.1 Semi-Dry Process Long Kilns
There are long kilns with heat exchanger crosses fed with nodules. This system was applied
by Italcementi and looks very similar to an installation for semi-wet feed material.
5.2 Semi-Dry Process Grate Pre-heater Kilns

The grate preheater kiln is by far the most popular semi-dry system.
The principle of the grate preheater system for the semi-dry process is identical to the one
used for the semi-wet process. What is different is the feed preparation:
The dry raw material is mixed with water (10 ... 12%) and nodulised in a drum or preferably
on a rotating plate (pan noduliser). This system can be used only for raw materials
containing plastic components enabling the formation of nodules that are resistant against
thermal shock and abrasion. The main factor influencing plasticity is the mineralogical
composition, especially the presence of montmorillonite.
On the grate, heat exchange from the gas to the nodules forming a fixed bed layer of approx.
20 cm thickness is excellent. In some grate preheaters, precalcination is done successfully,
often using even waste fuels (such as Fullers earth, acid sludge, waste lubricating oils etc.)
utilising secondary firing.
The only successfully working travelling grate pre-heater was available from Polysius and
became known under the name LEPOL system (American licensee: Allis-Chalmers, ACL
system).
Production of Cement by the Semi-Dry Process
Feed Bins
Quarriy (ies) Crushing Plant (s) Raw Materials Storage Corrective Materials
Electrostatic
Precipitator
Coal Water
Homogenizing and
Storage Silo
Granulating
Disc
Coal Raw Meal

Mill
Travelling Grate
Preheater
Rotary Kiln
Raw Mill
Mineral Filter Hot Gas
Clinker Cooler Gypsum Components Generator
Bulk Dispatch Packing Machine Bag Palletization
Clinker Cement Cement

Storage Silo Silo
Cement Mill
CC-99055.dsf Kma 17.2.99
This principle sketch shows a LEPOL kiln fed with nodules made out of dry raw meal.
LEPOL kilns built after 1945 are equipped with two-pass grates; i.e. the exhaust gas is led
twice through the nodule bed from top to bottom:

The hot kiln gas passes first through a bed of dry and preheated nodules and subsequently,
after an intermediary dedusting once again trough a layer of moist incoming nodules. It is
believed that the nodules survive throughout the process resulting in a clinker with very
uniform size.
Furthermore, dust loads in the kiln atmosphere and dust emission out of the system are low.
The nodules on the grate let only pass the fine dust while the coarse particles are retained.
In cases of increased trace compound concentrations (especially alkali) in the raw material,
the fine dust separated in the electrostatic precipitator is largely enriched with them. Only a
small amount of dust has to be discarded to reduce the balance of these compounds in the
kiln system. This effect makes the LEPOL kiln quite suitable to produce a low alkali clinker
with rather low heat consumption. For this reason, it has been chosen in many cases,
particularly in the USA.
The following limits and disadvantages have to be considered:
Only raw materials with good plastic properties can be used
(semi-wet: filter cake nodules -> good filtration properties are required)
The grate chain is subject to wear.
Uneven temperature distribution across the grate can cause difficulties.
Additional theoretical heat consumption due to the water content of the feed (partially
compensated by a low exit gas temperature).
Exhaust gases cannot be used in drying and grinding systems.
Characteristic data of a semi-dry grate pre-heater systems:
Feed nodules made from dry raw meal
Moisture content of the feed nodules 10 ... 12%
Specific heat consumption q 3450 kJ/kg cli (= 820 kcal/kg cli)
Exit gas temperature after grate 100 ... 120C
System pressure drop 2.6 kPa

Semi dry process kiln (with grate preheater "LEPOL")
100 C
- 8 mbar
8 %O2
2.2 Nm3/kg
dust: 0.01kg/kg
meal nodules
12% H2O
1,9 kg/kg
1.3 Nm3/kg Raw Meal

1300C
300C
5%O2
1.6Nm3kg
clinker : 1kg cli, 1400C
secondary air: 0.9 Nm3/kg, 860C
3400 kJ/kg
1000C
dust dust
0.02kg/kg 0.1 kg/kg
1 kg cli 80C
Capacity : 300 - 2000 t/d Kiln feed : meal nodules, 11 - 14 % H2O

Diameter : 3 - 4.8 m Kiln speed : 1 - 20 rpm
Length : 35 - 80 m Kiln slope : 3 - 4 %
Length/Diameter : 11 - 16 Spec.heat consumption : 3200 - 3600 kJ/kg cli
Grate load : 80 - 12 t/m2d
CC-99070.dsf / Kma 17.2.99

6. DRY PROCESS
6.1 Long Dry Kilns

Without internal heat exchange equipment
The simplest kind of dry process installation is the long dry kiln without any internal heat
exchange equipment (empty tube). With a heat consumption of 5100 kJ/kg cli (1200 kcal/kg
cli) or about 90% of the wet process it must be considered very uneconomical. Advantages
might be its simplicity and insensitivity to heavy circulation problems.
This kiln type is suitable to be used in combination with waste heat recovery steam boilers
for power generation. In that case, the waste heat contained in the hot kiln exhaust gases is
further used to produce valuable energy.
Characteristic kiln data:
Heat consumption q 4500 ... 6000 kJ/kg cli (1075 ... 1430 kcal/kg cli)
Kiln gas exit temperature 450 ... 500C
With internal heat exchange equipment
Long dry kilns with internal heat exchange equipment (chains or crosses from steel or
ceramic material) represent a more economical solution. Heat consumption of 4200 kJ/kg or
even less can be achieved. Other typical operating figures are contained in annex 10.
Heat consumption q 3800 ... 4500 kJ/kg cli (910 ... 1075 kcal/kg cli)
6.2 Raw Meal Suspension Preheater Kilns
6.2.1 General
During the last thirty years, the suspension preheater kiln became the dominant clinker
manufacturing system. This system is fed by dry raw meal that is preferably prepared in a
grinding and drying plant, using the kiln waste gases as a drying medium. This ground and
dried raw meal is homogenised and then fed to the preheater where it is suspended in the
kiln gas flow, where an extremely effective heat transfer takes place. More information is
contained in the special section Suspension Preheaters.

6.2.2 One and two Stage Cyclone Pre-heater Kilns
one stage: Heat consumption q 3750 ... 4000 kJ/kg cli (900 ... 950 kcal/kg cli)
two stages: Heat consumption q 3500 ... 3750 kJ/kg cli (850 ... 900 kcal/kg cli)
Production of Cement by the Dry Process

Feed Bins
Sampling Station
Quarry (ies) Crushing Plant (s) Raw Materials Storage and Preblending Corrective Materials
Electrostatic
Homogenizing and
Precipitator
Cooler Dedusting Storage Silo
Coal
Raw Meal
Water
Air-to-Air
Cooler Coal
Mill
Cyclone
Rotary Kiln Preheater Conditioning Raw Mill
Tower
Mineral Filter
Clinker Cooler Gypsum Components Bulk Dispatch Packing Machine Bag Palletization
Cement Cement
Clinker Silo Silo
Storage
Cement Mill
CC-99054.dsf Kma 17.2.99

6.2.3 Four Stage Cyclone Pre-heater Kilns
Until the mid 1980s, this arrangement belong to the systems with the lowest fuel
consumption. It was offered in several configurations with capacities up to 4500 t/d, most of
them being combinations of single or twin cyclone stages.
The kiln exit gas includes still enough heat to dry raw material up to moisture content of 8%
if the mill is running during all the kiln operation time. From this point of view, the remaining
relatively high exit gas temperature cannot be considered fully as a loss, because it can
substitute an auxiliary firing for raw material drying.
The preheater system is installed in a steel or concrete tower with a height of about 60 to
120 m (6 stages) above the kiln inlet, depending on capacity and concept.
The four to six stages preheater is most susceptible to circulation problems at presence of
excessive concentration of circulation compounds causing clogging problems in the pre-
heater system.
The sketch shows a conventional four stage cyclone preheater system. In the 1970s,
production lines with more than approx. 2000 t/d had to be built with two parallel preheater
strings. Today, one-string installations are possible for up to 4000 t/d.
Characteristic operating figures of 4-stage pre-heater kilns:
Heat consumption q
small units 3350 ... 3550 kJ/kg cli

(= 800 ... 850 kcal/kg cli)
large units 3150 ... 3350 kJ/kg cli
(= 750 ... 800 kcal/kg cli)
kiln exit gas temperature 320 ... 350C
kiln exit gas volume approx. 1,5 Nm3/kg cli
System pressure drop 4 ... 6 kPa
Dust loss relative to clinker 8 ... 15%
Transition chamber
kiln gas temperature approx. 1100C

Material temperature approx. 800C

Dry process kiln ( with 4-stage SP)
360 C
1.65kg/kg -50 mbar
50 C 3,5 %O2
1,41 Nm3/kg
dust: 0.1kg/kg
550C
1.45 Nm3/kg, 240C
720C
830C
3250kJ/kg
1050 C
- 3 mbar
20mbar 50mbar
2 %O2
1,2 Nm3/kg
= 30%
clinker : 1kg cli, 1400C
secondary air: 0.9 Nm3/kg, 860C
installed: 3.0Nm3/kg, 20C
1kg cli, operation: 2.3
80C
Capacity : 500 - 3200 t/d Kiln speed : 1.5 - 20 rpm
Diameter : 3.5 - 5.5 m Kiln slope : 3 - 4 %
Length : 50 - 90 m Spec.heat consumption : 3150 - 3500 kJ/kg cli
Length/Diameter : 14 - 16
CC-99073.dsf / Kma 17.2.99

6.2.4 Precalciner Kilns
For larger production capacities, a larger portion of the pyroprocess had to be relocated out
of the rotary kiln in order to maintain reasonable kiln diameters without excessive thermal
load of the burning zone.
The process of dissociation of CO2 (calcination) is suitable to take place in a static reactor
outside of the rotary kiln. Of the total heat consumption, 60 to 65% are required to achieve
about 90% of calcination. 100% calcination must be avoided because clogging problems will
seriously disturb kiln operation (beginning of clinker formation).
The development of this reactor started with a secondary firing in the kiln riser duct sufficient
for 35 to 40% calcination of the meal, combustion air still pulled through the kiln tube (=air
through). It was therefore referred to as precalciner (PC) type AT. Only when hot cooler air
(= tertiary air) for the PC fuel (= secondary fuel) was taken to the calciner in a separate duct,
the so called tertiary air duct, the full benefit of this technology could be used. Today, only
this type called PC-AS (=air separate) is considered a real precalciner. The elements of a
precalciner kiln system are explained in the sketch.
The strongest boost of calciner development was in the seventies in Japan, initiated by the
demand for very large units exceeding the potential of conventional kilns with suspension
preheaters. Only precalciner technology makes todays largest units of 10000 t/d possible.
Two process alternatives of precalciner are used:
in-line calciner (calciner installed in kiln gas flow)
separate-line calciner (calciner not passed by kiln gases)
More details on calciner technology are contained in a separate section.
The operating data are very close to the ones of the corresponding preheater kiln system.
In-line calciners have a tendency to higher gas exit temperature and system pressure drop;
however, modern units are equipped with 5 or 6 preheater stages to compensate for this.
Characteristic operating data of 4 to 6 stage precalciner kilns:
Heat consumption q
small units, 4 stage SP 3350 ... 3550 kJ/kg cli

(= 800 ... 850 kcal/kg cli)
large units, 5 stage SP 2900 ... 3200 kJ/kg cli
(= 700 ... 800 kcal/kg cli)
SP exit gas temp. 6 to 4 st. SP 290 ... 370C
SP exit gas volume approx. 1.3 to 1.5 Nm3/kg cli
System pressure drop 4 ... 6 kPa
Dust loss relative to clinker 8 ... 15%
Transition chamber:
kiln gas temperature approx. 1100C
Material temperature approx. 800C
More data of precalciner kiln systems are shown in the section Precalciners.

Dry process kiln (with 5-stage SP and PC) 310 C
-50 mbar
3 %
1,65 kg/kg 1.30 Nm3/kg
50C dust: 0.1 kg/kg
500C
1.45Nm3/kg, 240C
650C
1800 kJ/kg (60%) 770C
870C
2%O2
0.5Nm3/kg, 1100C
1200 kJ/kg
(40%)
1100 C
20mbar -3 mbar
50mbar 2 % O2
clinker : 1kg cli, 1400C 0,43 Nm3/kg
secondary air: 0.35 Nm3/kg, 1100C = 90%
Installed: 2.0Nm3/kg, Capacity : 1500 - 10'000 t/d Kiln speed : 2.5 - 4.5 rpm
1kg cli, 20C Diameter : 3.5 - 6.2 m Kiln slope : 3.5 - 4 %
110C operation: 1,8 Length : 40 - 105 m Spec.heat consumption : 2900 - 3500 kJ/kg cli
Length/Diameter : 10 - 16
Kiln load (inside lining) : 3 - 5.5 t/m3d
CC-99079.dsf / Kma 17.2.99

HEAT BALANCE
WET / SEMI-DRY / 4-ST. PREHEATER / 5-ST. PREHEATER-PRECALCINER
WET SEMI-DRY 4-STAGE SP 6-STAGE SP-
PROCESS LEPOL PC
nput kJ/kg % kJ/kg % kJ/kg % kJ/kg %
cli cli cli cli
Fuel kiln combustion 5560 96.7% 3343 97.6% 3150 97.7% 1180 39.2%
sensible heat 25 0.4% 15 0.4% 13 0.4% 5 0.2%
Fuel PC combustion 0 0.0% 0 0.0% 0 0.0% 1775 58.9%
sensible heat 0 0.0% 0 0.0% 0 0.0% 8 0.3%
Kiln sensible heat 25 0.4% 30 0.9% 54 1.7% 45 1.5%
feed
sensible heat of 73 1.3% 17 0.5% 0 0.0% 0 0.0%
water
Insufflated air (PA, 67 1.2% 20 0.6% 6 0.2% 0 0.0%
cooler)
Total inputs 5750 100% 3425 100% 3223 100% 3013 100%
Output kJ/kg % kJ/kg % kJ/kg % kJ/kg %

cli cli cli cli
Heat of formation 1750 30.4% 1750 51.1% 1750 54.3% 1750 58.1%
Water evaporation 2370 41.2% 506 14.8% 13 0.4% 8 0.3%
Exhaust gas sens. heat 754 13.1% 314 9.2% 636 19.7% 553 18.4%
Exh. gas dust sens. heat 25 0.4% 21 0.6% 18 0.6% 29 1.0%
Clinker 59 1.0% 50 1.5% 63 2.0% 83 2.8%
Cooler waste air 100 1.7% 276 8.1% 423 13.1% 288 9.6%
Radiation and convect :
- Preheater 0 0.0% 160 4.7% 77 2.4% 60 2.0%
- Precalciner (or bottom 0 0.0% 0 0.0% 20 0.6% 20 0.7%
stage)
- Kiln (+tertiary air duct) 530 9.2% 200 5.8% 200 6.2% 200 6.6%
- Cooler 10 0.2% 92 2.7% 10 0.3% 10 0.3%
Water cooling 0 0.0% 42 1.2% 0 0.0% 0 0.0%
Other outputs 0 0.0% 0 0.0% 0 0.0% 0 0.0%
Rest 152 2.6% 14 0.4% 13 0.4% 12 0.4%

Total outputs 5750 100% 3425 100% 3223 107% 3013 100%

HISTORICAL DEVELOPMENT Annex 1
The word cement is more than 2000 years old, but impure lime has been used much longer
as a building material. It is historically established, that the Phoenicians used a pozzolanic
lime about 700 B.C. and also the Romans produced some sort of cement or hard burned
lime. From the medieval ages, it is known that in Holland a type of hydraulic cement was
formed out of lime and tuff in dome shaped kilns.
Our cement, as we know it today, is now more than 200 years old, invented by the
Englishman John Smeaton in 1756. It was burned in bottle kilns. The better known inventor
of Portland cement was Joseph Aspdin, who patented his burning process in 1824. He
also used dome kilns of approx. 36 ft height and 17 ft diameter with a production of 90 bbl (=
15 t) per charge, each of which took several days to produce. Fuel consumption was 50% of
clinker weight in coal which corresponds to 15500 kJ/kg cli (= 3700 kcal/kg cli).
In 1880 an important step forward was made with the development of the continuously
working shaft kiln, which had a much better heat economy. An example of such a kiln was
the Dietzsche Etagenofen which is shown in Annex 1.
From 1877 experiments have been conducted with rotary kilns. In 1897 Hurry and Seaman
developed the first successfully operating unit of this type in America.
These first rotary kilns were wet process kilns with a daily capacity of 50 to 100 tons. Their
heat consumption was again very high (about 30% of clinker in coal = 9500 kJ/kg cli) and
they had an incredible dust emission (usually more than one third of the whole production).
In order to decrease heat consumption, chain systems were installed in wet kilns to improve
heat transfer during drying. Behind long dry kilns, waste heat steam boilers were arranged
for the same purpose.
It took almost another 30 years, before a further substantial reduction of heat consumption
could be achieved by reducing the water content of the feed and by a better heat exchange
in the preheating a calcining zone. In 1930 an officer of the army of the tsar, Dr. Lellep, took
an important step in this direction. He developed the travelling grate pre-heater, which is fed
with moist nodules. This invention was taken over by Polysius and got the name LEPOL kiln.
Some years later, there was a Czech patent of a cyclone raw meal pre-heater, and in 1953
Kloeckner-Humboldt-Deutz AG in Germany installed the first suspension pre-heater
system for raw meal. This type of kiln now became dominant because of its heat economy
and nowadays other systems are only chosen in special cases. In former years, the main
reason for the selection of the wet process was, that effective homogenisation of ground raw
material was not possible except in the form of slurry. In developing special techniques for
dry material homogenisation such as mix beds, mixing chamber silos etc., this factor could
be eliminated.
Utilising a rather old idea, since about 1966 especially Japanese cement machine
manufacturers have designed several successfully working precalcining kiln systems.
Calcination is already done in a stationary calciner system, where secondary firing is
installed. By this means, it is possible to design kiln systems with a comparatively small
rotary part diameter but a very large capacity up to more than 10000 t/d.
Kiln systems built after 1990 include 6-stage preheaters with up to 4000 t/d per string, pure
air calciners, designed for a variety of fuels and emission control. Using modern low primary
air burners, low pressure drop cyclone designs and high recuperation efficiency coolers
allow further reduction of heat and power consumption.

Rotary Kilns
U. Gasser
PT 98/14362/E
1. General ............................................................................................................................. 38
2. Kiln Dimensioning ........................................................................................................... 40
3. Mechanical Aspects of Rotary Kilns.............................................................................. 42
3.1 Riding Ring Fixation, Kiln Shell Ovality ............................................................... 42
3.2 Kiln Seals ............................................................................................................ 44
3.2.1 Kiln Inlet Seal................................................................................................... 44
3.2.2 Kiln Outlet Seal................................................................................................ 44
3.3 Kiln Drive............................................................................................................. 45

SUMMARY
After over 100 years, the rotary kiln is used in all cement plants for clinker production.
The following properties made it superior to other principles:
suitable to cope with high temperatures
easy to be lined with refractory bricks due to its shape
material transport behaviour
tight to ambient
mechanically relatively simple
large units possible
The rotary kiln must be designed for process, combustion and mechanical requirements.
Characteristic figures: Length L [m] , diameter D [m] and their ratio L/D [-]
Slope [], speed range [min-1] and drive [kWh]
Dimensioning criteria: Volume load [t/(d m3)]
Burning zone load [t/(d m2)]
Thermal burning zone load [MW/m2]
Important mechanical features are:

riding ring fixation
roller station / alignment
seals at inlet and outlet
drive
With modern precalciner technology, outputs exceeding 10000 t/d per kiln are possible with
diameters still below the 6.5 m of the largest wet kilns.
There is a trend towards short L/D kilns with only two piers mainly because of lower
investment.

1. GENERAL
Today, all clinker producing installations of industrial size use a rotary kiln. The rotary kiln is
still the only feasible way to manage this high temperature process with process material of
varying behaviour.
One exception is the vertical shaft kiln still used in some parts of the world, e.g. China,
however, for small unit capacities only. The other exceptions are few pilot installations based
on sintering in a fluidized bed reactor.
Like many other great ideas, the rotary kiln was invented towards the end of the 19th century
and has found application in many different industries. In 1987, Hurry and Seaman in the
USA developed the first successfully working rotary kiln to produce cement clinker.
The first rotary cement kilns were using the wet process with one very long kiln tube, making
it the dominating single piece of equipment of a plant. With technological progress, the kiln
sections used for for drying, heating-up and calcining have gradually been replaced by other
types of equipment, the rotary kiln remains to be the most suitable type of machine for the
clinkerization process.
The rotary kiln has to satisfy three types of requirements:
Combustion: as a combustion chamber for burning zone
fuel
Process: as a reactor for the clinker burning process ( retention time)
as a material conveyor ( slope, speed)
Mechanical: stability of shape, carrying load, thermal
flexibility, tightness
Remarks:
Even though the rotary kiln is a relatively simple piece of equipment, nobody has
developed a complete theoretical/mathematical model of its behaviour and process
which would allow correct process simulation and equipment design.
The rotary kiln is still the heart of the entire production line. Its OEE (overall equipment
efficiency) depending mainly on hourly output and availability, is decisive for the success
of a plant.
The rotary kiln is designed to operate 24 hours a day, and the rest of the equipment
upstream and downstream has to follow.
Being a major cause for production cost (mechanical maintenance, refractories), a well
managed kiln is vital for a successful plant.

Figure 1: Old and new kiln

2. KILN DIMENSIONING
The kiln dimensions are defined with diameter D (for kilns with different diameter: burning
zone D) and length L:
L [m] and D [m] resp. L/D [m]
For cement kilns, the actual L/D ratio range is:
from 40 (for long wet kilns) to 11 (for modern short kilns with precalciner)
The diameter D is the inner diameter Di of the kiln (steel-) shell.
Process technological dimensioning of a kiln is based on empirical figures and
experience from existing installations
One limiting factor for the diameter is the mechanical stability of the arch of the brick lining.
Maximum diameters which can be safely realised with standard size bricks are about 6,5 m.
The largest kiln in the Holcim group is 232 m (wet process, 3750 t/d).
The following process technological dimensioning criteria are mostly used:
Specific Volume Load Clinker Production [t/(d m3)]

Net Kiln Volume
Specific Zone Load Clinker Production [t/(d m2)]

Net Burning Zone Cross Section
Thermal Burning Zone Load Burning Zone Heat Input [MW/m2)]

Net Bruning Zone Cross Section
Specific volume load and thermal burning zone (BZ) load have no physical significance.
They are merely defined to make existing installations comparable.
The specific load is indirectly a gas velocity, because generating a certain amount of thermal
energy by fuel combustion results in a proportional gas flow which can be calculated.
The thermal BZ load per cross section is considered the limiting factor for a modern kiln
system. For a certain length/diameter ratio, which is typical for each kiln type, the thermal BZ
load it is proportional to the heat load on the inside of the lining surface which is one of the
main influencing factor on brick life. The limit usually respected is:
Max. Thermal BZ Load = 6 MW/m2 (=5.16 x 106 kcal/m2 h)
Other absolute limiting values of all the three factors are not known. Each supplier seems to
have his own rules of kiln dimensioning. Since no theoretical formulas have been derived to
calculate the kiln size on an analytical basis, it is possible, that the present limits of the
dimensioning criteria may be surpassed even for the conventional processes.

Figure 2: Long and short L/D kilns
Rotary kiln zones
L/D = 30 (long wet kiln)
IZ DZ PZ CZ SZ TZ BZ KZ
L/D = 15 (dry preheater kiln)
CZ SZ TZ BZ KZ
L/D = 11 (short dry PC kiln)
DZ = Drying zone
PZ = Preheater zone
CZ = Calcining zone IZ SZ TZ BZ KZ
SZ = Safety zone
TZ = Transition zone
BZ = Burning (sintering) zone
KZ = Kiln internal cooling zone

3. MECHANICAL ASPECTS OF ROTARY KILNS
The following aspects of kiln mechanical design are relevant for the process:
Riding ring fixation, kiln shell ovality
Kiln seals
kiln drive
refractory lining (separate paper)
nose ring (covered in refractory lining)
3.1 Riding Ring Fixation, Kiln Shell Ovality

A rotary kiln should be designed as cheaply as possible, yet it must still be rigid to guarantee
minimum wear of the lining. This requirement can be met, if the deformation of the kiln shell
is reduced to a tolerable limit.
The parameter expressing shell deformation at a certain point is the kiln shell ovality :
Definition of : =2 (a - b) with 2a and 2b as the main axis of an ellipse

Investigations have shown, that generally a maximum relative ovality of 0,3% is allowed
This ovality may be subdivided into two amounts:
a) Ovality of the riding ring 3 cm due to external forces allowed value:

<= 0.2%
dr
a) Ovality of the kiln shell due to deformations by its own weight in loose riding rings and
due to increased temperature.
The following two requirements must be met to keep the kiln ovality within the tolerable
limits:
The riding rings must be rigid enough
The clearance between the ring shoes and the riding ring should be minimum during
operation. The following table shows some practical values:
Riding Ring No. 1 2 3 4

Clearance during operation [mm] 3-4 3-4 4-6 5-6
maximum [mm] 10-15
Riding rings with splined fixation provide much better support of the kiln shell. Because the
kiln shell is laterally suspended in adequately designed carrying bars, ovality is much
reduced resulting in noticeably better brick life.
Such systems are currently available from Polysius and FLS, the latter one is lso offered as
retrofit. Splined tire fixations are integral part of gearless kiln drive systems.

Figure 3: Tire fixations

3.2 Kiln Seals
In order to avoid the danger of hot gases and dust leaking into the atmosphere, the entire
kiln system is operated at negative pressure. The pressure profile starts at ambient (grate
cooler: above first grate, satellite and tube cooler: fresh air inlet) and becomes increasingly
negative towards the kiln induced draft (ID) fan.
Instead of leaking out from within the process, there is now a problem with ambient air being
sucked into the system, called false air. Depending on the point of entry, false air has
different undesired effects. That is why a lot of effort is made to keep process systems tight.
3.2.1 Kiln Inlet Seal

The kiln inlet seal (inlet: referring to material flow) is at point with negative pressure of less
than 10 mmWG (modern 2-support kilns) up to 100 mmWG (long wet kilns with chains).
Modern kilns with low suction have high temperatures (up to 1300C) instead.
False air entering the system causes
Additional gas to be handled by kiln ID fan and dedusting system
Unnecessary cooling of hot process gases reducing value of heat
Kiln inlet seals:
Sealing force by pneumatic cylinders (pneumatic); sealing-rings
Sealing force by coil springs/levers or weights (mechanical); sealing-segments
Sealing force by leaf springs and rope with weight; lamella (fish scale)
Kiln inlet seals must be equipped with a dust return scoop ring to avoid spillage of kiln feed.
Note:
The inlet seal is designed to seal against cold fresh air from outside, but it can be damaged
if it must seal hot gas from inside to ambient in case of system overpressure! (this happens
sometimes during the heating-up phase)
3.2.2 Kiln Outlet Seal

With grate and tube coolers, the kiln outlet seal is installed between kiln head and rotary kiln
where pressure should be slightly negative. Kiln outlet seals used with grate coolers must be
designed to cope with pressure pulsation with occasional positive pressure. Outlet seal and
nosering (brick retainer) with cooling air fan can be considered one system.
Here, the loss generated by false air reduces recuperation from the clinker cooler. Cold
ambient air replaces hot secondary air from the cooler which has to be vented.
Outlet seals designed specifically for this application of the following type are available:
Pneumatic
Mechanical
Lamella (fish scale)
Labyrinth (outdated)
With planetary coolers, false air reduces the amount of cooling air resulting in higher clinker
temperatures. The outlet seal is smaller, at lower temperature and negative pressure only.

Figure 4: Kiln seals
3.3 Kiln Drive

Kiln drives are designed for speeds between 1.0 and 4.0 min-1, depending on slope, process
and kiln dimensions. Long wet kilns are typically operated at the low end of this speed range
where some new high performance kilns (short L/D with precalciners) are running at the
upper end.
For over 10 years, rotary kilns have been driven by girth and pinion type drives. Decisive for
their performance are:
Correct dimensioning
Correct alignment (even load distribution on the flanks of the teeth; no peaks)
Adequate lubrication system and lubricant quality

With the new two support short kilns (L/D < 13) with long overhangs, kiln shell deformation
and burning zone much closer to the drive, it became more difficult to ascertain correct
alignment. Because of the determined load distribution on two piers, it became possible to
avoid the girth drive by using the kiln rollers to transfer the torque to the riding ring: the
gearless drive (=friction drive) was introduced. It is currently available from Polysius
(POLRO) and FLS-Fuller (ROTAX).
The following elements are part of this system:
Two supports for defined load on the driven tire

Splined tire fixation for safe torque transmission to the shell
Self-aligning roller station for linear load pattern between roller and tire (friction)
Today, there are only few kilns with friction in operation; the first one was Lgerdorf 11 by
Polysius. Detail optimization and long term experience are yet to be awaited.
Most systems have hydraulic drives for two rollers. This provides smooth operation, but is
expensive, rather complex (hydraulic unit) and has higher power consumption. Electric direct
drive of only one roller has been installed in one case.

Figure 5: Kiln drives

Suspension Preheater
U. Gasser
PT 98/14363/E
1. General ............................................................................................................................. 50
1.1 History ................................................................................................................. 50
1.2 Trend................................................................................................................... 50
2. Heat Exchange in a Suspension Preheater................................................................... 52
2.1 Counter-Current Heat Exchange (Shaft Stage) .................................................. 52
2.2 Co-Current Heat Exchange (Cyclone Stage) ...................................................... 52
2.3 Thermodynamic Limits ........................................................................................ 52
3. Preheater Types............................................................................................................... 54
3.1 Preheaters with Shaft Stages.............................................................................. 54
3.1.1 Pure shaft preheaters: ..................................................................................... 54
3.1.2 Hybrid preheaters: ........................................................................................... 54
3.2 Preheaters with Cyclone Stages ......................................................................... 56
3.3 Economical Number of Stages for Cyclone Preheaters...................................... 58
3.4 Minimum Gas Velocity ........................................................................................ 58
4. Design Features of Preheater-Cyclones........................................................................ 61
4.1 General ............................................................................................................... 61
4.2 Dust Cycles ......................................................................................................... 61
4.3 Features .............................................................................................................. 61
4.3.1 Splash Box ...................................................................................................... 61
4.3.2 Dip Tube (Immersion Tube, Vortex Finder, Thimble) ...................................... 63
4.3.3 Meal Flap......................................................................................................... 63
4.3.4 Cyclone Shapes .............................................................................................. 63
5. Preheater Operation ........................................................................................................ 65
5.1 Operating Problems of Suspension Preheaters.................................................. 65
5.1.1 Circulation Phenomena. .................................................................................. 65
6. New Developments.......................................................................................................... 66
6.1 Horizontal Cyclone .............................................................................................. 66
6.2 TRS ..................................................................................................................... 66

SUMMARY
Practically all modern kiln systems are equipped with a cyclone suspension preheater.
New installations include a precalciner with tertiary air duct, so that the preheater and
precalciner have become one unit. However, the preheater has a specific task and is not
principally connected to the precalciner.
Modern low pressure drop cyclones are the result of a development which started in 1932.
Shaft Preheaters:
Counter-current heat exchange
Limited production (around 1000 t/d)
Disappointing heat exchange mainly due to poor meal distribution
Hybrid preheaters:
Combination of shaft and cyclone stages
Bhler-Miag, Polysius, Prerov, Humboldt
Shaft stages often replaced by cyclone stages
Cyclone preheaters:
Co-current heat exchange
Successful concept, predominantly used
Industrial installations of cyclone preheaters:
Precalciner kiln with 4 to 6 cyclone stages (contemporary technology):
Large capacities possible > 10000 t/d
Up to 4000 t/d in 1 string
Heat consumption < 3000 kJ/kg possible (6 stages)
4-stage cyclone pre-heater kiln (standard technology 1970 to 1980):
Cyclone for raw meal preheating
Large application world wide
Capacities of up to 4500 t/d technically possible
Heat consumption: 3150 to 3350 kJ/kg cli
2-stage cyclone pre-heater kiln:
Less sensitive to circulating elements than 4-stage pre-heater
Higher heat consumption than pre-heater with more stages
Most recent innovations:
Horizontal cyclone for low profile preheaters (Polysius)
Dip tube add-on RTS for 30% lower cyclone pressure drop

1. GENERAL
1.1 History
With dry process, the heat exchange for heating up and calcination takes place between hot
kiln gas and dry powder. Since the high dust losses from long dry kilns made it almost
impossible to achieve acceptable heat consumption: other heat exchange principles had to
be applied.
Since the temperature range to be covered is below 1000C, where the meal behaves
normally like dry powder, stationary reactors where the meal is in suspension with the hot
gas can be used.
The first patent for a suspension preheater using four co-current cyclone stages was applied
for in 1932 and issued in 1934 by the patent office in Prague to a Danish engineer employed
by FLS. Even though the concept was entirely described in the patent, it took another 20
years for industrial application in 1951 by the company Humboldt, now KHD.
Other developments using shaft stages have been abandoned and today, a suspension
preheater is actually a cyclone preheater.
1.2 Trend
All new kiln systems and the majority of the ones with start-up date after 1970 are equipped
with cyclone pre-heaters. Gradually, older plants with wet kilns or long dry kilns are shut
down for good due to their age as well as their high specific production cost
The portion of worlds cement produced with kilns using suspension pre-heaters is still
growing, as can be seen by the development of the Holcim plants. It looks as if it will exceed
95% one day because no feasible alternative solution changing this development is in sight.
In combination with pre-calciners, units of 10000 t/d have been built using up to four strings,
five stages. Typically, 3500 t/d can be handled in one string, in a recent project even 4000
t/d have been proposed.
Figure 1:
Kiln Systems in the Holcim Group II
140
120
100 Long wet
Long semi wet
80 Long dry
Grate preheater
60 SP
SP-PC
40
20
0
1981 1983 1985 1987 1989 1991 1993 1995 1997

Figure 2:
CaCO free
CaCO 3 C3S
C2 S
Surrite
C2AS
Quarz C3A C3A
Clays Liquid phase

CA C12A7
C2F C4AF C4AF
400 600 800 1000 1200 1400 1400 1200C

Heating Calcining Sintering Cooling
Holcim Cement Manufacturing Course Kma 24.2.98
CC-98051.dsf

2. HEAT EXCHANGE IN A SUSPENSION PREHEATER
2.1 Counter-Current Heat Exchange (Shaft Stage)

The most efficient type of heat exchange is the counter-current principle. The flows of the
heat releasing media and the heat absorbing media are in opposite directions. This provides
optimum the temperature difference (=temperature gradient, in theory allowing almost
complete heat exchange.
In case of a suspension preheater, where powder is suspended in a gas, the heat exchange
takes place in a reactor vessel where the hot gas enters from below and leaves at the top.
The meal to be preheated is fed at the top. The meal retention time depends on distribution
across the gas flow and the retention time, which is determined by the gas velocity.
In industrial installations, the heat exchange proved to be far below expected, because even
distribution of the meal was not achieved, particularly not with large units.
2.2 Co-Current Heat Exchange (Cyclone Stage)

Co-current heat exchange takes place if both heat exchanging media flow in the same
direction. Because of the rapidly decreasing temperature difference, the meal can never
reach gas inlet temperature.
Good and reproducible results in industrial installations with this type lead to the
predominance of this principle in the cement industry. The heat exchanger is a gas duct with
velocities from 10 to 20 m/s, equipped with good meal dispersion devices. The purpose of
the cyclone is primarily to separate meal from gas, and not to exchange heat!
2.3 Thermodynamic Limits

Regardless of the type of heat exchange, there is always a thermodynamic imbalance
between hot gases from kiln and calciner and cold raw meal. The heat contained available in
the hot gas leaving the rotary kiln exceeds the heat required for heating the meal to the
temperature levels required for calcination.
Another limit must be observed: Because the temperature gradient between gas and meal (T
gas > T meal) must always be maintained, a higher calcination degree than 30% cannot be
achieved without additional heat input.
The following heat balance estimate shall illustrate this:
Heat contained in the gas: 2300 kJ/kg cli
(1100C; 1.3 Nm3/kg cli)
Heat to preheat meal to 850C: 1300 kJ/kg cli
(1.6 kg meal /kg cli)
Heat required for 30% calcination: 650 kJ/kg cli
Rest (ideal heat exchange): 350 kJ/kg (corresponding to 200 C)
This shows that even if the heat of the gas above 850C is used for partial calcination (about
30%), there is still excessive heat in the gas which would correspond to 200C gas
temperature.
It is obvious that even with a very large number of stages (with accordingly small
temperature gradients), there will always be excess heat! This waste heat is lost only for the
kiln system, but not for the plant, since it can be used for raw material drying in the mill.

Figure 3:

3. PREHEATER TYPES
3.1 Preheaters with Shaft Stages

The rather disappointing performance of the shaft stage made it virtually disappear from the
market. Many hybrid preheaters were equipped with one or two cyclone stages replacing the
shaft stage.
Shaft stages at the kiln inlet have the advantage to be less sensitive to build-ups. This could
be an advantage in cases where elevated sulfur input in the kiln system must be expected.
Several Suppliers built preheaters using shaft stages. Two groups can be distinguished:
3.1.1 Pure shaft preheaters:

Polysius: GEPOL
Self-supporting structure (no tower required)
Vertical tube with restrictions
For small capacities (up to ca. 1000 t/d)
ZAB Dessau: Some applications in Eastern Europe

Similar to GEPOL, but not self-supporting
The Deuna plant had originally 4 ZAB shaft preheaters
3.1.2 Hybrid preheaters:

Several suppliers used a combination of shaft and cyclone stages:
Polysius: DOPOL preheater (first generation)
The central swirl-pot (second lowest stage) was shaft stage
Replaced by DOPOL 90 from 1990
Gradually developed into a cyclone preheater
Up to ca. 3000 t/d
Bhler-Miag: Lowest stage was shaft stage

Later often replaced by cyclone stage
Prerov: One large shaft stage with dedusting cyclone

Shaft stage selfsupporting
Additional cyclone stage possible
Separate meal duct to kiln
As sensitive to circulation phenomena as a cyclone type
MBM: Bottom shaft stage with 4 cyclone stages

Only hybrid design still on the market

Figure 4:

3.2 Preheaters with Cyclone Stages
A quasi counter-current heat exchange can be achieved by serial installation of several co-
current stages. The result is the multi-stage cyclone preheater as it is generally applied in
modern cement plants.
In the early years, one and two stage systems have been installed with long kilns, often to
avoid problems caused by circulating phenomena. A large number of plants are equipped
with four stages; the majority of them were built before 1990.
Today, five stage preheaters represent the economical optimum. High raw material moisture
leads occasionally to fewer stages, in combination with low temperature dedusting systems,
or in areas with high fuel cost, six stages can be more economical. Number of stages
depends thus on:
Raw material moisture (i.e. drying heat requirement)
Cost of thermal energy
Cost of electrical energy
Gas handling system (temperature limit, dew point)
Soil conditions (foundations, earthquake zone -> height of structure)
If raw material moisture shows significant seasonal variations, it can be economical to equip
preheaters with variable stages. This is achieved by feeding all or part of the meal to the
second highest stage or by skipping a stage.
Note: Numbering of stages is always from top to bottom: top stage = stage 1.
Exception: Polysius: bottom stage = stage 1

Figure 5:
Temperature C
~ 850C
~ 800C
as
ust g
Exha
al
me
w
Ra
Preheating Calcining
50C
Heat transferred (or heat content) [kJ/kg cli]
2 - Stage cyclone suspension preheater kiln
500C
Feed
Stage 1
Stage 2
To raw mill
950C
Holcim Cement Manufacturing Course Kma 24.2.98
CC-98053.dsf

3.3 Economical Number of Stages for Cyclone Preheaters
For many years, the pressure drop across one preheater stage was up to 1,5 kPa (15 mbar).
The reason for the 4-stage pre-heater being so widely used is, that it represented an
optimum between investment cost (structure height, foundation), pressure drop and heat
consumption. . The performance of comparable systems built in about the same period are
within a relatively narrow range.
About two third of the pressure drop of a stage occurs in the cyclone and depends on its
shape/design and the size, the latter being the determining cost factor.
New cyclone designs are now on the market with only 0,5 to 1,0 kPa (5-10 mbar) pressure
drop per stage. Considering increasing energy cost, it is justified to install 5 or 6 pre-heater
stages for new or modified kiln systems.
The following table indicated the estimated effect of a 5th and a 6th cyclone stage:
4 to 5st 5 to 6st
Heat consumption kJ/kg cli - 80 - 50
Exhaust gas temperature C - 40 to -50 - 20 to -30
Exhaust gas quantity Nm3/kg cli - 0,03 - 0,015
Drying capacity in RM % H2O from 8 - 6,5 From 6,5 - 5,5
3.4 Minimum Gas Velocity

Dimensioning of a cyclone preheater is a careful consideration of the importance of
separation efficiency, pressure drop, part load operation capability, size of the preheater and
cost of the project.
It must be mentioned that there is a lowest gas velocity in a cyclone preheater. If operation
results in lower figures, the meal will not be lifted by the gas anymore, resulting in poor heat
exchange and consequently high heat consumption, but also excessive temperatures .
Large dimensions give lower velocities with low pressure drop, but also limit the lowest
possible economical production.

Figure 6: Polysius

Figure 7: FLS

4. DESIGN FEATURES OF PREHEATER-CYCLONES
4.1 General
Modern preheaters are designed for low pressure drop using the new cyclone design which
must still provide good separation efficiency, particularly in the top and the bottom stage.
Cyclone inlet velocities are designed in the range of 10 to 15 m/s.
It has been found that the total pressure drop of one cyclone stage is caused by about 1/3
by the gas duct (i.e. lifting of the meal) and 2/3 by the cyclone. Since not much can be done
regarding lifting of the meal, efforts have been made to improve the cyclone design in order
to reduce total pressure drop: the low pressure (drop) cyclone was designed.
Cyclone design means to optimize between high separation efficiency, low pressure drop
and low cost (i.e. small size).
Other than having the correct design parameters, all stages should be equipped with
Dip Tubes (also called immersion tubes, thimbles or vortex finders)
Meal flaps
Splash boxes (or splash plates).
4.2 Dust Cycles

The entire kiln system is subject to dust cycles. Precondition is gas flow in opposite direction
of pulverized process materials. This causes wear, unnecessary material transport and heat
losses due to heat exchange in the wrong direction.
In the preheater, internal dust cycles due to poor separation efficiency of the cyclones result
in less than optimum preheating of meal. Unfortunately, it is almost impossible to measure
dust loss from lower cyclones in normal operation. The only indicator is the temperature
profiles of gas and meal, but even the meal temperature is not always easy tp measure.
4.3 Features
4.3.1 Splash Box

Early cyclone preheater designs had no splash boxes. Instead, the meal was fed into the
gas at a higher point against the gas flow, creating some turbulence with a certain
distribution effect.
Modern cyclone preheaters must be equipped with correctly designed splash boxes for
optimum meal distribution across the gas duct cross section. The principle is based on
impact on a plate. In some installations, the bottom plate of the splash box can be adjusted.
Note: No splash box must be installed at the kiln inlet! The hot meal from the bottom cyclone
must enter the rotary kiln as smoothly as possible. Meal is easily picked up by the kiln gas
and will create a dusty transition chamber.

Figure 8:

4.3.2 Dip Tube (Immersion Tube, Vortex Finder, Thimble)
This integral element of the cyclone has a decisive influence on separation and pressure
drop. It makes the gas to follow a 180 to 360 rotation thus creating the desired centrifugal
force for the separation effect.
In the colder upper stages (stage 1 to 3) it can be designed as simple extension of the outlet
gas duct, made from steel plate. These upper stage dip tubes create usually no problems
except when the preheater gets overheated, e.g. during start-up. Then, the dip tube can
collapse, causing excessive pressure drop.
In the hotter lower stages, mild steel ducts from one piece are not suitable. Several
segmented designs made from heat resistant steel or ceramic material (Hasle) are available
on the market. It is standard today that all stages are equipped with dip tubes.
Note: It appears that some designs of segmented dip tubes have a tendency to unhook
enabling individual elements to drop and to block the cyclone outlet!
For older plants, installing a segmented dip tube in the lower stages is a optimization
possibility which is often applied.
4.3.3 Meal Flap

In order to understand the purpose of the meal flap, the following two aspects must be
mentioned:
There is a pressure difference across a cyclone stage, i.e. between two subsequent
cyclone gas outlets (maintained by the ID fan).
Without meal, there are two ways the gas can flow from one stage to the next: through
gas duct and through meal duct
If there was an ideal kiln system, i.e. a system with 100% constant meal flow and never
changing operation parameters, the meal duct diameter could be designed for just the meal.
The meal would then fill the entire cross section, leaving no opening for the gas. In reality,
there are fluctuations of meal and dropping build-ups, requiring oversized meal ducts.
It is the purpose of the meal flap to close the free cross section not used by the meal, to
avoid gas bypass. There are designs that open only when a certain weight pushes them
open, thus creating meal fluctuations. Other designs (see figure) are adjustable so that they
move only in case of meal peaks or lumps.
Not operational meal flaps cause heat loss and allow build-up formation in meal ducts
(circulating elements)!
4.3.4 Cyclone Shapes

The separation efficiency of a cyclone gets better with longer dip tube and increasing
distance between swirl (cylinder) and dust collecting cone, i.e. with high and slim shapes.
The top stage of preheaters is designed for high separation efficiency. In order to save
height, most suppliers install twin cyclones with the drawback that meal and gas have to be
split. There are a few plants from FLS with only one top cyclone, avoiding this drawback.

Figure 9:

5. PREHEATER OPERATION
The performance of a preheater is assessed based on the criteria:
Temperature profile (first indicator: exit gas temperature)
Pressure profile
Oxygen profile
Table Typical Gas Temperature Profiles
4 stages 5 stages 6 stages

SP PC SP PC SP PC
Stage 1 C 350 360 300 310 270 280
Stage 2 C 540 570 490 500 440 460
Stage 3 C 710 740 630 650 580 600
Stage 4 C 840 870 750 770 690 710
Stage 5 C - - 840 870 770 800
Stage 6 C - - - - 840 870
5.1 Operating Problems of Suspension Preheaters

Some reasons for poor preheater performance frequently experienced:
Worn out or non existing immersion tubes (often in bottom stage)
Open inspection doors, leaky gaskets or holes in the pre-heater (cold false air leaks in,
can be detected by hissing sound)
Blocked or non existing meal flaps
No splash boxes (specially older preheaters), combined with not optimum position of
meal feed point (e.g. old DOPOL)
Excessive dust circulation due to poor separation efficiency of cyclones
5.1.1 Circulation Phenomena.

Cyclone preheaters are sensitive to circulation phenomena. Cyclone blockages cause kiln
stops resulting in production loss and dangerous cleaning actions. Possible causes are:
Excessive input via feed or fuel (Cl, S, 1 Na, K)
Chemical unbalance (sulphur, alkali ratio)
Unfavourable kiln/burner operation
Unfavourable design geometry of bottom stage and kiln gas riser duct area

Countermeasures known today allow to solve the problems are:
Change feed composition or fuel quality
Improve burning conditions
Install automatic cleaning (air cannon, big blasters) at critical locations
Change temperature profile by installing a small secondary burner
Install a kiln gas bypass* system
*A bypass system is not desirable since it is expensive and causes loss of heat and material.
It is therefore the last solution left and should be only considered if all other measures are
not sufficient.
The paper circulating phenomena contains more details on this rather complex subject.
6. NEW DEVELOPMENTS
6.1 Horizontal Cyclone

Polysius have developed a horizontal cyclone. (not to be mixed up with earlier designs of
Kawasaki!)
This cyclone is a modified version of the conventional cyclone with the major difference that
the gas outlet is also at the bottom, encircling the meal outlet.
The heat exchanger duct is still from bottom to top, but the stages can be arranged next to
each other instead on top of each other. This allows a significantly lower height of the
preheater structure.
It is expected that savings in civil cost can be achieved. Additional benefit is possible in
cases where the maximum height is restricted (earthquake zones, scenery protection).
Only top cyclones on conventional preheaters are in industrial operation, however. Any other
experience is yet to be made.
6.2 TRS
The Austrian company Zyklontechnik have introduced a dip tube add-on device which will
reduce pressure drop across the cyclone (not the entire stage!) by 30% at otherwise
unchanged performance.
The principle is to avoid the flow around the edge of the dip tube. Instead, the horizontal
swirl from the gas inlet is maintained and can continue into the dip tube through an
accurately shaped slot in the TRS.
Prerequisite is aerodynamically correct cyclone design and very accurate manufacturing of
the TRS, which cannot be made locally.
The device can be mounted to the bottom of a shortened dip tube. If the inspection opening
is large enough, the whole unit can be installed in one piece, otherwise it comes in pieces.
Several TRS are in operation (not in preheaters, however) with performance equal to or
exceeding the predicted figures.

Figure 10:

Figure 11:

Precalcining Systems
U. Gasser
VA 93/4055/E
1. INTRODUCTION ............................................................................................................... 71
2. THEORETICAL ASPECTS OF PRECALCINING............................................................. 72
2.1 Calcining of Raw Meal ........................................................................................ 72
2.2 Combustion in Precalciner .................................................................................. 74
2.3 Specific Heat Consumption................................................................................. 75
2.4 True and Apparent Calcination Degree............................................................... 76
3. BASIC ARRANGEMENTS OF PRECALCINING SYSTEMS ........................................... 77
3.1 AS and AT Systems ............................................................................................ 77
3.2 In-Line, Off-Line and Separate Line Calciners .................................................... 79
4. FEATURES OF PRECALCINERS .................................................................................... 81
4.1 Main Benefits of Precalciner Technology............................................................ 81
4.2 Limitations and Restrictions ................................................................................ 81
4.3 Tertiary Air Damper and Kiln Riser Orifice .......................................................... 83
4.4 Circulation Problems and Bypass with PC Kilns ................................................. 84
5. PRESENT STATE OF PRECALCINER DEVELOPMENT................................................ 85
5.1 Calciners from FCB............................................................................................. 85
5.2 Calciners from FLS - FULLER ............................................................................ 85
5.3 PYROCLON Calciners (KHD) ............................................................................. 86
5.4 PREPOL Calciners (Polysius) .......................................................................... 86
5.5 Prerov-Calciner ................................................................................................... 86
5.6 Conclusion .......................................................................................................... 87
6. SYNOPSIS OF PRECALCINERS ..................................................................................... 88
7. TEST QUESTIONS ........................................................................................................... 89

SUMMARY
When burning cement clinker in a suspension preheater kiln, about 2/3 of the total heat
consumed or about 2000 kJ/kg are required for the dissociation of CaCO3 also known as
calcination.
The idea of precalcination is, to let this process take place before the meal enters the rotary
kiln by introducing that part of the fuel, i.e. up to 65%, into a stationary reactor.
Because the combustion air (tertiary air) is drawn through a separate duct parallel to the kiln
directly from the cooler, the rotary kiln operates at significantly lower specific thermal load
and gas flow.
The main advantages of precalcination are:
More stable kiln operation due to better kiln control via two separate fuel feed/control
points
More stable kiln operation due to controlled meal conditions at kiln inlet
Reduced thermal load of burning zone
Higher kiln availability
Longer life of burning zone refractories
Larger capacity with given kiln dimensions, resp. smaller kiln for given capacity
Possibility of increasing capacity of existing kilns
More favorable conditions regarding circulating element problems
Allows shorter kilns (L/D <12, 2 supports)
Lower NOx emissions
The drawbacks of higher gas exit temperature after the bottom cyclone and the preheater
higher pressure drop can be compensated by five preheater stages and modern low
pressure drop cyclones.
There are three basic precalciner arrangements available from several suppliers: in-line, off-
line and separate line, all with separate tertiary air duct.
Being the key for complete combustion, the main design criteria is gas retention time: 2 to
3.5 sec minimum, depending on fuel reactivity, 0.5 to 1 sec more for in-line calciners.
Systems where 10 to 20% of the fuel is introduced to the riser duct are considered
secondary firings (SF) and not precalciners.

1. INTRODUCTION
The idea of separating the calcining process from the burning process was already
described in a patent as early as 1912.
However, the first industrial precalciner was built by Humboldt-Wedag (KHD) only in 1966
(Fig. 1). It was the Polysius kiln in Dotternhausen (Germany) which was equipped with a
special 5-stage suspension preheater with extended riser duct. This riser duct had a larger
diameter and the shape of a gooseneck to provide more length thus more gas retention time
enabling combustion of oilshale, a locally available fuel containing raw material. The
combustion air (tertiary air) was still drawn through the rotary kiln. Additional burners were
installed later at the bottom of the precalciner chamber.
Tube type calciners using the gooseneck design are still being used by KHD (Pyroclon) and
Polysius (Prepol).
So it is obvious that the precalciner (PC) kiln was developed from a straight suspension
preheater (SP) kiln. The process characteristics (heat balance etc.) of both SP and PC kiln
systems are quite similar, the main difference being the fact that in case of the PC kiln, 50 to
60% of the fuel (heat) is introduced via a chamber between kiln inlet and bottom cyclone.
This allows to match the process heat requirements more evenly leading to significant
improvements.
Since true precalciners with 50 to 60% PC fuel ratio require a separate tertiary air duct,
almost all PC kilns feature a grate cooler.
The demand for larger and larger capacities which started back in the 1970ies led to a rapid
development of the new precalciner technology. The fastest growing market asking for the
largest units was in Japan where most of the clinker is produced in PC kilns.
During that period, 12 competing suppliers developed their own precalciners, 8 of them were
Japanese (see para synopsis of precalciners).
After the home market for cement plants started to stagnate, the Japanese suppliers
exported their know-how via licenses as well as entire plants. During the late 1980ies, where
only few new plants have been constructed world-wide, the Japanese activities came to a
stop.
The latest development of precalciner technology was aimed at
Complete combustion, also for low reactivity fuels
Suitability for a wide range of fuels
Low emissions of NOx
Since the Japanese competitors have virtually disappeared on the international market, the
variety of precalciner systems is reduced. Five European suppliers (FCB, FLS-Fuller, KHD,
Polysius and Prerov) offer precalciners, some even a choice of alternative solutions.

2. THEORETICAL ASPECTS OF PRECALCINING
2.1 Calcining of Raw Meal

Among all reactions taking place when burning clinker, the calcining - also called
decarbonisation - requires the highest amount of energy: the dissociation of carbonates,
primarily calciumcarbonate according to the reaction
CaCO3 + heat CaO + CO2

in the raw meal requires approx. 1.3 MJ/kg raw meal corresponding to 2.0 MJ/kg cli. The
DTA-curves (Fig. 2) illustrate very well the importance of calcining within the clinker burning
process.
Fig. 2 Differential Thermo-Analysis (DTA)-curves of a typical cement raw meal
During the process of heating up a raw meal, the calcining does not happen suddenly at a
well defined temperature, but starts at about 600 - 700C and ends between 900 and
1000C, following a so-called S curve (Fig. 3). Exact shape and position of this curve vary
from raw meal to raw meal.

Fig. 3 General aspect of the calcining curve of a cement raw meal
Not only the temperature, but also the retention time of the raw meal is an important
parameter of calcining. While the heat transfer from gas to suspended raw meal in a
preheater stage is achieved a fraction of a second, the complete calcination at a
temperature of about 900C in suspension requires a reaction time in the range of 2 to 12
seconds. However, as only 90 to 95% of the calcining should take place in the precalciner in
order to avoid clogging problems, a residence time of about 1 to 3 seconds has proven to be
sufficient.

To perform both above mentioned tasks, i.e. to keep raw meal in suspension for a few
seconds at 850 to 900C in a stationary vessel without clogging, is the common process
target of all PC systems.
2.2 Combustion in Precalciner

The combustion in the precalciner takes place under quite different conditions compared to
the main firing because:
The temperature of the combustion environment is in the order of 850 to 900C (flame
temperature of the main firing: around 2000C).
Some PC systems (in-line systems) use an air-gas mixture for combustion (main firing:
pure primary and secondary air) while others use pure air (off-line and separate line
systems).
In all PC systems preheated raw meal is suspended in the combustion air or air-gas
mixture respectively in order to absorb the heat released thereby maintaining the
temperature at a comparatively low level. By all means must Sintering of material
avoided, as this would lead to clogging in the precalciner stage.
On the account of the less favorable combustion conditions complete combustion is not
always readily obtained, it requires a certain experience to achieve optimum performance.
Of the various parameters influencing the combustion performance, the following are
perhaps the more important ones:
Good mixing of the fuel with the available oxygen. (This is particularly difficult to achieve
with in-line calciners.) Optimum fuel dispersion into the gas flow (liquid fuel: atomization)
is essential.
Retention time for combustion has to be sufficient. The combustion must be completed in
the PC stage. Otherwise, it will continue in the next stage (post-combustion) where the
temperature level is lower and therefore less favorable for the calcination (see S-curve).
This results in not optimum utilization of the heat which leads eventually to higher fuel
consumption.
The flow pattern of the air/gas mixture (resp. tertiary air) has to be favorable for the
combustion.
The meal distribution in the combustion zone has to be optimum, i.e. causing minimum
distortion of the combustion. (CaCO3 as well as CO2 can also react with C - carbon from
the fuel - to produce CO!).
It is known from experience that too high concentrations of raw meal can seriously impede
the complete combustion.
With the introduction of the separate air duct for the combustion air for the calciner, the new
term of tertiary air had to be introduced:
Primary air: Air introduced via kiln burner

Secondary air: Air from cooler to kiln burning zone
Tertiary air: Air from cooler to PC for combustion

Introduction of fuel between kiln inlet and bottom cyclone - as secondary firing or via
precalciner - necessarily increases the temperature level. The gas exit temperatures from
the lowest stage of a straight preheater kiln is only 790 to 820C as compared to precalciner
kilns where this temperature increases by some ten degrees to 840 to 870C. Therefore, the
preheater exit temperature is also somewhat higher entailing an increased heat loss, which
is more pronounced with 4-stage preheaters.
The performance of PC systems can primarily be judged on two characteristic values:
The temperature difference between gas and material ex precalcining stage should be
as low as possible, so as to minimize the heat losses of the exit gas. The reaction
temperature in the precalciner depends of course on the raw meal and the required
precalcining degree as well as tolerated NOx level.
Complete combustion must be achieved as this directly influences the overall heat
consumption of the system. It must be mentioned that this is strongly influenced by the
excess of air.
Note: Stating the amount of unburned matter in the gas is therefore only meaningful
to assess a calciner system, if the amount of oxygen in the gas is indicated as well.
Solid, liquid and gaseous fuels are successfully fired in PC kiln systems. However, the
location and position of the burners in the precalciners have to be adapted to the fuel
particularities. This is specially important for gaseous fuels, which seem to be more difficult
to burn in the PC chamber than other fuels.
2.3 Specific Heat Consumption

From the above mentioned it can be concluded that PC systems have a tendency to slightly
increased heat consumption, unless countermeasures are taken such as:
Although equipping existing preheater kilns with precalcination usually results in a slight
increase of the heat consumption, the average (annual) heat consumption may be equal
or even lower on account of a more regular kiln performance.
Also for new installations the heat consumption is about 50 - 100 kJ/kg cli higher than for
conventional preheater kilns with 4 stages.
Where the somewhat higher exit gas temperature cannot be fully used, say for raw material
drying - then it has become standard to install one or two additional preheater stages to
reduce the heat consumption to a figure slightly, for 6 stages noticeably, below that of a
conventional 4st SP kiln. The first PC kiln in Dotternhausen was in fact equipped with a 5-
stage preheater.

2.4 True and Apparent Calcination Degree
An important parameter for controlling the precalciner operation is the calcination degree. It
is important to know the meaning of the true and the apparent calcination degree.
True calcination degree:
Degree to which the calcination is completed, i.e. extent to which the CO2 is dissociated from
the CaCO3.
Extremes: Raw meal 0% (LOI = 35%)
Clinker 100% (LOI = 0%)
In reality, the calcination degree is determined using a hot meal sample taken from the meal
duct of the bottom cyclone. Because of always present dust cycles between kiln / kiln inlet /
kiln riser / bottom cyclone, this sample contains a certain amount of dust which was already
in the kiln calcining zone and is higher or even fully calcined. This sample is therefore a
mixture consisting of fresh meal and dust circulated back and has a higher calcination
degree than the pure fresh hot meal.
This means: The higher the dust concentration near the kiln inlet resp. the dust cycle, the
higher the apparent calcination degree.
Apparent calcination degree:
The calcination degree determined from a hot meal sample taken from the meal duct of the
bottom cyclone.
Fig. 5 True and Apparent Calcination Degree, PC Fuel, Dust

3. BASIC ARRANGEMENTS OF PRECALCINING SYSTEMS
3.1 AS and AT Systems

The first precalciner in Dotternhausen used combustion air which was drawn through the kiln
as excess air. This technology was maintained for several years and is known as AT
System. However, in reality only up to 35% fuel could be introduced to the precalciner thus
limiting its benefits. The AT type is thus no longer considered a precalciner; it is rather used
for secondary firings where a high calcination degree at the kiln inlet is not the main target.
Fig. 7 AS and AT Systems

Today, all precalciners are AS Systems using tertiary air which is extracted from the kiln
hood or from the cooler roof and drawn via a separate tertiary air duct parallel to the kiln to
the precalciner. This means that planetary coolers are not compatible with precalcination
technology (i.e. AS systems).
Table 1 Comparison of AS and AT System

Item AS AT
Portion of fuel to the up to 65% max. 35%
precalciner
Largest kiln in operation 8500 t/d, 6.2 x 105 m 4700 t/d, 5.2 x 80 m
Kiln for given capacity (st SP approx. 75-80% approx. 85-90%
= 100%)
Suitable type of cooler only grate or rotary all types
Suitable for extension of poor (cooler, tertiary air very good for low PC rates
existing SP kiln duct)
Burning conditions in rotary normal flame temperature lower flame temperature and
kiln (normal excess air) stable operation due to high
excess air
Thermal load in burning zone approx. 60-70% at 60% approx. 85-90% at 30% PC
(4st SP = 100%) PC
Behavior regarding circulating like 4st SP kiln due to the high O2-content of
elements the kiln atmosphere, reduced
volatility of sulfur and
therefore decrease of
encrustation in transition zone
and riser pipe
Heat loss at 10% bypass (4st approx. 40% (bypass will approx. 90% (bypass will have
SP = 100%) be smaller than in 4st SP same size as in 4st SP kiln)
kiln)
Exhaust gas temperature (4st higher than 4st SP higher than 4st SP
preheater)
Heat consumption slightly higher than 4st SP slightly higher than 4st SP
Pressure loss over preheater higher than 4st SP slightly higher than 4st SP

3.2 In-Line, Off-Line and Separate Line Calciners
This criteria refers to the position of the precalciner in the kiln system installation and is
illustrated with Fig. 8 below.
In-Line Calciners are installed in the kiln exhaust gas flow which means that the
combustion takes place in an air/kiln gas mix. This precalciner can be considered an
enlarged kiln riser duct.
Off-Line Calciners are installed off the kiln exhaust gas flow. The combustion takes place
in pure (tertiary) air which is also responsible for lifting up the meal.
Separate Line Calciners are off-line calciners with a separate preheater string.
Fig. 8 Precalciner Arrangements

Table 2 Comparison of Calciner Arrangements
In-Line Off-Line Separate Line

PC arrangement Extended riser duct Parallel to riser duct Parallel to riser
duct
Combustion Kiln gas and air mix Pure air Pure air
atmosphere
Preheater string 1 to 4 of same size 1 to 4 of same size 2 to 4, 2 different
sizes
Advantages Low NOx version Suitable for Two independent
(reducing kiln NOx) modification combustions
Easy combustion
control
Excess air used for Good combustion Good combustion
combustion
Suitable for lump Suitable for Suitable for
fuel modifications modifications
Weak points Mixing of air with Higher peak Higher peak
gas temperature (NOx!) temperature
(NOx!)
Larger volume PC drop-out can fill PC drop-out can
required TAD fill TAD
Incomplete Asymmetry
combustion regarding
circulating
elements
Height requirement Requires 2 strings
(depending on (not feasible for
type/design) <3000 t/d)
Strings of different
sizes (problem
>7000 t/d)

4. FEATURES OF PRECALCINERS
4.1 Main Benefits of Precalciner Technology

There are many advantages of precalciner technology which made it state of the art today.
Some of them are listed here:
1) More stable kiln operation due to better kiln control via two separate fuel feed/control
points.
2) More stable kiln operation due to controlled meal conditions at kiln inlet.
3) Reduced thermal load of burning zone.
4) Lower brick consumption as a result of 1. and 3.
5) More than double capacities possible with given kiln (10000 t/d with 6 m x 95 m kiln).
6) Possibility of increasing capacity of existing kilns.
7) Reduced volatilization of circulating elements.
8) Reduction of cycles (S, Cl, Na20, K2O) with smaller bypass rate, i.e. lower losses.
9) Makes short kilns possible with 2 stations, L/D < 12
10) Possibilities of NOx reduction.
11) Lump fuel utilization in some cases.
4.2 Limitations and Restrictions

Even though the advantages of precalciner systems are doubtlessly convincing, not all types
can be used in all cases. Limitations are:
Additional installation (fuel dosing, calciner, tertiary air duct) as well as the relatively
smaller rotary kiln sets a lower economical limit to PC systems for new plants at around
1200 t/d.
Alternative fuels containing hazardous components can only be used in the main firing
due to the high temperature level there. The potential to use such fuels is then lower for
PC kilns.
Higher exhaust gas temperature and higher pressure drop can be a drawback in specific
cases.
Separate line calciners for new installations are only feasible if a 2-string arrangement is
required for the capacity, i.e. above 3500 t/d.

Fig. 9 Comparison of wet, SP and PC Kilns (average curves)

4.3 Tertiary Air Damper and Kiln Riser Orifice
Off-line calciners as well as in-line calciners are usually equipped with one kiln ID fan. In
order to allow control of the tertiary air/secondary air ratio, there must be a control device in
at least one of the two gas paths (kiln resp. tertiary air duct).
For efficient warming up of the preheater, a damper is usually installed in the tertiary air duct
to avoid fresh air to bypass the main flame. Very often, this damper is used also, for
controlling the tertiary air flow (Fig. 10a). However, experience shows that high temperature
and clinker dust require a quite refined design of this tertiary air damper. In many cases, this
damper operates only for a short period without problems.
Another approach is to install the control device in the other path. Some suppliers (e.g. FLS
and Kawasaki) have developed a kiln riser orifice which is successfully operating in several
plants. This solution (Fig. 10b) is generally more expensive than the TA damper above, but
performs well.
Fig. 10 TA Damper and Kiln Orifice

4.4 Circulation Problems and Bypass with PC Kilns
Precalciner kiln systems have two major advantages regarding circulation problems.
Reduced volatilization in the rotary kiln because less than 50% of the heat is introduced
in the burning zone.
Less than 50% thermal and dust losses in case of a bypass compared to a straight SP
kiln.
The volatilization of circulating elements occurs primarily in the rotary kiln. The percentage of
the volatilized elements which can be extracted with a bypass depends on
volatilization rate in the kiln, and
amount of kiln gas extracted via bypass (= bypass rate) which is expressed by the ratio:
bypass gas
gas at kiln inlet
The highest possible reduction of circulating elements at a given volatilization rate would be
if 100% of the gases at the kiln inlet could be extracted. this is only possible in the case of a
precalciner but not with a straight preheater kiln. Accordingly are the heat losses approx. 50
to 60% lower at a given reduction because the concentration of volatilized circulating
elements in the gas at the kiln inlet is much higher than for a SP kiln.
Fig. 11 Bypass for PC Kilns

5. PRESENT STATE OF PRECALCINER DEVELOPMENT
5.1 Calciners from FCB

FCB have been IHI licensees since the mid seventies for in-line calciners resulting in 8
operating installations and 6 under construction. The highest capacity is 3300 t/d (Tourah,
Egypt).
Together with Ciments Franais FCB have designed a new type of calciner with low
emissions suitable for low grade fuels called the FCB low NOx PC (Fig. 12).
In combination with a low NOx kiln burner, FCB expect to achieve 150 - 350 ppm NOx at the
stack with their new calciner. The first industrial prototype is scheduled for 1992.
The FCB calciner looks like a vertical reactor with one three channel burner on the top.
Tertiary air is introduced from the top as well as with the meal on two sides. Flow is vertical
from top to bottom. Meal can be proportioned via three points on two levels. FCB claim to
achieve:
Hot spot
Reducing atmosphere zone NOx reduction
Controlled flame
No separation of coal and meal
5.2 Calciners from FLS - FULLER

The FLS range of calciners will be marketed by both FLS and FULLER.
Three basic air separate calciner systems are available: ILC, SLC-S and SLC (Fig. 13).
All these use a vessel type calciner which provides retention time by means of volume.
Experiences with this system made no conceptual changes necessary.
The only modification to be mentioned is the new tangential tertiary air inlet for the ILC
system which allows larger calciner volume without requiring more height.
Main features of the FLS calciner systems presently available are:
Variable kiln orifice (Fig. 14) for the SLC-S calciner to control the ratio of secondary to
tertiary air in place of the often troublesome damper in the tertiary air duct.
Low NOx version by splitting the tertiary air creating a controlled area of reducing
atmosphere in the lower part of the PC which is horizontally divided in two zones by an
orifice.
Variation of the calciner outlet temperature with the SLC-S system without changing the
preheater temperature profile providing a temperature window for NH3 injection.

5.3 PYROCLON Calciners (KHD)
The calciner systems by KHD (and Polysius) are based on the 1965 Dottenhausen goose
neck design, a tube type calciner. As PYROCLON-R, a whole range of versions has been
developed (Fig. 15).
A low NOx version of the RP version is not available. KHD tackle the problem of CO from
incomplete combustion with coal firing by focusing on improved coal dosification.
Incomplete mixture of waste gases from kiln and calciner is often found with tube type
calciners. In order to achieve a good mixture, an essential prerequisite for low NOx systems
using excess fuel zones in the precalciner, the 180C elbow is substituted by a new reaction
chamber, called PYROTOP (Fig. 16).A PYROCLON-R Low NOx with PYROTOP allows:
Complete combustion of the calciner fuel
Temperature controlled zones (NH3 injection)
Improved mixing of gases
Reduction of NOx
5.4 PREPOL Calciners (Polysius)

Polysius calciners are all of the air separate (AS) h-line-type. It is generally accepted today
that the calcination process takes place within a few seconds making the fuel reactivity the
decisive design criteria for the calciner size.
The goose neck-tube type calciner PREPOL by Polysius is presently available in three
basic configurations (Fig. 17).
Several Polysius calciners have been modified by the company CLE who added an RSP
type pre-combustion chamber. The same principle is now incorporated in the PREPOL AS-
CC calciner by Polysius.
Polysius started in 1985 to develop their NOx reducing technology called MSC based on
experience available from power stations with staged combustion. They have adapted this
method to the requirements of the clinker burning process. Trial operation on cement plants
have shown 35 - 45% reduction of NOx.
The idea is to create a limited zone of reducing atmosphere near the transition chamber by
adding a small amount of fuel to the rotary kiln exhaust gas via a small burner in the riser
duct. For the NOx from the calciner fuel, the same principle is applied resulting in a second
reducing zone. Such a system would have the following fuel inputs:
< 50% main burner
< 10% via primary DeNOx burner
> 30% via precalciner
< 10% via secondary DeNOx burner
Experience on an industrial scale only will prove the capability of this system. One of the
difficulties is how to control the kiln atmosphere without the gas analysis sampled near the
kiln inlet.
5.5 Prerov-Calciner
The Czek company Prerov have developed a new precalciner (Fig. 18). It consists of a
precombustion chamber (KKS) and a reaction chamber (KKN) with a vortex chamber and is
comparable to Polysius PREPOL-AS CC. During 1992, the first installation will be
commissioned in Southern Italy.

5.6 Conclusion
The development of tube type calciners and vessel type calciners has moved them closer to
each other. The tube type calciners have received a swirl pot or a pre-combustion chamber
for improved mixing and fuel burning and the vessel type calciners have become longer.
The calciner without separate air duct also known as air through actually operating only
with 10 - 20% of the total fuel never fulfilled the expectations and has virtually disappeared,
together with the planetary cooler.
Low NOx calciners have been developed based on the principle of locally reducing
atmosphere by means of fuel excess zones. It can be expected that NOx from precalciner
combustion can be reduced to around 700 - 800 ppm. Calciners can be designed to reduce
NOx generated in the burning zone, or to keep NOx generated in the calciner low, or both.
Since further NOx reduction to lower levels require methods such as NH3 injection,
temperature control is very important.
A modern calciner can be described as follows:
Type: in-line with pre-combustion chamber

Fuel ratio: 50 - 60% (include. low NOx fuel in case of staged combustion
Fuel dosing: low fluctuation
Fuel types: various, including alternative fuels
Combustion environment: pure air or air/kiln gas mix
Calciner size criteria: fuel reactivity
gas retention time (up to 4 - 5 sec.)
Feature: enhanced turbulence
Tertiary air: staged for reducing zone

6. SYNOPSIS OF PRECALCINERS
The different PC systems as well as their developers and suppliers are summarized in Table
3. During the 1970ies the cement manufacturers greatly contributed to the development of
the Japanese PC systems:
Until 1985, ot 304 kilns with PC, 83 were located in Japan, totaling 35% of the capacity. This
shows the explosive expansion of PC systems in Japan back than. Today, all new kilns have
precalciner with tertiary air duct.
Table 3 Synopsis of PC Systems

Trade Name Signification Developer & Licenser Plant Supplier & Licensee
PASEC Voert Alpine / SKET ACT
SLC Separate Line Calciner F.L. Smidth F.L. Smidth
SLC-S Separate Line Calciner Special
ILC In-Line Calciner
ILC-D In-Line Calciner Downdraft
ILC-E* In-Line Calciner, Excess Air
Prepol AS Air Separate Krupp-Polysius Krupp-Polysius
Prepol AS-CC Controlled Combustion
Prepol AS-MSC Multi Stage Combustion
Prepol AT* Air Through
Pyroclon R Regular = Air Separate KHD Humboldt Wedag KHD Humboldt Wedag
Pyroclon RP Regular Parallel
Pyroclon R Low NOx
Pyroclon R Low Nox
with Pyrotop
Pyroclon S* Special = Air Through
EVS-PC Echangeur voie sche avec Fives-Cail Babcock Fives-Cail Babcock
(only fuel - oil) prcalcination
KKS-KKN n.a. Prero Prerov
SF Suspension Flash Calciner Ishikawajima-Harima Heavy Ishikawajima-Harima Heavy Ind.
Ind. Fuller Company /
Chichibu Cement Fives-Cail Babcock
NSF New SF
RSP Reinforced Suspension Onoda Cement Onoda Engineering & Consulting
Preheater Kawasaki Heavy Industries
Allis-Chalmers
CLE-Technip
KSV Kawasaki Spouted Bed and Kawasaki Heavy Industries Kawasaki Heavy Industries
Vortex Chamber
NKSV New KSV
MFC Mitsubishi Fluidized Calciner Mitsubishi Mining & Cement Mitsubishi Heavy Industries
GG Reduction Gas Generator Mitsubishi Heavy Industries
DDF Dual Combustion and Denitration Nihon Cement Kobe Steel
Furnance
CSF (CFF) Chichibu Suspension Flash Chichibu Cement Chichibu Cement (own plants)
Calciner
SCS Sumitomo Cross Suspension Sumitomo Cement Kawasaki Heavy Industries
Preheater and Spouted Furnace Ishikawajima-Harima Heavy
Process Industries
*Air through: secondary firing systems

7. TEST QUESTIONS
1) Which is the chemical reaction with the highest heat consumption within the clinker
burning process? How much does it consume in absolute terms (kJ/kg clinker) and in
percent of the total heat consumption of a modern kiln system?
2) Which are the three basic precalciner arrangements and what are their differences?
3) At what temperature does the calcination take place and how much CO2 is totally
dissociated from the CaCO3?
4) Which are the benefits of precalciner technology?
5) Which is the most important design criteria for precalciner dimensioning?
6) Explain the term apparent calcination degree. How can it be determined and what is its
significance?
7) How do the effects of a bypass compare in case of a straight preheater kiln and a
precalciner kiln?

Fig. 1 Sketch of Dotternhausen Kiln, the first Precalciner (KHD, 1966)

Table 4: Temperatures and Process Steps for Clinker Burning
Temperature [C] Process Step, Type of Reaction Heat

20 - 100 Evaporation of free H2O Endo
100 - 300 Loss of physically absorbed H2O Endo
400 - 900 Removal of structural water Endo
> 500 Structural changes in silicate minerals Exo
600 - 900 Dissociation of CO2 from CaCO3 Endo
> 800 Formation of intermediate products Exo
Belite, Aluminate and Ferrite
> 1250 Formation of liquid phase (aluminate Endo
and ferrite melt)
Formation of alite Exo
1300 - 1240 Crystallization of liquid phase into mainly Exo
aluminate and ferrite
For numerical calculations, an approximate quantity of CO2 from the raw material
(dissociated from the calcites) can be used, regardless of the exact chemical composition.
CO2 from raw mat = 0.28 Nm3/kg cli

Table 5: Energy Balance of Process Steps for Clinker Burning
Endothermic Processes: kJ/kg cli kcal/kg cli

Dehydration of clays 165 40
Decarbonisation of calcite 1990 475
Heat of melting 105 25
Heating of raw materials (0 to 1450C) 2050 490
Total endothermic 4310 1030
Exothermic Processes: kJ/kg cli kcal/kg cli
Recrystallization of dehydrated clay 40 10
Heat of formation of clinker minerals 420 100
Crystallization of melt 105 25
Cooling of clinker 1400 335
Cooling of CO2 (ex calcite) 500 120
Cooling and condensation of H2O 85 20
Total exothermic 2550 610
Net Theor. Heat of Clinker Formation: kJ/kg cli kcal/kg cli
Endothermic - exothermic 1760 420
Heat consumption of Kiln System: kJ/kg cli kcal/kg cli

Average 4-stage SP system 3300 790
Modern 6-stage SP system 3000 720
Rel. Heat Requirement of Calcination:
Average 4-stage SP system 60%
Modern 6-stage SP system 66%

Fig. 12 FCB Low-NOx Precalciner

Fig. 13 FLS

Fig. 14 FLS Adjustable Kiln Orifice

Fig. 15 Pyroclon

Fig. 16 Pyrotop

Fig. 17 Polysius

Fig. 18 Prerov

Fig. 19 EVS-PC
Features of EVS-PC PC system

Supplier: Fives-Cail Babcock

Fig. 20 SF / NSF

Fig. 21 RSP
Features of RSP PC system

Suppliers: Onoda Engineering & Consulting
Kawasaki Heavy Industries
Allis Chalmers
Creusot - Loire Entreprises

Fig. 22 KSV / NKSV
Features of KSV/NKSV PC system

Supplier: Kawasaki Heavy Industries

Fig. 23 MFC

Fig. 24 GG
Features of GG PC system
Supplier: Mitsubishi Heavy Industries
System abandoned

Fig. 25 DD

Fig. 26 CSF
Features of CSF PC system

Supplier: Chichibu Cement in own plants

Fig. 27 Voest Alpine PASEC System

Fig. 28 FLS

Clinker Coolers
U. Gasser / D. Brassel
PT 97/14232/E (Revision 1, February 1999)
1. INTRODUCTION ............................................................................................................. 113

2. GENERAL CONSIDERATIONS...................................................................................... 113
2.1 Heat Flow in a Kiln System ............................................................................... 113
2.2 Definitions ......................................................................................................... 115
2.3 Calculations....................................................................................................... 117
3. GRATE COOLERS ......................................................................................................... 120
3.1 The Reciprocating Grate Cooler ....................................................................... 120
3.1.1 Principle......................................................................................................... 120
3.1.2 History ........................................................................................................... 122
3.1.3 Conventional Grate Coolers (1980s) ............................................................ 123
3.1.4 Typical Grate Cooler Problems ..................................................................... 126
3.1.5 Modern Grate Coolers (1990s) ..................................................................... 128
3.1.6 Design Highlights of Modern Grate Coolers .................................................. 131
3.1.7 Clinker Crushers............................................................................................ 139
3.1.8 Cooler control ................................................................................................ 143
3.1.9 Cooler Dedusting........................................................................................... 146
3.1.10 Developments................................................................................................ 148
3.2 The Cross Bar Cooler ....................................................................................... 150
3.2.1 Principle......................................................................................................... 150
3.2.2 Main features................................................................................................. 150
3.2.3 Strengths and Weaknesses........................................................................... 152
3.3 The Travelling Grate Cooler.............................................................................. 154
3.3.1 Principle......................................................................................................... 154
3.3.2 Strengths and Weaknesses........................................................................... 156
4. ROTATING COOLERS ................................................................................................... 158
4.1 The Rotary Cooler or Tube Cooler.................................................................... 158
4.1.1 Principle......................................................................................................... 158
4.1.2 Design Features ............................................................................................ 158
4.1.3 Cooling performance ..................................................................................... 160
4.1.4 Strengths / Weaknesses................................................................................ 160
4.2 The Planetary Cooler ........................................................................................ 161

4.2.1 Principle......................................................................................................... 161
4.2.2 Historical........................................................................................................ 162
4.2.3 Design features ............................................................................................. 163
4.2.4 Internal heat transfer equipment (see Fig. 26) .............................................. 163
5. VERTICAL COOLERS .................................................................................................... 166
5.1 The Gravity Cooler (G - Cooler) ........................................................................ 166
5.2 The Shaft Cooler ............................................................................................... 166

SUMMARY
Clinker coolers have two tasks to fulfil:
Recuperate as much heat as possible from the hot clinker by heating up the air used
for combustion
Cool the clinker from 1400C to temperatures adequate for the subsequent process
equipment, normally to 100 - 200C.
There are mainly two different types of clinker coolers in operation with the following
features:
Grate coolers
Crossflow heat exchange through horizontal clinker bed with cold air from below.
Cooling airflow exceeding combustion air requirement allows low clinker temperatures,
but necessitates excess (waste) air dedusting.
Modern cooler technology with sophisticated plates and forced aeration systems allow
combustion air temperatures exceeding 1000C.
Trend to wider and fewer grates, less cooling air and fixed inlets
Largest units: 10000 t/d
Travelling grate (Recupol): last unit built around 1980
Rotating coolers
Rotary tube coolers with separate drive or planetary cooler attached to kiln shell
Quasi counter-current flow heat exchange
Cooling air determined by combustion air, no waste air
Heat exchange (recuperation) determined by condition of internal heat transfer
equipment
Limited unit size, up to 3000 t/d
Planetary cooler not suitable for precalciner technology
Practically no new installation built anymore

1. INTRODUCTION
The clinker cooler is a vital part of the kiln system and has a decisive influence on the
performance of the plant. Three key indicators characterize a good cooler:
Maximum heat recuperation
Minimum cooling air flow
Unrestricted availability
There have been periodic changes in trends during the past decades. Grate coolers were
first introduced by Fuller Company (USA) around 1930. While its design was continuously
being optimized, the grate cooler became the predominant type in the 1950's. In the late
1960's, the planetary cooler gained popularity which reached its peak in the 1970's, mainly
due to its simplicity. Larger unit capacities with precalciner technology made the grate cooler
the preferred solution again. A wave of grate cooler reengineering starting in the mid 1980's
has generated a much improved grate cooler technology as well as a new situation on the
suppliers' side. New problems were experienced and have been or are being solved.
Since cement plants have life cycles of 40 years and more, numerous units of each cooler
type, planetary, rotary or grate cooler of old or new designs, will remain in operation for
many more years.
2. GENERAL CONSIDERATIONS
The clinker cooler has the following tasks to fulfil:
Process internal heat recuperation by heat transfer from clinker to combustion air
Reduce clinker temperature to facilitate clinker handling and storage
Provide maximum cooling velocity to avoid unfavorable clinker phases and crystal size
2.1 Heat Flow in a Kiln System

The importance of the cooler as a heat recuperator can be well demonstrated with a heat
flow (Sanki) diagram.

Figure 1 Clinker cooler and kiln system
Figure 2 Energy turnover (Grate cooler)

2.2 Definitions
As for other components of the kiln system, specific figures for clinker coolers refer to
1 kg of clinker. This eliminates the influence of plant size and allows direct comparison
of clinker coolers of different types and sizes.
Cooling air is the air which passes the clinker thus being heated up while cooling the
clinker. It corresponds approximately to the combustion air requirement, only grate
coolers allow additional air for better cooling.
Primary air is the air which is required for proper functioning of the burner. Ambient air
insufflated by a separate small fan plus the air from a pneumatic transport system,
amounting from < 10% up to > 30% of the air required to combust that fuel. Some
precalciner burners are equipped with primary air fans (for cooling) as well.
Secondary air is the hot air entering the rotary kiln via clinker cooler. Its flow is
determined by the combustion of the burning zone fuel. While cooling the clinker, it
reaches temperatures of 600 to over 1000C, depending on type and condition of the
cooler.
Tertiary air is that part of the combustion air which is required for combusting the
precalciner fuel. It is extracted from kiln hood or cooler roof, and then taken along a duct
(=tertiary air duct) parallel to the kiln to the precalciner. It reaches temperatures near or
equal to the level of the secondary air.
Middle air (grate cooler only) is extracted from the cooler roof if drying of process
materials requires a temperature level which is higher than the waste air. If the quantity
is small, up to 450C can be expected at normal cooler operation.
Waste air (grate cooler only) is also called cooler exit air or cooler excess air. The total
cooling airflow from the fans is normally higher than the flow required for combustion
(=tertiary + secondary air). The extra air, which has normally a temperature of 200 to
300C, must be vented to ambient via a dedusting system.
False air is cold air entering the system via kiln outlet seal, burner opening, casing or
clinker discharge. It either dilutes secondary air thus reducing recuperated heat or adds
load to the waste air system of grate coolers.
Specific air volumes are airflows per kg of clinker (m3/kg cli, Nm3/kg cli). Independent of
the kiln size, airflows of cooler systems can be directly compared.
Specific loads express the relation of clinker production to a characteristic dimension of
the cooler (t/d m, t/d m2, t/d m3). Exact definitions vary with cooler type.
Radiation losses from the cooler casing/shell are particularly important for planetary
coolers, where they actively support the cooling of the clinker.
Efficiency expresses the quality of heat transfer from clinker to the air which is used for
combustion in the burning zone and precalciner firing.
Remark: Since the heat recuperated is proportional to hot air used for combustion and
temperature, an efficiency figure is only meaningful if it is related to a heat
consumption figure (resp. a combustion airflow).

Figure 3 Clinker coolers - Definitions

2.3 Calculations
The calculations below are examples of heat balance investigations:
Heat in hot clinker Qcli :

Qcli = mcli* cpcli* (tcli - t ref) Example with mcli =1 kg/h:
tcli = 1400C:
Qcli = 1 kg/h * 1.090 kJ/kgC * (1400C-20C) = 1504 kJ/h
Heat in hot air Qair :
Qair = Vair* cpair* (tair - t ref) Example with V air = 1Nm3/h:
tair = 1066C:
Qair = 1 Nm3/h * 1.421 kJ/Nm3C * (1066C-20C) = 1486 kJ/h
Radiation loss Qrad :
Q rad =CR * * A {(t/100)4 - Grate cooler
4
(t0/100) } Qrad = 20 kJ/kg cli (from experience)
Cooler efficiency cooler
cooler =
Q combustion air
= 1
Qloss
Q clinker from kiln Q clinker from kiln
The secondary (+ tertiary) air requirements are dictated by the amount of fuel fed to the
burners. Per this definition, the efficiency of a cooler is getting better with increasing kiln heat
consumption. It is thus obvious that a cooler efficiency figure is only meaningful if the
corresponding heat consumption (or airflow) is indicated.

Example: production 5000 t/d
heat consumption 3000 kJ/kg cli
secondary and tertiary 1066C
air temperatures
Primary air main burner 10%
PC fuel ratio 60%
False air and excess air neglected (not realistic!)
Q comb air:
V Comb air = 3000 MJ/kg cli * 0.26 Nm3/MJ * 5000/24*103 kg/h * (1-0.4*0.1)
= 156'000 Nm3/h
t comb air = 1066C q combustion air = 1.421 kJ/Nm3 * (1066-20) = 1486 kJ/Nm3
Q comb air = V comb air * q comb air = 1486*156'000 kJ/h = 231'816 GJ/h
Q clinker:
m clinker = 5000 t/d /24 h/d *103 kg/t = 208'333 kg/h
t clinker from kiln = 1400C q clinker from kiln = 1.09 kJ/kg * (1400-20) = 1504 kJ/kg
Q clinker = 208'333 * 1504 kJ/kg = 313'333 GJ/h
Efficiency = 231'816 / 313'333 * 100% = 74.0%

Figure 4 Clinker cooler typical data (4-stage SP Kiln, 2000 t/d)

3. GRATE COOLERS
3.1 The Reciprocating Grate Cooler

The reciprocating grate cooler is the most widely applied type and is exclusively used for
new plants.
3.1.1 Principle
The following major system components can be distinguished:
Casing with kiln hood and connections for air at different temperature levels
Reciprocating grate with drive system
Aeration system with fans, undergrate compartments and direct air ducts
Riddling (= fall through) extraction system with hoppers, gates and transport
Clinker crusher
Material transport
The clinker is pushed by the vertical part of the front edge of the preceding plate. The
entire grate consists of a combination of fixed and moving rows which results in a quasi-
continuous motion of the clinker bed.
Heat exchange
Heat exchange from clinker to air is according to the cross current principle. The
cooling air penetrates the clinker bed which is laying on the grate from underneath and
leaves it at the surface. While passing through the hot clinker, the air is accumulating
heat which is transferred from the clinker.
Cooling air
Normally, ambient air is blown to underneath of the grate plates loaded with clinker by a
number of cooling air fans. Delivery pressure must be sufficient to penetrate the clinker
bed and to compensate for the expansion (increase of actual volume) of the air from
heating it up
Under ideal conditions, the required cooling air depends directly from the desired clinker
temperature. One part of the cooling air is used for combustion in the kiln, the rest is
cleaned and vented to ambient, unless it is further used, e.g. for drying.
Cooling curve
A simplified mathematical model for clinker cooling in a conventional, optimized grate
cooler gives the relation between cooling air quantity and clinker temperature as follows:
T cli Tamb
= exp[ ( Vair / 0.77)]
Tcli in Tamb
with T cli in = clinker temperature at cooler inlet C
T amb = ambient temperature C
V air = cooling air quantity Nm3/kg cli
The above approximation (curve Fig. 17: Tcli = 1400C) has been found to give satisfactory
results for conventional grate coolers from various suppliers.

Figure 5 Reciprocating Grate Cooler: Design Features

3.1.2 History
It was the Fuller Company (USA) who introduced the first reciprocating grate cooler in the
late 1930's with a grate slope of 15.
Fluidized material running down the grate leads to 10 grate inclination. The 10 cooler was
predominantly used until the mid 1950's. Problems were encountered with those 10 coolers
when the clinker was fine and started to fluidize. As an attempt to solve this problem, wedge
grate plates were used. Another drawback of those 10 coolers was the building height
required for larger units.
In the mid 1950's, the first horizontal grate coolers were introduced. They were initially just
10 grates installed horizontally with accordingly reduced conveying capacity. Some of these
coolers were severely damaged by overheating, due to fluidization and accumulation of hot
fine clinker at the feed end.
This drawback of the horizontal cooler lead to the development of the so-called combi
cooler. Is has one (or formerly two) inclined grates with normally 3 slope, followed by one
or two horizontal grates. Not all suppliers followed the same philosophies, so all three
concepts (all horizontal, combi and all inclined) can be found all over the world.
The planetary cooler boom period in the 1970's came to an end, when large production
capacities were in demand. Precalciner technology required grate coolers which eventually
needed to be reengineered again. Problems related to the clinker distribution, growing
awareness of heat and power consumption as well as the demand for higher availability
forced the suppliers to introduce new solutions. Initiated by the new company IKN, the grate
cooler technology underwent significant changes since the mid 1980's. Modern grate
plates, forced (direct) aeration and better gap design were introduced by all cooler
makers helping to reduce cooling airflow and cooler size.
The new approach lead to better recuperation in most cases. However, serious wear
problems with the new systems forced most of the companies to modify their solutions once
again. Today, in the mid 1990's, we are still gaining experience with latest designs.
The ultimate solution would be the waste air free grate cooler with unlimited flexibility and
availability. However, right now the cement industry would be happy with smooth operation,
high recuperation, low cooling air and no cooler related kiln stops.

Figure 6 Various configurations of reciprocating grate coolers
3.1.3 Conventional Grate Coolers (1980s)
3.1.3.1 Typical Design Features

Grate plates with round holes
Two to three grates, depending on size
Grate slope 0 or 3 or both, depending on supplier
Mechanical excenter drives for reciprocating grate
Chamber aeration
Fan pressure 45 mbar (first) to 25 mbar (last)

Smaller compartments at inlet, larger towards outlet
Clinker riddling extraction with hoppers, gates and dragchain (some earlier designs:
internal drag chain without hoppers)
Hammer crusher at cooler discharge
Worlds largest kilns (10'000 t/d in Thailand) are equipped with conventional grate coolers
from CPAG with 4 grates.
3.1.3.2 Strengths and Weaknesses of Conventional Grate Coolers
Strenghts Weaknesses
Lower clinker end temperature due to Waste air handling system

higher amount of cooling air (dedusting, fan) required
Possibility of adjusting cooling air and More complex cooler requires higher
grate speed provides higher flexibility capital investment
Optimization possibilities during Higher power consumption than
operation planetary or tube cooler
Uneven clinker discharge /
segregation leads to several
problems
Red river
Snowmen
Air breakthrough (bubbling, geyser)
Reduced plate life
Excessive clinker fall through
between gaps
Causes and mechanism of those problems are further explained in the next paragraph.

Figure 7 Conventional grate coolers: Design features

3.1.4 Typical Grate Cooler Problems
Most grate coolers show a tendency to one or more of the system inherent problems, and in
many cases there is no real cure. Investigations of the causes lead to the development of
the modern cooler technology.
Segregation:
Due to its physical properties, the clinker is lifted by the kiln rotation before it is
discharged into the cooler. Installation of the grate axis offset from the cooler axis should
compensate for this effect. However, since discharge behavior of finer and coarser
clinker particles differ from each other, the clinker fractions are not evenly distributed
across the grate. Fines are discharged later and are thus found predominantly on the
rising side of the kiln shell (Fig. 8a).
Thin clinker bed in recuperation zone:
With a conventional grate cooler with chamber aeration, the clinker bed thickness is
limited directly by the installed cooling fan pressure and indirectly by the quality of
compartment seals and distribution of the clinker across the width. In order to avoid
overheated plates, the operator will set the bed not higher than allowed to guarantee
airflow through the plate carrying the clinker with the highest bed resistance.
Thin bed operation leads to unfavorably high air to clinker ratio and poor heat exchange
on the sides with consequently low recuperation efficiency.
Red river:
The infamous red river is one of the most feared problems with grate coolers.
Due to segregation, fine clinker has always its preferred side (see above).
Different bed resistance on either side and only one air chamber across the entire width
often cause fluidization of the fine clinker laying on top. This fluidized clinker does no
longer follow the speed of the grate, but shoots much faster towards the cooler discharge
end. Because the residence time of that fine clinker is much reduced, it does not follow
the general cooling curve and forms a red hot layer on top of the regularly cooled,
already black clinker. Hence the term "red river".
It is not the missed heat recuperation, but the red hot material being in touch with cooler
walls, plates and side seals in the colder area where such temperatures should normally
not occur. Premature destruction of those pieces results in poor availability, high
maintenance and ultimately in loss of production and sales revenues.
Snowman:
The sticky consistence of the hot clinker leaving the kiln combined with the compaction
at the drop point often leads to formation of solid clinker mountains on the grate. Not
permeable for cooling air, they grow larger and disturb the flow pattern of the clinker in
this anyway critical inlet area.
Air breaking through:
Due to the different resistance of the clinker bed and the fear of overheated plates, too
much air is put on the first grate compared to the clinker bed. The result is air shooting
through the bed, hardly taking any heat and thus not contributing to the heat exchange.
In addition to that, the clinker is mixed which can be seen by the bubbling action, and the
layered clinker bed (colder clinker below, hotter on top) is destroyed thus disturbing the
cross flow heat exchange pattern.
The results are low recuperation and too much heat going to the aftercooling zone.

Figure 8a: Segregation at cooler inlet Figure 8b: Clinker bed depth effect on
cooling

Figure 8c Red River Figure 8d Snowman
3.1.5 Modern Grate Coolers (1990s)
3.1.5.1 Design Features

The successful clinker cooler has:
Correct allocation of cooling air to clinker

Sustainable gap widths in the entire cooler
All new or redesigned clinker coolers are aiming at the above two goals:
Modern grate plates, designed to cope with high temperature differences
Inclined inlet section without moving rows
Pattern of zones for individually adjustable aeration in recuperation zone
Modern plates for a tight grate in the after cooling zone
New, improved side seal plate design for tight gaps and low wear
Careful undergrate compartment sealing
Adequate seal air system with correct control
Wider and shorter coolers; lower number of grates
Improved and wear protected moving grate support and guidance
Hydraulic grate drive with optimized control system

Cooling air fans with inlet vane control and inlet nozzle for measuring flow
Roller crusher
3.1.5.2 Strengths and Weaknesses of Modern Grate Coolers

Strenghts Weaknesses
More constant heat recuperation More complicated mechanical
improved, smoother kiln operation installation (varies with supplier)
Cooler inlet: improved clinker Higher secondary air temp. increases
distribution across grate width wear of nose ring and burner
Elimination / control of red river refractories
Significantly reduced grate riddlings Higher actual (m3/h) tertiary air flow
(clinker fall through) can increase dust entertainment at
take off point
Higher waste air temperature (valuable
for drying) Teething problems with new designs -
> design changes still in progress
Lower heat consumption due to higher
heat recuperation (cooler efficiency)
Reduced power consumption due to
less waste air
Lower civil cost due to more compact
cooler
Lower investment due to smaller
waste air system
Reduced cost for maintenance

Figure 9 Modern Grate Coolers: Design features

3.1.6 Design Highlights of Modern Grate Coolers
3.1.6.1 Modern Grate Plates

In the mid 1980's, the first modern grate plates were installed in grate coolers by IKN and
CPAG. They were designed for the following targets:
Allow for lower air/clinker ratio in the recuperation zone for higher recuperation
Improve clinker distribution across the grate width
Assure that all grate plates are always sufficiently cooled by air
The above targets were reached using the following ideas:
Higher built-in pressure drop
Similar to the effect of thick bed operation, a higher pressure drop across the plate
reduces the relative influence of variations in permeability of the clinker bed.
No more fine clinker falling through
Fine clinker falling through means loss of heat and thermal stress on the drag chain.
For forced aeration (below) it is mandatory that no material can fall in the air ducts
where it would cut off the air supply.
Forced (direct) aeration via air ducts
In order to ensure that all plates get enough air, to allow individual allocation of air to
different areas and to avoid that air escapes through gaps, groups of plates are
supplied with air directly via a special duct system
Tight gaps between plates and plates/casing
Not only through the grate surface, but also through gaps between plates within the
same row as well as from one row to the next, fine clinker can fall through. Those
gaps have to be sealed as well, e.g. by interlinked steps in the plate sides (Fuller,
Polysius) or by bolting them together as packages (IKN).
The modern grate plates are the basis of modern cooler technology. Problems experienced
with the first generation of modern grate plates lead to several detail modifications:
Cracks in corners of air outlet openings
Solution: modified shape
Plastic deformation caused premature failure with many designs
Solution: thermally flexible plates built from two or more pieces
Preferred plate internal airflow left plates locally uncooled
Solution: plate internal guide vanes, optimized air channelling
Modern grate cooler, as the IKN Pendulum Cooler, use also Pneumatic Hopper Drains
(PHD) to withdraw the fine clinker fall through.

Figure 10 Modern grate plates

3.1.6.2 Air Ducts
The concept of forced aeration, i.e. the idea to bring the air directly to the grate plates
requires a flexible air connection between the (stationary) fan and the moving rows.
Initially, the most obvious and simple approach was chosen: flexible hoses or bellows. IKN,
CPAG, Polysius and Fuller used this solution at the beginning.
However, experience showed that those hoses were sensitive to design (geometry),
installation and material qualities. While many coolers operated without any problem, others
showed frequent rupture of those hoses, very often causing severe plate damage and
consequently kiln downtime.
Meanwhile, all suppliers developed new solutions. Only KHD avoided these problems by
using telescopic ducts from the beginning.
The individual suppliers are now using the following standard solutions:
Telescopic air connector (BMH-CPAG, KHD)
Ball and socket type air connector (FLS, Fuller)
Gate type air connector (Polysius)
Open air beam (IKN)

Figure 11 Forced (direct) aeration to moving rows: Flexible ducts

3.1.6.3 Aeration Concept
It was soon recognized that only a few (6 to 8) rows of direct and individual aeration are not
sufficient to improve clinker distribution or to eliminate/control red river formation. The
number of rows with direct aeration was gradually increased and soon the suppliers started
to equip the entire recuperation zone or even the entire cooler with direct aeration. Indeed,
this improved the control possibilities, but created the following new drawbacks:
Complicated and expensive equipment
More parameters to control
Difficult access underneath grate
High number of potential problem areas (flexible hoses!)
Ways had to be found to reduce the number of air ducts to the individually aerated cooler
zones. There are two ways to achieve this:
Reduce number of individually aerated zones
Modify the air duct system
Today, the following different solutions with varying degrees of experience are presently
available from the suppliers:
No moving rows requiring flexible air connectors in inlet section
Longitudinal structural beams designed as air ducts
Short air ducts from one moving row to the next (Air bridge)
Direct aeration for fixed rows only (hybrid aeration)
Full chamber aeration with modern grate plates

Figure 12 Aeration patterns

3.1.6.4 Seal Air (Confining Air)
When direct plate aeration was introduced, the significance of the seal air or confining air
was not properly investigated. It was expected that direct individual aeration of the plates
alone would be enough to get the desired improvement due to better air to clinker allocation.
If the cooler grates were tight and had no or very narrow gaps between moving and fixed
rows or between grate and cooler casing, this would indeed be true. However, real grates
have large gaps, which is one of the reasons why direct aeration was introduced.
The effect of insufficient seal air pressure for direct aerated grates can be explained as
follows:
High resistance in clinker bed (bed thickness, kiln upset, granulometry)
Cooling air sneaks around plate edge to undergrate compartment instead
Clinker dust carried in this air abrasion / wear
Gap becomes larger seal air can escape more sneak air
Stops for repair reduce availability and increase operating cost
Today it is generally accepted that partition, sealing and pressurizing of the undergrate
compartments is even more important than with chamber aerated coolers.
Ideally, the partition of the undergrate compartments should repeat the pattern of the
individually aerated grate zones of the grate itself. Since this would lead to very complicated
and expensive designs with difficult access, simpler solutions had to be found.
One of the most common countermeasures is, to install larger seal air fans. It was
interesting to observe the installed cooling air to be gradually increased with each new
project. This did not only lead to larger waste air systems but also to higher cooling fan
motor power which partially offset the savings expected from modern coolers.
The suppliers have proposed the following improvements:
Larger seal air fans
Seal air branched off from cooling air fans
Seal air from booster fan using air from cooling air fans
Undergrate pressure controlled by cooling air fan pressure
Careful sealing of undergrate compartments
No more moving rows in hot inlet zone

Figure 13 Seal air systems
3.1.6.5 Side Seal Systems

Extremely serious wear problems occurred along the side seal plates on each side of the
grate. Excessive fall through along the sides and shockingly short lifetime of the side seal
plates, mainly in the recuperation zone, were the result. The main reasons for this problem
can be listed as follows:
The same seal element used for lateral and longitudinal movement
Side seal plates fixed to cooler casing
Entire thermal expansion to be compensated by (cold) gap on each side
Side plates used for lateral guidance of the grate (older designs)
More lateral thermal expansion of wider grates for large units
The following new solutions have been developed and are now part of the contemporary
standards:
Entirely new side seal plate concepts
Side seal plates bolted to cross beams of fixed rows (no longer to cooler casing)
Joints for thermal lateral expansion and mechanical longitudinal movement between
moving rows and casing separated
Center grate guide for large coolers

Figure 14 Side seal designs
3.1.7 Clinker Crushers

All kiln systems produce larger than normal clinker lumps more or less frequently. Large
balls of material enter the cooler when coating drops during kiln upsets.
Such large clinker masses can only be cooled superficially and contain a lot of heat. Before
being discharged to the clinker conveyor, they must at least be crushed to smaller particles.

All clinker coolers, regardless of the type, are equipped with a clinker crusher. Traditionally,
this is a hammer crusher which has proven to be reliable.
In order to cool large clinker lumps, they must be crushed within the cooler. In reality, this
means installing the crusher before the last grate. Early trials with hammer crushers were
not successful, however.
Based on the idea and experience with roller grate bottoms in shaft kilns (and shaft coolers),
CPAG developed the roller crusher to be used as intermediate crusher in a step cooler.
The advantages of the roller crusher make it also superior at the cooler outlet. Hydraulic or
electric drives as well as different combinations of reversing rollers are available from
various suppliers.
Compared to the hammer crusher, the roller crusher is rated as follows:
Strengths Weaknesses
low speed higher initial investment
low wear chokes easier
low dust generation more difficult to design
equalization of material rushes
suitable for high temperatures
lower power consumption
Figure 15a Hammer crusher

Figure 15b Roller crusher

Figure 16 Heat and air balance of a modern Grate cooler

Figure 17 Optimization
3.1.8 Cooler control

One of the advantages of the reciprocating grate cooler is its high flexibility, due to operating
variables adjustable independently from kiln operation. Usually three main variables are
controlled automatically.
a) Grate speed

In order to prevent the clinker bed resistance from exceeding the pressure capabilities of the
cooling fans (which would mean too little cooling air and danger of heat damage), the bed
resistance on the grate should be kept constant.
To do this, each grate section drive is controlled by the undergrate pressure of the first or
second compartment in each grate section. An increase in pressure indicates an increase in
bed resistance (either more material in the cooler or finer material). The reaction is an
increase of the grate speed, causing the bed to become thinner. If the undergrate pressure
decreases, the drive slows down and the bed becomes thicker.
Another possibility is to control only the first grate by the undergrate pressure, and to keep
the speed of the following grates proportional to the speed of the first grate.
More sophisticated control systems use the weighted average of several undergrate
pressures to control first grate speed. In many cases, however, control systems amplify
fluctuations from the kiln instead of smoothening them. Increasing the bandwidth of the
control system has shown good results in several cases.
b) Airflow
This control is complementary to the grate speed control. It maintains a constant volume of
cooling air entering the cooler independently from the grate underpressure.
Each cooling fan is equipped with a piezometer sensor which will recognize an increase or
decrease of the airflow and cause the cooling fan damper to close or open (in case of inlet
vane damper control) or the fan motor speed to decrease or increase (in case of variable
speed fan drives).
During normal conditions the cooling fans operate at about 2/3 to 3/4 of their maximum
performance so that enough spare capacity is left to cope with eventual kiln rushes.
Together, grate speed and air flow control will on one hand ensure a sufficient cooling air
supply to the cooler and, on the other hand, tend to provide more uniform combustion air
temperature to the kiln.
c) Hood draft
The third component of the cooler control system is the hood draft control.
An automatically controlled grate cooler can improve the whole kiln operation and allows the
operator to concentrate on other problems.
The kiln hood pressure is used to regulate the cooler vent air fan speed to maintain a
constant pre-set draft. As the draft tends to become positive, the cooler vent fan speed is
increased. This takes more air from the cooler and maintains the draft setpoint. As with the
other controls, reaction in the opposite direction is just as important.
Coolers with radiation walls (IKN) allow hood draft control by one of the first cooling air fans.

Figure 18 Cooler control

3.1.9 Cooler Dedusting
While dedusting of kiln exhaust gas can be commonly solved by using one type of dust
collector only (electrostatic precipitator), the choice of the most adequate system for
dedusting clinker cooler vent air raises quite often many discussions. This choice problem is
basically a result of the special and fluctuating conditions of the vent air to be dedusted:
normal operation kiln upset

airflow (actual volume) % 100 up to 150
air temperature C 200 - 250 up to 450
air dew point C 5 - 20 5 - 20
dust load g/Nm3 5 - 15 25 - 35
The dust particle size distribution can vary in a wide range depending on the burning
conditions in the kiln.
Dimensioning of the dedusting equipment must take into account the worst conditions, in
order to maintain the required clean gas dust content even at kiln upset condition.
The types of dust collectors for this application are compared below. Today's trend is:
multiclones will no longer be tolerated in new and many existing plants
gravel bed filters have proved to be inefficient and expensive
use of electrostatic precipitators is possible without restriction
bag filters with cooling of the vent air in a heat exchanger are often used nowadays
Type of collector Strengths Weaknesses

multiclone simple poor efficiency for particles
low investment cost < 20 m
low space requirement efficiency sensitive to gas
flow fluctuation
not sensitive to
temperature peaks comparatively high
pressure loss
high operating cost
electrostatic precipitator low pressure loss big unit required or use of
low operating cost pulse generator -> high
investment cost
low maintenance cost
possibly water injection
required
gravel bed filter not sensitive to highest investment cost
temperature peaks highest pressure loss
high operating cost
bag filter high efficiency no bags for temperatures
relatively low investment up to 450C precooling
cost required
high pressure loss
high operating cost
high maintenance cost

Figure 19 Grate cooler dedusting

3.1.10 Developments
Air recirculating (Duotherm) cooler
A patent has been taken out in 1970 by the "Socit des Ciments Franais" concerning the
recirculation of the vent air after sending it through a heat exchanger.
The first application of the unconventional system has been realized in 1970 at the
Beaucaire plant of the above mentioned company, on a 1500 t/d Fuller cooler.
Initial experience gained with this installation was very satisfactory.
Only few installations using this principle have been realized, e.g. in the Ulco plant. The
main advantages and disadvantages of this system are:
no dust emission at all possible wear of fan blades
simple (preventative measures
necessary)
low investment cost
maintenance and operating
heat recovery possible (at various costs higher than conventional
temperature levels) cooler dedusting system with EP
extension possible by adding further
heat exchange units
Modern cooler technology and problems in some cases have pushed this idea in the
background. However, it might be reactivated if it can be combined with modern cooler
systems.
Dual pass cooler

A completely new principle of cooling in a grate cooler has been introduced by Polysius in
1994: the dual pass cooler or REPOL-ZS.
This cooler can be considered a two-grate cooler with intermediate crusher where grate 1
and 2 are identical.
The hot, 1400C clinker from the kiln is fed on top of a layer of colder clinker already laying
on the cooler grate. At the end of the grate, the now cold lower clinker layer is extracted via
a special system consisting of reciprocating bars and a hopper. The upper layer which has
reached about 500C passes a roller crusher and is then returned to a intermediate hopper
below the kiln from where it is fed onto the empty grate to pass the cooling air a second
time, this time below the fresh hot clinker.
One 1400 t/d unit is in operation in Germany using Jet-Ring technology. With less than
1.6 Nm3/kg cooling air, extremely low clinker temperatures have been reported. The crucial
problems of this solution are intermediate transport and storage.
In spite of the compact size, high cooling degree with low air flow and low plate
temperatures, this cooler will only be successful if the intermediate temperature level can be
increased and the heat losses reduced.

Figure 20a Non venting cooler
Figure 20b Dual pass cooler (Polysius)

3.2 The Cross Bar Cooler
3.2.1 Principle
F.L.Smidth and Fuller developed together the new SF (Smidth - Fuller) Cross Bar Cooler
representing a completely new concept.
The basic idea was to develop a cooler in which conveying of clinker and air distribution
systems are separated. The SF cooler has a clinker conveying device installed above an
entirely fixed grate.
In addition the cooler should be less complicated, more efficient and easier to operate than
other grate coolers on the market. Sealing air is eliminated and the distribution of air is
optimized for all modes of operation
The thermal behavior of the SF cooler (e.g. heat balance, recuperation) is similar to the
other grate coolers.
3.2.2 Main features

One inclined fixed grate.
Clinker conveying by cross bars, separate from air distribution.
No thermal stress of grate.
Minimum wear on grateplates due to a dead layer of clinker (50 mm) protecting the grate
surface. The thickness is given by the space between the cross bars and the grate.
(Anticipated service life time at least 5 years)
Dynamic flow control unit (mechanical flow regulator) for each grate plate. The
mechanical flow regulator maintains a constant airflow through the grate and clinker
bed, irrespective of the clinker bed height, particle size distribution, temperature, etc.
No fall through of clinker to the undergrate compartment.
Eliminating undergrate clinker transport resulting in low installation height
for new plants.
Easy cooler operation by elimination of sealing air and automatic control of air
distribution.
Modularized cooler concept short delivery and installation time.
Different drive speeds across the cooler possible.
Additional control of clinker distribution.
Fewer and less expensive wear parts (easy to replace).
Easy visual inspection of undergrate compartment (clean undergrate, windows).
Sustainably high thermal cooler efficiency throughout the lifetime of the cooler.
Reduced system heat consumption.

Figure 21a: SF Cross Bar Cooler
Figure 21b: SF cooler grate with cross bars

3.2.3 Strengths and Weaknesses
No clinker fall through (no hoppers, no The clinker bed seems to be influenced
dragchain). by the conveying reciprocating cross
bar, resulting in disturbed clinker layers.
The grate is protected from
overheating. In case of fine clinker and coating
drops, air breakthroughs can occur.
Very high availability is expected.
The performance of the mechanical
Wear and tear affects only the
flow regulator (amount of cooling air)
conveying system and not the air
and its distribution is yet to be
distribution system.
assessed.
For each plate, the cooling air is
Airflow through the fixed grate at the
individually controlled.
cooler inlet (CIS) can generate dust
The amount of cooling air is about 1.6 and dust cycle.
to 1.8 Nm3/kg.
Reduced height and maintenance
required since the undergrate clinker
transport can be dropped.
Time for installation is short due to
modular concept.
Remark: So far, no SF Cross Bar Cooler is in use within the Holcim group and therefore no
first hand experience is available. Worldwide, there are only three SF cross bar
coolers installed. Two of a capacity of 450 t/d and one of 2000 t/d. (as of January
1999)
Figure 22a: Cross Bars: Easy to replace wear parts

Figure 22b: Mechanical flow regulator
Figure 22c: Modular concept: One module

3.3 The Travelling Grate Cooler
3.3.1 Principle
The traveling grate cooler (Recupol) was originally developed by Polysius for use in
combination with grate preheater (Lepol) kilns. Using the same principle and similar
technology, it uses the same wear parts. The following main components can be
distinguished:
Casing with kiln hood and connections for air at different temperature levels
Inlet with water cooled chute (2nd generation) and pulsator
Traveling grate with return carrying idlers and drive system
Aeration system with fans, undergrate compartments
Riddling extraction system with chutes, flap gates, hoppers and transport
Clinker crusher
Material transport
The clinker is carried by a horizontal traveling grate which works like a stationary
caterpillar chain with perforated chain plates. In contrast to the reciprocating grate cooler,
the clinker does not tumble over plate edges, but remains as undisturbed layered bed
from inlet to discharge.
Heat exchange
Heat exchange takes place, like for the reciprocating grate according to the
cross current principle. Because the layers remain, it should be even better, at least
theoretically.
Cooling air
Ambient air is blown by a number of cooling air fans to underneath of the travelling grate
plates carrying the clinker. Pressure and flow criteria of cooling air are basically as for
the reciprocating grate cooler.
Water cooled inlet chute
In order to achieve rapid cooling in the inlet section, but also to protect the travelling
grate from the highest clinker temperatures, Recupol coolers were equipped with a water
cooled inlet chute.
Key figures / KPI
Specific grate loading: 25 - 30 t/d m2 (design)
Largest units: 3000 t/d (Lgerdorf kiln 10)

Figure 23 Travelling grate cooler

3.3.2 Strengths and Weaknesses
Travelling grate cooler compared to reciprocating coolers:
Possibility of replacing grate plates Larger machine for the same
during operation (on the returning part) grate area equipment requiring
Undisturbed, layered clinker bed is more space and higher civil cost
better for optimum heat exchange Lower specific grate loadings
adding further to overall size
More expensive to build than a
reciprocating grate cooler
The absence of clinker
movement (see above) was
often considered a disadvantage
because of cases where a solid
(fritted) layer on top of the
clinker bed made it impermeable
for air. For this reason, pulsators
were installed for first cooling
fans.
Much higher maintenance
requirement with ageing
equipment
Heat loss via cooling water for
inlet chute
Due to the mentioned weaknesses, Polysius eventually decided to develop their own
reciprocating grate cooler (Repol) around 1980:

Figure 24 Travelling grate cooler: Design details

4. ROTATING COOLERS
4.1 The Rotary Cooler or Tube Cooler
4.1.1 Principle
The rotary cooler consists mainly of a rotating cylinder, similar to a rotary kiln.
The clinker is fed through the inlet chute and is then cooled by air while being transported
towards the outlet end. Cooling is performed in countercurrent flow. The tube is equipped
with internal lifters which improve the heat transfer. About 2/3 (66%) of the cooler length is
lined with refractory bricks.
The rotary cooler is of simple design and is the oldest type of clinker coolers. It was seldom
used for modern, large kiln systems. Therefore comparatively little design and operating
experience is nowadays available for rotary coolers above 2000 t/d. However, the
application of rotary coolers still offers certain advantages. Presently units up to 4500 t/d
(dimensions dia 6.3/6.0 x 80 m) are in operation. It will be interesting to follow the future
development of large rotary coolers.
4.1.2 Design Features

Arrangement of the rotary cooler is normally in the extension of the kiln axis; in many
cases the reverse manner (underneath the kiln) has been applied.
The diameter of the cooler is similar to that of a corresponding suspension preheater
kiln. Likewise the rotating speed is in the same range as for the kiln (max. 3 rpm).
Length/diameter ratio: L/D ~ 10.
Many cooler tubes are designed with an extension in diameter in order to reduce air velocity.
The inclination is comparatively high (in the order of 5%).
Like for all rotating coolers, the internal heat transfer equipment is an important part of
the rotary cooler. Its task is to generate additional area by scattering the clinker without
generating too much dust. Basically a similar design may be applied as in a planetary
cooler tube (see next chapter) however the following differences must be considered:
The clinker falling heights are larger. Wear protection of shell and lining is essential.
At a comparative length position the clinker in a rotary cooler is hotter than in a
planetary cooler.

Figure 25 Rotary cooler

The following zones can typically be distinguished in a rotary cooler (simplified):
A Lined inlet zone
B Lined crushing teeth zone
(metallic teeth)
C Lined cast lifter zone, lining protected by wearing plates
(at least in the second half)
D Cast lifter zone, shell protected by wearing plates
(having air gap, giving also insulating effect)
E Sheet metal zone with wearing plates
Construction materials have to be selected according to the high temperature and wear
requirements.
4.1.3 Cooling performance

Depending on the design and the shape of the lifters clinker outlet temperature usually tends
to be high. In many cases it is necessary to enhance the cooling by injecting water into the
tube (up to 60 g/kg clinker) in order to reach reasonably low clinker temperatures of 100 to
150C.
The cooling efficiency (heat recuperation) is equal or even slightly better than on a planetary
cooler.
4.1.4 Strengths / Weaknesses
Simplicity of cooler design, robust Not recommended for large units
piece of equipment. (above 2000 t/d)
No special mechanical problems Formation of build-ups ("snowmen") in
comparable to a rotary kiln. the inlet chute. A water-cooled chute or
a dislodging device is required in such
No control loops.
case.
Easy commissioning.
Clinker outlet temperatures tend
No waste air and therefore no to be high and therefore water injection
dedusting equipment required is usually required.
Electrical energy consumption up to 5 Due to large falling height wear
kWh/t lower compared to grate cooler. protection in the tube must be
reinforced (compared to a planetary
Rotational speed can be adjusted and cooler).
therefore upset kiln conditions can be
handled easier than with a planetary High kiln foundations are required.
cooler.
Cooler inlet seal can contribute to
Suitable for AS type precalcining additional false air inlet.
system tertiary (extraction of hot air is
possible).

Figure 26 Internal transfer equipment for rotary and planetary coolers
4.2 The Planetary Cooler
4.2.1 Principle
The planetary cooler is based on the same cooling principle as the rotary cooler in the
preceding chapter. However, the essential difference of a planetary cooler is the number of
individual cooling tubes. The flow of clinker is subdivided into 9 to 11 (usually 10) cooling
tubes which are installed around the kiln circumference at the kiln outlet (see Fig. 15).
Therefore the planetary tubes follow the kiln rotation. Because of their connection to the kiln
rotation, planetary coolers do not need a separate drive. This fact already illustrates one
main advantage of the planetary cooler: its simplicity in operation.
Strictly speaking the cooling of clinker does not only start in the cooling tubes but already in
the kiln. In the case of a planetary cooler the kiln burner pipe is always inserted into the
rotary kiln so that a cooling zone behind the flame of 1.5 to 2.5 kiln diameters is created.
This zone is called the "kiln internal cooling" zone and must be considered as an integral
part of any planetary cooler. In this zone the temperature of the clinker drops from 1450 to
1200 - 1300C. This temperature reduction is important for the protection of the inlet
opening, the elbow and the first section of the cooling tubes.

After this first cooling in the kiln internal cooling zone the clinker falls into the elbows when
they reach their lowest point of kiln rotation. The hot clinker is then cooled by air in
counterflow (the amount of air equals the amount of secondary air). The air is heated up to
approx. 700C. The clinker reaches final temperatures which are typically in the range of
140 to 240C.
A considerable amount of heat is also transferred to ambient by radiation and convection
since approx. 75% of the cooler shell is not insulated.
4.2.2 Historical
Planetary coolers have been used since 1920. When large kiln units and grate coolers were
developed planetary coolers were abandoned for many years. But about 1966 planetary
coolers of large capacities were introduced. At that stage serious mechanical problems
occurred on these first large planetary coolers. As a consequence a lot of work had to be
done in order to improve the mechanical design of planetary coolers. As a result of extensive
computer calculations and operating experience the planetary cooler became a
mechanically reliable piece of equipment.
In the late 1970's, the design had reached a high standard and a considerable level of
perfection. Units of up to 5000 t/d were envisaged. With the demand for permanently larger
units using precalciner technology with separate tertiary air dusts, the boom period of the
planetary coolers came to an end.
Figure 27 Planetary cooler

4.2.3 Design features
Planetary coolers in the late 1970's had the following design features:
Shell extension:
The kiln shell is extended beyond the cooling tube outlets and is supported by an
additional roller station.
Fixation of cooling tubes:
Fixed support of cooling tubes near inlet and loose support near outlet end.
With larger coolers, the cooling tubes can consist of two separate sections requiring three
supports. In that case two fixed supports are located near inlet and near outlet and a loose
support is located at the interconnection point in the middle.
Design of cooler supports:
The kiln shell is reinforced (high thickness) where the cooler support structure for the
cooler is welded on. The support structure (base and brackets) itself is of heavy design
consisting of reinforcement ribs and box beams.
Cooler length:
Length/diameter ratio of tubes is approx. 10:1
Inlet openings:
The inlet openings to the cooler elbows weaken the kiln shell and high mechanical and
thermal stresses occur in that zone.
The openings are made of oval shape and the kiln shell is considerably reinforced in its
thickness (up to 140 mm in large kilns) in order to compensate for the weakening.
In some cases a diagonal retaining bar (made of high heat resistant steel) is
incorporated in the opening in order to avoid that large lumps can enter the cooler.
Kiln-to-elbow joint:
This joint is designed in a manner that no forces due to thermal expansion and
deformation are transmitted from elbow to kiln.
Elbow:
In order to prevent that clinker is falling back into the kiln while the opening is on top
position, the position of the cooling tube is displaced back against the direction of
rotation. The elbow design must avoid excessive dust backspillage and wear.
4.2.4 Internal heat transfer equipment (see Fig. 26)

Cooling performance depends strongly on efficient lifters of solid and durable design. Since
high heat resistant metallic lifers are available on the market also the high temperature
zones can be adequately equipped. Special high temperature alloys can be used for this
purpose. They can withstand maximum temperatures of up to 1150C. These alloys are
usually characterized by a high chromium content of approx. 30% Cr. Other elements as Ni
or Mo can occur in various proportions. Fig. 26 shows a typical arrangement of heat transfer
internals. Breaking teeth are applied in the hottest zone. They are able to crush large lumps
of clinker and create also a tumbling effect, which improves the heat transfer. They are of
heavy design and mounted on separate supports.
The first rows of lifters must be carefully selected regarding design and material. Their
functioning is very important since they also protect the following lifters from overheating.

Figure 28a Temperature profile in planetary cooler

Figure 28b Water cooling for planetary coolers

5. VERTICAL COOLERS
5.1 The Gravity Cooler (G - Cooler)

The Claudius Peters Company have developed the g-cooler. The letter "g" stands for
gravity since clinker movement is performed by gravity.
This cooler is designed as an after cooler and can therefore only be used in connection with
a primary cooler such as a short grate cooler or a planetary cooler. The installation together
with a grate cooler is shown in Fig. 29.
An intermediary crusher reduces the clinker size to 20 - 30 mm. The material of approx.
400C is then filled by a drag chain into a vertical shaft. Cooling is performed by horizontal
rows of tubes which are cooled by internal air flow. The heat is therefore exchanged
indirectly and the air remains dust-free. The clinker slowly drops down (at a speed of 20
30 mm/s) and reaches final temperatures of approx. 100C at the discharge.
There is no dedusting equipment required for the cooling air. However, the system according
to Fig. 29 as a whole is usually not free from dusty waste air. In case of a suspension
preheater kiln system there is still some waste air required on the grate cooler since the kiln
cannot take all the hot air produced during the first cooling step. In addition, a marginal
amount of dusty air is produced by the g-cooler itself (top and discharge).
The application of this cooler type is often considered for kiln extension projects. If an
existing grate cooler (or a planetary cooler) has to be operated at higher capacity the new
clinker outlet temperature can become too high. In this case the clinker temperature can be
reduced by a g-cooler used as an aftercooler.
5.2 The Shaft Cooler

A shaft cooler can be operated waste-air-free and theoretically offers an ideal countercurrent
heat exchange and thus high recuperating efficiency. Based on the idea the first large shaft
cooler was designed and constructed on a 3000 t/d kiln in 1973.
The experience gained in the plant shows that it is possible to operate such equipment but
some serious disadvantages have to be taken into account:
All depends of the clinker granulometry! Theoretically, an extremely uniform clinker
granulometry having no fines and no coarse material would be required. This is hardly
achievable in a cement kiln. Therefore, fluctuations occur.
High cooling air quantity (= secondary air) of 1.05 Nm3/kg cli is required but even so the
clinker exit temperature of 350C is very high.
High power consumption (10 kWh/t)
For the above reasons, the technical realization is not yet solved. The shaft cooler so far is
not a reasonable alternative to the conventional clinker coolers.

Figure 29 Gravity cooler (g-cooler, CPAG)
Figure 30 Shaft cooler

Figure 31 Claudius Peters CPAG: Combi Cooler
Figure 32 FLS: Coolax Grate Cooler

Figure 33 Fuller: Controlled Flow Grate (CFG) Cooler

Figure 34 IKN: Pendulum Cooler
Figure 35 KHD: Pyrostep Cooler

Figure 36 Polysius: Repol RS Cooler

Chapter 2
Internal Kiln Fittings

Kiln Chain Systems
A. Obrist
PT 96/ 14036 / E
1. INTRODUCTION ............................................................................................................. 177

2. Functions of a Kiln Chain System ............................................................................... 177
2.1 Heat Exchange.................................................................................................. 177
2.2 Cleaning of the Kiln Shell .................................................................................. 178
2.3 Transport of Material ......................................................................................... 178
2.4 Reduce Dust Emissions .................................................................................... 178
3. Individual Zones of a Chain System ............................................................................ 179
3.1 Free Zone of the Kiln Inlet................................................................................. 179
3.2 Dust Curtain Zone ............................................................................................. 179
3.3 Plastic Zone ...................................................................................................... 180
3.4 Granular Zone (Preheating Zone) ..................................................................... 180
3.5 Heat Resistant Zone ......................................................................................... 180
3.6 Main Characteristic Data of the Individual Chain System Zones ...................... 181
4. Arrangement of Chains................................................................................................. 181
4.1 Straight Curtain ................................................................................................. 181
4.2 Spiral Zone........................................................................................................ 182
4.3 Multiple Spiral Curtain ....................................................................................... 183
4.4 Triangular Curtain (Z-Curtain) ........................................................................... 184
4.5 Garlands............................................................................................................ 185
4.6 Festoons ........................................................................................................... 186
4.7 Spiral Garlands ................................................................................................. 186
4.8 Thermochains ................................................................................................... 187
5. Types of Chain Links .................................................................................................... 188
5.1 Round Links ...................................................................................................... 188
5.2 Long Links......................................................................................................... 189
5.3 Oval Links ......................................................................................................... 189
5.4 Other Types of Chain Links............................................................................... 189
6. Chain Material ................................................................................................................ 190
6.1 Mild Steel Chains .............................................................................................. 190
6.2 Heat Resistant Alloy Chains.............................................................................. 190
7. Chain Hangers ............................................................................................................... 191

8. Main Characteristic Data of Chain Systems................................................................ 194
9. ANNEXES ....................................................................................................................... 195
1. LITERATURE.................................................................................................................. 202
2. Test Questions............................................................................................................... 203

Summary:
A kiln chain system has four main functions:
It helps to increase the heat exchange between gas and raw meal
It keeps the kiln shell (lining surface) clean
It assists the transport of material through the kiln tube
It helps to reduce the dust emission
A properly designed chain system must respect the changing properties of material passing
through the kiln tube. In a wet process kiln the material is fed as a liquid slurry and changes it
properties subsequently in several steps inside the chain system to dry preheated granules.
In accordance with the changing material properties different arrangement of chains (straight
curtains, spirals, garlands, etc.) have to be used for individual parts of the system to satisfy
the specific requirements. Also the chain densities and the height of the free tunnel below the
chains have to be selected carefully in order to reach the maximum efficiency.
The chain links can have different shapes (round, long, oval etc.), preferably round links. The
chemical composition of the chains' alloy and its physical treatment (hardening) strongly
influence the life time of the system.
Different types of chain hangers can be used (single or multiple hangers, with or without
shackles etc.). They have to guarantee a sufficient stability, to enable an easy installation and
they should as far as possible assist the function of the chains.

NOMENCLATURE
Just a few symbols and names are to be explained before starting this lecture, the other ones
will be explained in the respective chapters.
Figure:
Dis Diameter inside kiln steel shell

DIL Diameter inside kiln lining
hfr Theoretical free height under the chains (see attached sketch),
expressed in mm or as % of DIL
density of chains m2/m3 is calculated for individual parts (zones) of the system as the total
surface area of chains in the respective zone divided by the
volume inside lining of this zone
density of chains kg/m3 similar to the above mentioned density, but concerns the weight of
chains instead of their surface

1. INTRODUCTION
Wet process kilns cannot be successfully operated without internal kiln fittings, among which
the kiln chains are the most typical and most frequently used ones. The number of existing
wet process kilns is still high (~33% in Holcim Group) and a conversion from wet to dry
process is very expensive. By improving the existing chain systems or, where necessary, by
installing a completely new chain system, the kiln operation can be upgraded considerably
with relatively moderate investment costs.
2. FUNCTIONS OF A KILN CHAIN SYSTEM

The kiln chain system has 4 main functions:
2.1 Heat Exchange

The heat exchange between hot gases and the raw material depends on the surface area
exposed to the hot medium. In the parts of kiln where no chains are installed, this surface
area consists of the surface of the material layer on the kiln bottom and of the surface of the
remaining part of the kiln shell (resp. lining). By installing the chains a large additional surface
area can be gained, exceeding that one mentioned above several times (up to 10 times and
more) in the respective part of the kiln. By improving the heat exchange the specific heat
demand is reduced and the kiln output is increased.
In Fig. 1 different positions of a chain during one kiln rotation are shown. In position 1 the
chain is exposed to the stream of hot gases and thus heated up. The cooling of the chain
(which passes its heat to the layer of material) starts in position 2, continues in position 3 and
ends in position 4.

Figure 1:
2.2 Cleaning of the Kiln Shell

In the upper part of the kiln the characteristics of the wet, sticky raw material favors the
formation of mud coating and mud rings. This would reduce the free kiln cross sectional area
and thus obstruct the flow of material and gases. Growing mud rings make the kiln operation
difficult. It is one of the main functions of the chain system to keep the internal kiln shell
surface clean, free of coating or rings. Due to the kiln rotation the chains slide on the kiln shell
(resp. lining) and destroy the rings and the coating. The sliding movement of a chain cleaning
the kiln shell is shown in Fig. 1 (position 3).
2.3 Transport of Material

The properties of material in different parts of the kiln differ considerably. In some sections of
the upper part of the kiln, where the material is sticky and plastic, its transporting is more
difficult than in other sections.
As a regular flow of material is of an eminent importance for a smooth kiln operation, it is
necessary to install material flow assisting devices in some sections. Some special
arrangements of chains can help to draw the material through the critical sections. This can
be achieved by chains moving in the desired direction (garlands) or by a screw shaped
arrangement of the chain fastening points. Other arrangements of chains can be an
obstruction to the flow of material and should therefore never be used in the critical sections.
2.4 Reduce Dust Emissions

The gases leaving the kiln contain a certain amount of dust consisting mainly of partly
calcined, hot raw material. The dust load of gases depends on the properties of the raw
material and on the specific conditions of the kiln operation. Dust loss should be kept small, it

means a loss of heat and material. The kiln chain system, mainly its upper part, can help to
reduce the dust emission. Dust particles carried by the stream of gases stick to the wet
surface of chains and later when these chains are emerged into the layer of material, this
dust is passed over to the slurry.
3. INDIVIDUAL ZONES OF A CHAIN SYSTEM

The material passing the chain system changes subsequently its properties - it loses water
and is heated up. According to the different material properties the total chain system can be
divided into several zones. These zones are:
3.1 Free Zone of the Kiln Inlet

This short zone is considered to be a part of the chain system in spite of the fact that no
chains are installed here. A sufficient amount of slurry should be accumulated in this zone in
order to guarantee a constant and regular flow into the lower parts of the system.
Good results have been obtained with the zone length of 1 to 1.5 kiln diameters.
3.2 Dust Curtain Zone

The dust curtain zone is relatively short, its length does not exceed 0.5 DIL under normal
conditions.
The material entering this zone still has the relatively good flow properties of the kiln feed
(slurry). When leaving this zone, the material has a lower water content and becomes more
"plastic", essentially due to the inter-mixing of the dust previously retained by the chains in
this zone.
In order to achieve a good dust catching efficiency, the density of chains must be high (some
8 to 15 m2/m3) and the free height below the chains should be 18 - 27% of DIL.

3.3 Plastic Zone
The length of this zone depends on properties of raw material, slurry moisture, characteristics
of the kiln operation etc. and can vary in a wide range (approx. between 1.5 and 4 DIL). The
material in this zone is plastic and sticky, still relatively cold and wet and because of these
properties it favors mud coating and mud ring formation. The transport of material through
this zone is the most difficult one among all the zones of the chain system.
Due to the material properties mentioned above the chains in this zone must have a good
shell cleaning and material transporting efficiency. The density of chains should be relatively
low, some 5 to 8 m2/m3. As to achieve a big free tunnel under the chains, the free height hfr
should be approx. 30% or, if garland chains are installed in this zone, some 40%. Heavier
(thick wire) chains should be installed.
In order to be sure that the zone of plastic material will always stay inside the zone of chains
which can treat it successfully, the respective arrangement of chains should be slightly
extended in the downstream direction as to obtain a sufficient safety.
3.4 Granular Zone (Preheating Zone)

The recommendable length of this zone depends on the desired material temperature and the
rest water content at its discharge end. Good results have been achieved with a zone length
between 2 and 4,5 DIL.
The material entering this zone is not plastic any more, it forms granules which are easy to be
transported and do not favor a mud ring formation. The granules should be dried and heated
up in this zone.
The chains should heat the material gently without unnecessary dust generation, they should
enable a good heat exchange. Lighter (thin wire) chains should therefore be installed. A chain
density of approx. 6 to 10 m2/m3 and a free height of approx. 25 to 30% can be
recommended. This zone is sometimes divided into an upper and a lower part. Both parts
have the same (or at least a similar) arrangement, but the lower part has a higher density of
chains than the upper one. The damming effect of the lower part, caused by the thicker layer
of chains on the kiln bottom, helps to increase the material retention time and improves the
heat exchange.
3.5 Heat Resistant Zone

This zone is relatively short, its length does not exceed 1,5 DIL. The material, dry and hot
granules, can easily be transported. A very gentle treatment of the material is required in
order to keep the dust creation as low as possible.
The main function of the chains is to protect the upstream part of the system against heat
radiation and too high a gas temperature. Chains made of heat resistant steel should be
installed in this zone. Lighter (thin wire) chains should be preferred.

3.6 Main Characteristic Data of the Individual Chain System Zones
Table 1
Zone
Free Dust Plastic Granular Heat
Curtain Upper Lower Resistant
Length DIL 1,5 0.5 1.5 to 4 1 to 3 1,5 1,5
hfr % DIL 18 - 27 40 / 30 25 to 30
Density m2/m3 8 to 15 5 to 8 6 to 10
Moisture % 30 to 40 15 - 25
Material temp. C 20 100 <600
Chain temp. C <400 <1'000
Gas temp. C 160 to 240 <1'100
Material flow
4. ARRANGEMENT OF CHAINS
Each individual zone of a chain system must have its own specific properties in order to
satisfy the requirements mentioned previously.
4.1 Straight Curtain

As shown in Fig. 2, this arrangement of chains is very simple. The chain fastening points form
a ring. Several straight curtains are combined to a straight curtain zone. The distance
between the individual straight curtains (rings or rows) should not be too long, this would lead
- in order to achieve a sufficient density - to too great a number of chains per ring and thus to
big heaps of chains on the kiln bottom obstructing the flow of material.
Figure 2: Straight Curtain
The main advantages of a straight curtain are its simple design and installation as well as an
easy maintenance. Its main disadvantages are the poor shell cleaning efficiency and the fact

that this arrangement does not assist the transport of material. The straight curtains should
therefore not be used in the plastic zone or in the dust curtain, but they can be recommended
for the granular zone.
4.2 Spiral Zone

In a spiral curtain the chain fastening points follow the screw lines on the kiln shell.
The spiral (screw) lines should have an inclination of approx. 30 (see Fig. 3). In order to
assist the transport of material, the inclination must have the proper sense, i.e. the sense of
rotation must be taken into consideration.
Figure 3: Spiral Curtain (4-start spiral)
The arrangement with 4 screw lines per circumference, called a 4-start

spiral, is most frequently used. This arrangement allows for a good shell cleaning effect and
at the same time overlapping of chains can be avoided, as shown in Fig. 4a. The
recommended chain length is namely approx. 0.7 D, and the circumferential distance
between adjacent screw lines is D/4 = 0.78 D.
The benefits of this solution become clear if we compare it with other arrangements having a
different number of screw lines. An arrangement with less than 4 screw lines per
circumference (Fig. 4b) does not enable a full shell cleaning effect, because some parts of
the shell have no contact with a chain. An arrangement with more than 4 screw lines leads
necessarily to overlapping of chains (Fig. 4c) and thus to a higher layer of chains on the kiln
bottom which is not favorable for the transport of material.

Figure 4a-c:
The very good material transporting and the good shell cleaning efficiency are the main
advantages of a spiral curtain. Its disadvantages are a slightly smaller heat exchanging
efficiency, a more difficult brick lining work and a limited maximum chain density. For these
reasons, the spiral curtain arrangement should not be used in the granular zone but it can be
recommended for the plastic zone.
4.3 Multiple Spiral Curtain

A spiral curtain having 8 or more spirals per circumference is called a multiple spiral curtain.
Such an arrangement enables to achieve high chain densities and the passages between
chains are narrow, therefore, it can be used in the dust curtain zone. its material transporting
efficiency is not as good as that one of a 4-start spiral, but it is still better than that one of any
other arrangement enabling a similarly high chain density. A multiple spiral arrangement can
therefore be recommended for dust curtain zones in such cases where the kiln feed flow
properties are poor.

4.4 Triangular Curtain (Z-Curtain)
In a triangular curtain, also called Z-curtain, the chain fastening points follow a zig-zag line
(see Fig. 5). This arrangement enables to reach high chain densities without obstructing the
flow of material and gases too much. Because of the equal distribution of chains the heap of
chains on the kiln bottom is not too high in spite of the high chain density (see Fig. 6).
Figure 5 & 6
The arrangement with angles of 60 and with 8 "triangles" per kiln circumference has proved
to be the most successful one. In such an arrangement the layer of chains on the kiln bottom
is only approx. 4 chains high (because 8 triangles have all together 16 fastening lines and the
chain length of approx. 0.73 D corresponds to D/4, therefore, each chain passes 16/4 = 4
lines).
The dust catching efficiency of such an arrangement is very good, it was therefore used in
the dust curtain zone in such cases where the kiln feed flow properties are relatively good.
Frequently one single triangular curtain was installed, but two are also possible.
Nowadays, for dust curtains a multiple start spiral arrangement is preferred to a triangular
curtain because of its material transport efficiency.

4.5 Garlands
Chains having their both ends fastened to the kiln shell are called garlands (Fig. 7) The
fastening points form straight rings in the kiln shell. The shape of a garland is characterized
by the chain length, by the axial distance LAX and by the angle (see Fig. 7 and 8). Good
results have been achieved with an angle 90 to 110 and a distance LAX 0.5 to 0.9 D.
Figure 7 & 8:
Due to the sliding movement of the garland chains on the kiln shell (resp. lining) surface, their
shell cleaning and material transporting efficiency is very good and their heat exchange
efficiency is approx. 1.5 times higher than that one of pending chains (see Fig. 9).
Figure 9
As can be seen in Fig. 9 the garland chain should be hung in a "reverse" sense, i.e. in a
sense differing from that one of a screw line, in order to assist the material transport.

Due to the properties mentioned above, the garland chains have been used mainly in the
plastic zone.
In spite of the advantages mentioned above CS/PT does usually not recommend the
installation of garlands because of their disadvantages, namely:
complicated installation
difficult maintenance
relatively short life time
at the ends of the garland zone the shall cleaning efficiency is very poor (no movement of
chain)
4.6 Festoons
Garlands without overlapping chains in the axial direction are called festoons (Fig. 10). The
installation and maintenance of festoons are less difficult than that of garlands but their shell
cleaning efficiency is poorer (the areas between the individual bays of festoons are not
cleaned reliably).
Figure 10
4.7 Spiral Garlands

An arrangement of garlands where the chain fastening points form screw lines ("spirals") is
called spiral garlands. Its material transporting efficiency is even better than that one of
normal straight garland zone, but the erection and maintenance are more complicated.

4.8 Thermochains
Thermochains are a special type of festoons, meeting the following conditions:
The axial distance LAX between the two fastening points is short (approx. 0.1 to 0.15 DIL).
The angle between the two fastening points is approx. between 60 and 120,
preferably 90 to 120.
The free height under the chain hfr is approx. between 0.4 DIL (for = 120) and 0.6 DIL
(for = 90).
The chain (shackles included, if used) is approx. 1.1 times longer than the distance
between the two fastening points, measured on the lining surface (see Fig. 11a and 11b).
A good heat transfer and at the same time a gentle treatment of material preventing an
unnecessary dust creation are the main advantages of thermochains. Their disadvantages
are a low shell cleaning ability and a very limited material transporting efficiency.
Thermochains only have a limited sliding movement on the kiln lining compared to garlands
and for this reason the sense of hanging (reverse or non-reverse) does not make too much
difference. Thermochains cannot be used in the upper and central part of the chain system
where the shell cleaning efficiency is of an eminent importance. They should be used in the
lowest (hot) part of the system in such cases when another type of chain arrangement
enabling the same heat exchange would lead to an excessively high dust emission.
Figure 11a & b:

5. TYPES OF CHAIN LINKS
5.1 Round Links

The round links (Fig. 12a) can rotate slowly when kiln is in operation which has two
advantages:
the wear is distributed equally over the whole link circumference, and therefore, in
comparison with other types, round links have a longer life time.
due to the rotation of links their surface is kept clean which enables a good heat
exchange (links are not surrounded by an insulating mud layer).
Because of the properties mentioned above round links should be preferred to other types
mainly in those zones where the material is wet and sticky.
For calculations of the chain surface area and chain weight, the following formula can be
applied:
surface area of 1 link O = 2S ( s + d )

volume of 1 link s
V = 0
4
weigt of 1 link G = V ( 7.8 t m 3 )
The above formulas are valid for round chain links with a round wire cross section.
Some chain suppliers express the mentioned specific properties as chain surface area per 1
m of chain and chain weight per 1 m of chain. These values are formulated as follows:
01m = 1000mmm
01m link
d (mm )
G1m = 1000mmm
G1m link
d (mm )
Therefore, if the chain weight and surface area are to be calculated from the 1m specific
data, the chain length LCH should be measured as shown in the following sketch:
Figure

5.2 Long Links
The long links (Fig. 12b) cannot rotate like the round ones, their life time is shorter and their
self-cleaning ability poorer.
For calculations of the chain surface area and chain weight, the following formulas can be
applied:
surface area of 1 link O = s{2(l d ) + (s + d )}
volume of 1 link s
V = 0
4
weigt of 1 link G = V ( 7.8 t m 3 )
The above formulas are valid for long chain links with a round wire cross section.
5.3 Oval Links

The oval links (Fig. 12c) are similar to long links, they have similar properties and their
surface area and weight can be calculated (with a negligible mistake) by means of the
formulas mentioned in the part 5.2.
5.4 Other Types of Chain Links

Besides the links types mentioned above, the suppliers occasionally offer various other types
of chain links such as asymmetric links, overlapping links (Fig. 12e), links with a non-round
wire cross section (Fig. 12d) etc. These types are not to be recommended for cement kilns
and are very rarely used, except the links with a non-round wire cross section.

Figure 12: Type of Chain Links
6. CHAIN MATERIAL
6.1 Mild Steel Chains

A big majority of the kiln chains are made of mild steel. In order to withstand the friction
between chains, between chain and raw material, between chain and hanger etc. they have
to be made of a "through hardened" steel. This hardening (result of a thermal treatment) is
one of the important chain properties and only experienced kiln chain suppliers are in
possession of the necessary know-how for this procedure. For this reason, even the mild
steel chains should be purchased from experienced suppliers.
6.2 Heat Resistant Alloy Chains

Due to the thermal load of the kiln and the chain temperature at the hot end of the system,
the portion of the heat resistant chains should be some 15% of the total weight of chains. Mild
steel chains should not be installed in that part of the system where a chain temperature of
450C or more is to be expected.
The chain suppliers offer a lot of various heat resistant steel qualities. Besides the thermal
treatment the chemical composition of the respective alloy is the most important criteria. The
two main components are nickel and chromium.
Nickel increases the alloy resistivity against reducing kiln atmosphere, but a higher nickel
content becomes dangerous if the raw material or kiln gases contain sulphur compounds
which could react with it.
Chromium increases the alloy resistivity against high temperatures, but a higher content of
chromium makes the alloy sensible to sudden changes of temperature.

Above 20% Cr and at operating temperatures between 600 - 900C an intermetallic
compound can occur (-phase) which makes the alloy very brittle and causes destruction.
Whether this phenomenon occurs or not depends also on the Ni-content and on other
elements. Alloys which are sensitive to -phase formation must be used at working
temperatures above 900C.
Because of the properties mentioned above, it is necessary to find a compromise. Good
results have been achieved with heat resistant alloys containing approx. 18 to 25% Cr and
approx. 5 to 13% Ni.
7. CHAIN HANGERS
Chain hangers can be divided into 2 groups, namely single chain hangers and multiple chain
hangers.
The single hangers have only one chain fastening point, the multiple hangers have several
fastening points. The single chain hangers should be preferred in cases, where larger
distances between the chain hanging points are desired - this solution enables to keep the
weight of hangers as low as possible. (Under favorable circumstances, the weight of hangers
should not exceed some 15% of the weight of the chains).
The chain can be fixed to the hanger by means of a shackle (Fig. 15). Specially designed
hangers enable shackleless hanging of chains.
In Fig. 13 and 14 some examples of the chain hangers are presented, namely:
Fig. 13 a, b, c single hangers with shackles
Fig. 13 d single hanger, shackleless
Fig. 13 e, f
Fig. 14 b, c multiple hangers with shackles
Fig. 13 g
Fig. 14 a multiple hangers, shackleless

Figure13:

Figure 14:

The gap between the plate of a multiple hanger and the steel kiln shell should not be too
wide. Gaps exceeding some 20 mm enable, when the kiln lining is worn out, penetration of
chains into the space between kiln shell and hanger. This leads to tangling of chains and
finally to a destruction of chains and hangers.
Figure 15: Shackle
8. MAIN CHARACTERISTIC DATA OF CHAIN SYSTEMS

The following average data are to be understood only as a very rough guideline.
Length of the total chain system 18 to 35% of total kiln length
(some 6 to 10 D)
Total weight of chains:
- smaller kilns (< 1000 t/d) 9 to 12% of daily kiln output
- bigger kilns 11 to 20% of daily kiln output
Total surface area of chains:
- smaller kilns Fe/Fii*) = 1.1 to 1.8
- bigger kilns Fe/Fi = 1.5 to 2.6
* Fe = total surface area of chains
Fi = total surface are of kiln shell (inside lining)

9. ANNEXES
Annex 1: Example of chain system record keeping (DA K5)
Annex 2: Example of material sampling port arrangement on wet kiln (BP K1)
Annex 3: Example material sampling and mass balance (BP K1)
Annex 4: Example of material sampling port on wet kiln (BS K1)

Annex 1: Example of chain system record keeping (DA K5)

Annex 2: Example of material sampling port arrangement on wet kiln (BP K1)

Annex 3: Example material sampling and mass balance (BP K1)

Annex 4: Example of material sampling port on wet kiln (BS K1)

10. LITERATURE
P. Weber Wrmebergang und Wasserverdampfung beim Nassdrehofen

Zement-Kalk-Gips (1959), No. 5, p. 208 ff
Legrand: Calcul des coeffients de perte de charge et de filtration d'un rideau
de chaines
Rew. Mater. Constr. (1961), No. 549, p. 327 - 332
De Beus, Cement Technology: Design of Kiln Chain Systems Narzymski:
Rock Products 69 (1966), No. 7, p. 77 ff
Bennet, C.S.: Chain Experience in Wet Process Kilns
Minerals Processing, Vol. 8 (1967), No. 10, p. 18 - 19
De Beus, A.J.: Mind your Chain Dollar
Minerals Processing, Vol. 8 (1967), No. 10, pa. 12 - 17
Feiser, C.F.: Comments on Kiln Chain Developments in the Cement Industry
Minerals Processing, Vol. 8 (1967), No. 9, p. 11 - 13
Drayton, W.E.: Know your Kiln's Chain System
Rock Products (1972), H. 5, p. 88 ff
Numerous TC-VA Reports
Numerous Reports of Holcim Group Plants
"Datenbank-Blatt" Collection of TC-VA

11. TEST QUESTIONS
1) Name the main functions of a chain system and explain them!
2) According to the changing properties of material passing along the kiln tube, which
individual zones do we distinguish inside of a chain system? Describe them, define the
main requirements which the chains have to meet in each individual zone!
3) Name the different arrangements of chains and describe them!
4) Each arrangement has advantages as well as disadvantages, please list them!
5) Follow the stream of material in the kiln and define which chain arrangements can be
used for the individual parts of the system! Explain why!
6) Explain the reverse sense of hanging garlands! Compare it to the sense of hanging
thermochains
7) Would you recommend garlands for the downstream (hot) end of the system,
thermochains for the upstream (cold) end? Please explain why!
8) What are the specific advantages of round link chains?
9) Describe the material and gas temperature profile along the system!
10) What portion of heat resistant steel chains would you recommend for a system? Define
the main criteria for estimating this amount!
11) Which basic types of hangers do you know? In which case would you prefer single
chain hangers?
12) Name the main characteristic data of a chain system: Length, total weight and total
surface area of chains?

Chapter 3
Refractory Linings

Refractory Lining of Cement Kiln Systems
1. IMPORTANCE OF REFRACTORIES FOR CEMENT PRODUCTION ........................... 207

2. SUBDIVISION OF THE BURNING PROCESS AND SELECTION OF REFRACTORY
QUALITIES ..................................................................................................................... 208
2.1 Drying Zone (applies only for wet and semi-wet process kilns) ........................ 209
2.2 Chain Zone........................................................................................................ 209
2.3 Preheating Zone................................................................................................ 210
2.4 Calcining Zone .................................................................................................. 210
2.5 Transition Zone ................................................................................................. 210
2.6 Sintering Zone................................................................................................... 211
2.7 Cooling Zone..................................................................................................... 212
3. IMPORTANT FEATURES OF REFRACTORIES INSTALLATION ................................ 213
3.1 General ............................................................................................................. 213
3.2 Brick Joints and Jointing Materials.................................................................... 216
3.3 Thermal Expansion Compensation ................................................................... 217
3.4 Lining Methods.................................................................................................. 221
3.5 Stripping and Removing of Old Lining............................................................... 230
4. LINING OF PREHEATERS, KILN HOODS AND COOLERS ......................................... 232
4.1 Suspension and Grate Preheaters.................................................................... 232
4.2 Kiln Hood........................................................................................................... 233
4.3 Clinker Coolers.................................................................................................. 233
5. OPERATIONAL ASPECTS REGARDING KILN LININGS............................................. 235
5.1 Heat-Up of Rotary Kiln Systems ....................................................................... 235
5.2 Kiln Shell Deformation....................................................................................... 239
5.3 Fluctuating Process Parameters ....................................................................... 241
6. CONCLUSIONS .............................................................................................................. 242
7. TEST QUESTIONS ......................................................................................................... 243

SUMMARY
Since the sole costs of refractory materials per ton of clinker produced play only a minor role
in the overall manufacturing costs, the prior task of refractory lining optimization is to prolong
the useful life of the installed materials which fact in turn increases the kiln availability for
production.
Selecting of the appropriate refractory qualities for the different kiln zones is of particular
importance in various problem areas of the kiln system. As an example, alkali corrosion in
calcining zones may be reduced by selecting acidic types of bricks, whereas eutectic
reactions in the safety zone generally can only be countered by installing basic types of
refractories.
Apart from selecting the correct refractory materials, the work quality of lining installation is a
key element in obtaining a long refractory life. Tight lining and adequate joint dimensions are
important features.
The operational influences on refractories materials, the work quality of lining installation is a
key element in obtaining a long refractory life. Tight lining and adequate joint dimensions are
important features.
The operational influences on refractories performance are of thermal, chemical and
mechanical nature. By considering an adequate heating-up procedure, thermal and
mechanical damage can be avoided during start up. Measures to provide an optimum raw
material composition on one hand and to avoid kiln feed fluctuations on the other hand may
be necessary to decrease wear of chemical nature and to improve coating stability. In the
area of kiln tyres, mechanical destruction of bricks can be caused by excessive kiln shell
deformation. Continuous control of the mechanical condition of the kiln shell in the tyre areas
allows to detect possible deterioration and to take appropriate measures in an early stage in
order to prevent damage to the lining from this side.
Refractory lifetime is generally not a matter of one single criterion. The influences described
in this paper are nearly always jointly responsible for the results achieved and should be
considered as an entirety.

1. IMPORTANCE OF REFRACTORIES FOR CEMENT PRODUCTION
Cement plans represent quite big amounts of invested capital which fact demands for a
production as continuous as possible in order to guarantee a corresponding return.
The sole cost of the refractory lining of a cement kiln, including the outlay for removal of the
old and installation of the new materials amounts to barley 2 to 4% of the cement production
costs. The losses caused by interruption of production, on the other hand, are already higher
than the annual lining costs when the kiln has been idle for 10 to 15 days. Improvement or
optimisation of the useful life of refractories with the object of impairing the availability of the
kiln as little as possible by shut-down for relining, is therefore one of the major objectives for
cement producers.
The durability of refractory linings is mainly influenced by three factors:
The choice of the quality of materials employed in the various zones.
The installation of the lining with due consideration of the methods of placing, size of
joints and jointing materials.
Due attention to operational criteria which affect the durability of the lining, namely the
correct procedure for heating-up and for cooling down of a kiln system on the one hand
and the minimisation of process fluctuations in order to maintain the continuous
operation of the kiln on the other hand.
This paper contains accordingly remarks and information on how to consider the above
criteria in order to improve refractory performance and kiln availability.
For types and classification of refractory materials and other aspects related to materials
technology, reference is made to the MA Cement Course Documents Vol. 2.

2. SUBDIVISION OF THE BURNING PROCESS AND SELECTION OF
REFRACTORY QUALITIES
With regard to the requirements on refractory materials, the kiln system can be subdivided
into various zones according to specific operating conditions (Fig. 1). The designation of the
various zones refers to the respective reactions in the burning process which, however, are
anything but clearly defined as they overlap in both directions. The point, at which the
change from one refractory quality to another is made can usually be determined only by
observing the behaviour of the lining in operation.
Fig. 1
The main points of the following explanations are also summarised in Table 1.

Table 1
2.1 Drying Zone (applies only for wet and semi-wet process kilns)
In the drying zone, the water content of slurry or nodules is evaporated. This reaction is
almost terminated at material temperatures of 100C.
In wet process kiln, the drying zone is generally a part of the chain zone. In grate pre-heater
kilns, drying takes place in the first pre-heater chamber.
2.2 Chain Zone

In wet process kilns, the chain zone can be subdivided into a drying and a preheating zone.
The material temperature at the outlet of the chain zone depends on arrangement and
qualities of chains as well as type of system, i.e. nodule or dust kiln, and is in the order of
magnitude of 200 to 400C.
The length of the chain zone ranges from 7 to 9 kiln diameters in wet kilns and from 3 to 5
kiln diameters in long dry kilns.
The lining of the chain zone should be of a dense, low alumina firebrick with high abrasion
resistance to withstand the abrasive action of the chains.
An air setting mortar should be used as there is insufficient temperature to vitrify a heat
setting mortar. In wet kilns, furthermore, the mortar must be water resistant.
Since the arrangement of the chain hangers requires extensive modification work,
sometimes dense fireclay castables are installed, which, however, should be carefully
compacted in order to obtain high abrasion resistance.

2.3 Preheating Zone
In the preheating zone, hydrate-water is driven off and the raw material is heated up to
approx. 700C. The length of the preheating zone in long kilns is approx. 4 to 8 kiln
diameters (excluding chain zone). In short pre-heater kilns the preheating zone is a part of
the pre-heater.
Regarding the lining of preheaters, reference is made to paragraph 4.1: Lining of
Suspension and Grate Preheaters.
The preheating zone of long kilns is usually lined with low alumina firebricks or, for better
thermal insulation, with lightweight firebricks. With respect of lining stability, two layer lining
is generally not recommended in rotary kilns of more than 3.5 m diameter.
2.4 Calcining Zone

The calcining reaction starts already at material temperatures below 600C and is completed
at approx. 1200C. Since, however, the most part of calcination takes place between raw
material temperatures of 700 to 900C, usually this range is designated as calcining zone.
The length of the calcining zone in long kilns is in the order of magnitude of 6 to 9 kiln
diameters. In short pre-heater kilns calcining takes partially place in the pre-heater. There,
the length of the calcining zone in the kiln amounts to approx. 4 to 6, in kilns with
precalciners 1 to 3 kiln diameters.
From the refractory point of view, the calcining zone can still be lined with normal fireclay
bricks or, for better insulation, with fireclay lightweight bricks. In case of alkali attack good
operational results have been achieved with acidic light weight bricks with an SiO2 content
above 65%. These bricks form with alkalis a vitreous layer of 2 to 3 mm thickness on the
surface, which prevents the further alkali infiltration.
Two layer linings are generally not recommended. There are, however, two layer bricks on
the market, consisting of a dense working part and a porous insulating part. Such bricks are
generally installed for heat saving purpose, if lightweight bricks show unsatisfactory
operating results.
2.5 Transition Zone

The transition zones are located on both sides of the sintering zone. Since the length of the
sintering zone varies with process fluctuations, the transition zones are characterised in
particular by unstable coating formation.
Usually, the inlet-side transition zone is further subdivided into a colder section, so-called
safety zone, and into a hot section, the proper transition zone.
The safety zone has a length of approx. 2 kiln diameters and is usually lined with alumina
rich bricks with Al2O3 content of 50 to 60%. The bricks must have good thermal shock
resistance and low porosity in order to have good resistivity against chemical attack. The
application limit of alumina rich or high alumina bricks in the transition zone is generally
determined by eutectic reactions in the system Al2O3 - CaO - SiO2 or by alkali-spalling. In
such cases, lining lifetime can be improved only by installing basic bricks.
The transition zones are often exposed to considerably higher loads than the sintering zone
itself. Quite often no or only unstable coating is formed. Thus, the bricks are exposed directly
to the flame radiation and considerable temperature variations. The lengths of the transition
zones vary from 2 to 4 kiln diameters.They are usually lined with chrome free magnesia-
spinell bricks produced of very pure, synthetic materials or with magnesia-chrome bricks
containing approx. 60 to 70% MgO.

2.6 Sintering Zone
Although this zone is often referred to as burning zone preference is given to the expression
sintering zone on account that this better describes the mechanism of reactions taking place.
The sintering zone is generally covered with a more or less stable coating, formed with
clinker and liquid phase. Liquid phase starts to form at raw material temperatures above
approx. 1250C. However, as the lining surface temperature is higher than the one of the
raw material, coating formation starts already at raw material temperature above 1050 to
1150C.
The term sintering zone can also be explained as zone of increased material diffusion with
formation of C3S modifications, the latter starting at temperatures above approx. 1100C.
The maximum material temperature in the sintering zone is in the order of magnitude of
1400 to 1500C at the beginning of the cooling zone.
The length of the sintering zone is generally between 3 to 5 kiln diameters and depends
largely on the shape of the flame and type of fuel. Coal flames give generally short, oil
flames medium and gas flames longer sintering zones. In kilns with precalciner, the sintering
zone length amounts to 5 to 8 kiln diameters due to the higher specific material throughout.
The bricks in the sintering zone are exposed to chemical attack by the liquid phase of the
clinker and by alkali sulphates, high temperatures and, depending on coating stability, high
thermal shocks. These conditions are best met by basic bricks due to their high
refractoriness and good resistance against chemical attack.
Thus, generally, chrome free magnesia-spinell bricks, magnesia-chrome or dolomite bricks
are installed. When using chrome-free magnesia-spinell bricks, qualities particularly
developed to improve coating adhesion should be chosen (qualities based on natural sinter).
Dolomite bricks have generally good operating performance in zones with coating formation.
The price for dolomite bricks is only approx. 60% of that of magnesite. A disadvantage of
dolomite bricks is, however, its sensitivity to moisture. Thus, for longer storing times, these
bricks are to be stored air-tight. During longer kiln stops the dolomite lining is to be protected
against moisture by tightening the kiln tube and putting a hygroscopic agent (e.g. burned
lime) in the kiln.
Raw material analysis and tests can give some indications on selection of brick quality to be
installed. The final decision, however, is often to be based on operating experience, i.e. by
trial and error.
Insulation of the burning zone with insulating back-lining is generally not recommended. By
insulating, basically the hot face temperature of the bricks will be increased, resulting in
reduced coating thickness and higher chemical and thermal load of the bricks. Furthermore,
two layer lining is less stable and leads, particularly in big kilns, often to early failure due to
relative movement and loosening of the lining.
In cases, where a coating does not form, insulation may be helpful in reducing heat losses
and protecting the kiln shell, particularly in the tyre area. In such specific cases basic bricks
with back-linings of 40 mm hard fireclay slabs are sometimes installed. However, generally
installation of two layer linings is not recommended due to reduced lining stability.

2.7 Cooling Zone
The cooling zone in the rotary kiln reaches approx. from the burner nozzle to the kiln outlet.
In this zone, the clinker is cooled down from its maximum burning temperature of 1400 to
1500C to approx. 1350C in kilns with grate, rotary or shaft coolers and to approx. 1250C
in kilns with planetary coolers.
The cooling zone in kilns with planetary coolers has a length of approx. 1.5 to 2 kiln
diameters. It consists generally of a cam lining for better cooling efficiency and a dam ring
for equalising of clinker distribution to the individual cooler inlet openings. For camlining and
damring, dense high alumina bricks with approx. 80% Al2O3 and considerable thermal shock
resistance are generally used. The backing of the damring can be made of hard fireclay
bricks with adequate mechanical strength.
The length of the cooling zone of kilns with grate, rotary or shaft coolers is generally 0.5 to 1
kiln diameter. It is usually lined with dense, abrasion resistant high alumina bricks containing
80% Al2O3. In case of heavy chemical attack it can be necessary to line this zone with basic
bricks, which however, should have high thermal shock resistance (magnesia-spinell bricks).
Since the outlet zone is often free of coating, kiln shell temperature will increase due to the
higher thermal conductivity of basic bricks. Due to the high thermal and mechanical load of
the bricks in this zone, insulating back-lining is not recommended.
The end of the cooling zone, the nose ring, is one of the most critical points of cement kiln
linings. Its lifetime is often lower than the one of high loaded sintering zones. Shape and
quality of nose ring bricks requires therefore careful consideration.
Basically, complicated special shapes should be avoided since special shaped bricks are
often hand-made and can have a much poorer quality than machine made bricks of the
same composition. Nose ring design according to Fig. 2a and 2b would allow to use
standard brick sizes with minor modifications and is to be preferred against the design
according to Fig. 2c.
Fig. 2a Nose Ring Design Using Bricks

Fig. 2b Nose Ring Design Using Bricks
Fig. 2c Nose Ring Design Using Bricks
3. IMPORTANT FEATURES OF REFRACTORIES INSTALLATION
3.1 General
Depending on specific requirements and local conditions, refractory brick linings in rotary
kilns are installed dry or with mortar. For specific applications, unshaped materials are used,
which are installed by casting, gunning or ramming.

Apart from selecting the correct refractory materials, the quality of the installation procedure
is a key element in obtaining a durable refractory lining and long service life.
High refractory lining quality requires:
Tight installation of brick
Correct compensation for thermal expansion with adequate joints
Selection of suitable mortar for mortar joints
Selection of the optimum lining method, so that kiln rotation in the cold state can be
minimised
Loose linings may result in spiral displacement of entire lining sections (see Fig. 3). The
bricks jam in a twisted position and can no longer expand freely when hot. Result: spalling
due to excessive mechanical stresses.
Fig. 3
The lining ring must run true to the vertical kiln axis. Using a welding seam as reference,
alignment lines can be drawn on the shell at regular intervals.
Installation of closure bricks requires special care.

Procedure (Fig. 4)
1) Tension the ring with a hydraulic spreader jack.
2) Insert the most tightly fitting combination of key bricks
3) With basic lining: drive in key plates between axial joints of the ring until ring becomes
completely stable. Use only key plate per joint.
With linings containing Al2O3: closing with key plates should be avoided because iron
reacts with Al2O3 at high temperature to form a compound with a low melting point. For
this lining quality, the combination of key bricks must fit particularly well.
Fig. 4
For closing lining rings, key bricks of various dimensions are available which can be
combined to obtain optimum closures. If necessary, standard formats can be cut to the
desired widths. The brick lengths of the last lining ring generally will have to be cut as well.
Rotary kiln bricks may only be cut to 2/3 of their original length or width; if necessary two
rows must be cut.

Offset brick linings as shown in Fig. 5 may not be installed in rotary kilns. Each ring must be
self-supporting in order to avoid shearing stress occurring as a result of relative lining
motion.
Fig. 5
3.2 Brick Joints and Jointing Materials

Bricks laid in mortar generally give better performance than dry placed bricks provided that
the mortar used is of the correct quality. Mortar joints must be even and not thicker than 1.5
to 3 mm, preferably 2 mm.
To achieve the desired lining quality with mortar, qualified and experienced bricklayers are
required, who regrettably are becoming a rarity in many countries.
In contrast, dry bricking can be executed in less time by less specialised personnel.
Dry lining is therefore the standard procedure for basic bricks. They are currently almost
exclusively laid dry with or without steel plates in every radial joint. The steel plates react at
high temperatures with the magnesia brick to form a highly refractory combination of
magnesiaferrite which is claimed to give increased stability to the ring. Clench lining (dry
lining without steel plates) is becoming more and more popular. Its main advantages are:
tighter lining

no handling of steel plates
no separation of steel plates required when recycling old bricks
Expansion allowances in axial and circumferential joints of the brick linings are sometimes
necessary in order to compensate for the thermal expansion of the brick. With insufficient
compensation of thermal expansion, the bricks can be mechanically overloaded at operating
temperature. Excess compensation conversely leads to loose lining or even lining
displacement and collapse of brick rings. Therefore the correct dimensioning of thermal
expansion allowance is a most important feature of ensuring a stable lining.
3.3 Thermal Expansion Compensation

The thermal expansion of a refractory lining (re: Fig. 6 and 7) is partially compensated by the
following factors:
thermal expansion of kiln shell
burning out of glue if sheet metal shims are glued to the brick surface
lining inaccuracy
compressibility and pyroplastic behaviour of the bricks

Fig. 6

Fig. 7

The thermal expansion which cannot be absorbed by the above factors has to be
compensated by:
the elasticity and pyroplasticity of mortars
the elasticity and softening of steel plates
combustible materials (cardboard shims)
The thermal expansion of refractory linings is strongly dependent on materials and operating
conditions. For the correct dimensioning of the expansion allowance, the recommendations
of the manufacturers of each brick type should be strictly followed.
For dry lining consisting of basic bricks, most manufacturers recommend a longitudinal
expansion allowance of 1%, i.e. installation of a 2 mm cardboard in each circumferential joint
for 200 mm brick length (for designation of joints see Fig. 8). Most qualities of basic bricks
are normally laid without cardboards in the axial joints in order to achieve the tightest
possible lining at working temperature.
Fig. 8

For some particularly dense, high fired qualities additional expansion allowances within the
rings are recommended. In addition to the normal steel plate, a 1 mm thick, 50 mm wide
cardboard is to be installed near the hot face of the axial joints after every 4th to 8th brick.
To prevent the cardboards from slipping-in, they are bent at right angles to the brick face
(see Fig. 9).
Fig. 9
If basic bricks are laid in mortar, additional expansion allowance in the axial joints is not
necessary. In the circumferential joints an allowance of about 0.2% is made, i.e. a 2 mm
cardboard is inserted between each fifth ring without mortar.
Alumina and fireclay bricks as a rule are installed without expansion allowance in the interest
of a stable lining. This is possible due to the lower thermal expansion of these bricks
compared to the basic materials and due to good pyroplastic properties which permit
accommodation of the stresses at high temperatures.
Steel plates are not used with Alumina bricks, since steel and aluminium silicates form low
melting compound at high temperature, which can lead to lining damage.
3.4 Lining Methods

The following factors have to be considered for kiln lining methods:
safety
tight installation of the bricks
time required for lining
The methods applied can be generally classified as:
rotating methods, i.e. the kiln has to be turned during lining
formwork methods: the bricks are installed on a curved formwork without turning of the
kiln
3.4.1 Rotating Methods

With the rotating methods, kiln zones of 5 to 20 m length are lined at one time. Work is
carried out only in the bottom of the kiln, which is rotated intermittently as the work of lining
proceeds. The bricks must be fastened to the shell, e.g. by means of screw jacks or glues,
as the ring of bricks is rotated into the overhead position before being closed.
3.4.1.1 Screw-Jack Method (Fig. 10)

The lining is fastened by means of screw-jacks, usually of steel. The bricks are clamped in
position by means of wooden wedges. The screw-jacks should be furnished with pressure
control facilities as e.g. discsprings or load cells. The screw-jack method should not be
employed for kilns more than 4 m in diameter.
Fig. 10

Lining Procedure
1) Lining of slightly more than the bottom half of the kiln.
2) Fastening of the lining by means of screw-jacks and wooden wedges. The distance of
the screw-jacks is 0.5 - 1 m, depending on kiln diameter.
3) Turning the kiln by one quarter of circumference.
4) Lining of the third quarter.
5) Fastening according to point 2.
6) Turning of the kiln until the last section is on the bottom
7) Lining of the last section and keying with closure bricks.
Disadvantages of the Screw-Jack Method
Kiln has to be turned during lining.
With large kilns there is a danger of the screw-jacks slipping and the shell being
distorted.
Average Lining Speed with the Screw-Jack Method
1 to 3.5 m2/h or 0.17 to 0.25 m2/man-hour
3.4.1.2 Gluing Method

With the gluing method the bricks are glued to the shell in an alternative manner, following a
specific pattern according e.g. Fig. 11.

The glues used are two component expoxy or polyurethane resins and for very low
temperatures down to minus 25C there are synthetic resins based on polyesters or
polyacrylates.
Fig. 11
The glues decompose after heating up above 200C and lose their effect.

Lining procedure:
1) Cleaning of the shell down to the bright metal, if necessary with the aid of a grinder. The
surface must be free from dust or grease.
2) Gluing of the first lining sector of approx. 5 brick rows.
3) Installation of the next sector without glue etc. The number of bricks between the glued
sectors depend on brick weight, kiln diameter, glue properties, etc. Generally the brick
and/or glue suppliers specify the gluing pattern. As a rule approx. 20% of the bricks laid
are glued. Glue consumption amounts to approx. 0.4 to 0.8 kg/m2 of lined area or
approx. 2 kg/m2 glued area.
Using the gluing method, the following points require careful consideration:
The start to be glued must be fitted together before the adhesives start to set.
The kiln should not be turned before the glue has developed sufficient strength.
The glue must be carefully prepared according to the suppliers specifications, particularly
with respect to mixing ratio.
Disadvantages of the gluing method:
Kiln has to be turned during lining.
Tightening of the brick rings before keying is not possible, but only the bricks between
two glued sectors. Thus, lining mistakes are not easy to be localised.
Safety risk if kiln is turned before glue has developed sufficient strength.
With the gluing method, average lining speeds of 3 to 7 m2/h or 0.4 to 1.2 m2/man-hour can
be achieved.
3.4.2 Formwork Methods

The formwork methods can nowadays be considered as the standard lining method. They
are safe in operation and permit very accurate installation of the bricks.
With the curved formwork methods, first the lower half shell of the kiln is lined without any
particular aid, and then, the upper half ring by ring with the aid of the curved formwork. There
is a wide choice of equipment, form simple wooden forms to hydraulically operated
mechanised forms.
3.4.2.1 Pogo Stick Method

The main equipment for the pogo stick method is:
supporting construction
working scaffold
wooden curved drum
spring loaded pogo sticks
The working scaffold is installed upon the lined bottom half of the kiln. Then the bricks of the
upper half section are installed and fixed by means of the pogo sticks. Generally two
bricklayers and two helpers line from each side towards the top position. After keying, the
pogo sticks are removed and the formwork is put into the new position.

3.4.2.2 Wooden Curved Form Method
The main equipment for the wooden curved form is:
supporting framework
working platform
wooden curved formwork
wooden wedges
The supporting framework is so constructed to permit a fork lift vehicle to move beneath the
working platform, lifting the brick pallets to the platform. It is installed on the lined bottom half
of the kiln. The bricks of the upper section are then laid on the wooden arch and pressed
against the shell by means of wooden wedges. Work is done from both sides towards the
top by two men on each side. After keying of a ring the curved formwork is moved into the
next position, thereby, normally the wooden wedges are loosening themselves.
3.4.2.3 Mechanised Curved Forms (Fig. 12)

Mechanised curved forms are mechanised variations of the simple arch forms. They
basically consist of a travelling working platform on a supporting framework. The curved form
is installed on the working platform and is equipped with mechanically or pneumatically
operated pressure tools by which the bricks are pressed against the shell. The supporting
frame is so constructed as to allow the passage of a fork lift vehicle.
The most used mechanised lining machines are the Mult-O-Ring, the DAT-Appartus and the
Reintjes lining machines. The Mult-O-Ring is equipped with two parallel rows of pressure
tools which almost doubles the speed of work.
The lining procedure is similar to that with simple curved forms, except that the fastening of
the bricks is mechanised.
With mechanised lining machines very high lining speeds of up to 6.5 m2/h or 0.5 m2/man-
hour for mortared linings and up to 9.5 m2/h or 0.75 m2/man-hour for dry lining can be
achieved.

Fig. 12

3.4.3 Lining with Unshaped Materials
Unshaped refractory materials can be installed by casting, gunning or ramming. The
installed masses are fixed by metallic anchors which provide a good lining stability even in
cases of local damages or spallings. Usually V-type anchors, generally delivered by the
refractory supplier, are welded in a cross pattern on the shell. About 15 to 25 anchors per m2
are required, depending on lining thickness.
3.4.3.1 Castables
Castables or refractory concrete contain a hydraulic setting agent (high alumina cement).
They are installed generally in suitable formworks, e.g. by the rotocasting method (Fig. 13)
and compacted by vibrating. In order to avoid spalling due to quick water evaporation during
heating up, it can be advisable to perforate the surface by pushing steel wires of 5 mm
thickness into the freshly placed castable at a distance of approx. 5 cm. The steel wires are
to be removed after hardening of the castables. During the hardening period just after
installation of the castable, its surface should be protected against too quick evaporation of
water by means of wet cloths or water spray.
Fig. 13

3.4.3.2 Guniting Materials
Guniting materials are supplied either with chemically or with hydraulically setting bonding
agents. The materials are placed pneumatically, passing through a hose to a mixing nozzle
where water is introduced (Fig. 14).
Fig. 14
Guniting should always be done by specialised personnel, since quality of the lining as well
as amount of losses in rebounding is highly depending on the ability of the operator.
Depending on shape and position of the working surface losses in rebound amount from 10
to 50%.
During the hardening period, the gunned surface must be protected against water
evaporation similar to castables. Also perforation of the surface by means of steel wires can
be advisable.
3.4.3.3 Ramming Material

Ramming materials are plastic or granular materials, generally chemically, seldom
hydraulically bonded.
Due to the low water content, rammed refractories generally have low porosity, high volume
stability and strength and good resistance against chemical attack.

The masses are placed by compacting layer by layer of approx. 25 to 50 mm thickness each
by means of a vibrating hammer fitted on a compressed airhammer. Appropriate and
homogeneous compression is necessary for good operating performance of rammed
materials, therefore, highly qualified personnel is required for placing.
Chemically bonded materials should be installed only shortly before heating up. Spare parts
lined with chemically bonded rammed materials must be tempered at approx. 250C. After
tempering, the lining surface must be protected against humidity by bituminous paint, in
order that they are not damaged by absorbing water during storage.
3.5 Stripping and Removing of Old Lining

Stripping and removing of old lining and coating manually requires approx. the same time as
the relining of the same zone with new bricks. By use of mechanised stripping machines,
this dangerous and time consuming operation can be done much safer in considerably
shorter time with less personnel.
Manually, 6 to 7 men remove approx. 3 to 5 m2/h of old lining. By means of mechanised
stripping machines, 2 to 3 men remove approx. 15 to 18 m2/h.
The dismantling unit consists generally of a drilling machine chassis with an all-round
pivoting arm on the end of which a compressed airhammer or drilling device is mounted (Fig.
15). For removal, small front-end-loaders are used. The shovel is shaped to fit the curve of
the kiln shell.

Fig. 15
The stripping procedure is generally done in the following steps:

1) If the coating is more than 25 cm thick it should be cleared away before the actual lining
is to be opened up.
2) In small kilns of < 4 m diameter the removed coating must be transported away before
the stripping of lining starts.
3) The process of opening up a slit should be done very careful in order not to damage the
steel shell.
4) When the lining has been opened up, actual stripping can commence. Due to the
opening already made in the lining, the stripping tool can more or less work parallel to
the shell, thereby causing no damage.
5) After the upper half of the kiln is stripped, the kiln is rotated so that the remaining
brickwork fall down.
6) The rubble can now be carted away by means of the front-end-loader.

4. LINING OF PREHEATERS, KILN HOODS AND COOLERS
4.1 Suspension and Grate Preheaters

The lining of preheaters should have good insulating properties, therefore, basically two or
three layer linings are installed. The working lining generally consists of abrasion resistant
fireclay bricks with low porosity. In case of alkali attack acid firebricks with SiO2-content
above 65% are installed. Monolithic linings are used particularly for complicated shapes
where the installation of bricks is difficult and expensive. For plane walls and gas ducts lining
with bricks generally is easier and cheaper than lining with unshaped materials.
A three layer insulating lining is generally composed of a backing of ceramic fibre board, an
intermediate layer of insulting firebricks and the dense working lining.
The bricks of the working lining are usually placed with chemically setting mortars in order to
obtain gas tight linings. For compensation of thermal expansion, expansion joints are
installed horizontally and vertically at distances of approx. 2 m, as well as in corners. The
expansion joints have a width of approx. 15 to 20 mm and should be staggered and filled
with ceramic fibre mats in order to avoid penetration of dust (Fig. 16).
Fig. 16

For stabilising of the linings, anchor bricks and holding irons which are flexibly fitted with the
steel shell are used.
Roofs are either lined vaulted without use of anchors or flat with hanger bricks and holding
irons. Suspended roofs are often lined combined with hanger bricks and unshaped
materials.
Pre-heater cyclones usually are lined with bricks and insulating back-lining, except the
complicated shaped parts as e.g. cyclone inlet and control openings, which generally are
lined with unshaped materials.
Meal chutes should be lined with highly abrasion resistant materials. Usually dense
castables with metallic anchors or prefabricated materials are installed.
4.2 Kiln Hood

For kilns equipped with planetary coolers, refractory lining in the hood section is limited to
the hinged or otherwise movable kiln door with openings for burners, measuring instruments
and observation. The most frequently used lining material is dense castable fireclay together
with insulating backing.
The lining of the kiln hood of kilns equipped with grate or rotary coolers generally consists of
a dense working face and an insulating backing. The working lining must possess high
thermal shock resistance and abrasion resistance. In the upper region, where the lining is
exposed to direct flame radiation, alumina refractories with Al2O3-contents of up to 70% are
used. The lower region is usually lined with dense fireclay. As insulating linings conventional
insulating bricks as well as insulating gunning mixes or castables are used.
4.3 Clinker Coolers

(Fig. 17)
Approximately 15 to 35% of the length of planetary coolers is refractory lined. This lining
should have a high resistivity against thermal shock and abrasion. For the cam lining section
dense, high alumina bricks with high hardness and strength are used. The inlet bends are
usually lined with dense high alumina, or even pure Corundum castables. For simpler
geometric shapes conventional bricks of suitable quality can also be used.
In rotary coolers, about 50% of the tube length is lined with refractory materials. In the inlet
zones, similar as in planetary coolers, refractories of high strength fireclay are normally
installed.
The refractory lining of grate coolers, with the exception of the areas immediately above the
grate (where direct contact with clinker occurs), is exposed to far a lower extent to thermal
fluctuations and abrasion. Normally a two or three layer refractory wall and roof is installed,
consisting of an insulating rear and a dense fireclay working lining.
The wear zones immediately above the grate consist of particularly abrasion resistant bricks
or castables.

Fig. 17

5. OPERATIONAL ASPECTS REGARDING KILN LININGS
5.1 Heat-Up of Rotary Kiln Systems

The heating-up program, which mainly determines the amount of kiln temperature increase
per unit of time and the timing of the raw material feed, must consider a number of factors
such as type of refractory material, design of kiln system, mechanical systems etc.. The
optimum heating program constitutes a compromise which best satisfies the various
requirements.
Optimum Heating Period:
Refractory Materials
Material quality
Type of bond
Installation method
Riding Ring
Riding ring clearance
Riding ring dimensions
Monitoring capabilities
Exhaust Gas Flow
Temperature profile in kiln
Permissible temperatures
Environmental regulations
5.1.1 Requirements from a Refractory Materials Viewpoint

The minimum heating-up duration from the refractory materials viewpoint is given by the type
of special bricks and castables installed. Generally, the heating-up specifications of the
suppliers must be respected. The minimum heating-up time is indicated by that material
which requires the longest heating-up period because the minimum heating-up time can
generally be extended but not shortened without causing damage. Exception: basic brick
linings tend to hydrate when kept at temperatures between 100 and 450C for several days.
Generally, the heating-up time is influenced by the following materials:
Castable refractories with hydraulic bonding (Fig. 18)
Approx. 24 hrs setting time at room temperature
Approx. 24 hrs drying time at 100 to 200C
Heating-up at the rate of 25C/h up to 500C (expels crystal water)
In subsequent heating-up cycles, operational temperatures may be reached more
rapidly. The speed of reaching operational temperatures might be limited by other
factors as e.g. the tyre situation.
Castable refractories with chemical bonding
Approx. 8 to 16 hrs drying time at 100 to 200C
Heating-up at the rate of 20 to 30C/h up to 700C (until chemical bonding is
completed)
In subsequent heating-up cycles, operational temperatures may be reached more
rapidly, as explained above.

Untempered, phosphate bonded bricks
Heating-up to 700C (termination of chemical bonding) at the rate of 30C/h
In subsequent heating-up cycles, operational temperature may be reached more
rapidly, as explained above
Jointing materials (mortars)
Theoretically the same rules would apply as for casting mixes; however, the situation
is not critical because of the limited joint thickness.
Fig. 18
Based on the above listed criteria, most suppliers of refractory materials recommend for the
start-up of new kilns, and after major overhauls with monolithic linings, a maximum heating-
up rate of approx. 25C/h, in the sintering zone, up to a sintering zone temperature of 900C.
(Whereby a previous drying of heat exchanger systems and monolithically lined kiln zones is
assumed.)
After reaching 900C, heating -up can continue at the rate of 50C/h up to working
temperature.
After minor repairs and after shut-downs not caused by defective lining, refractory materials
suppliers generally recommend a heating-up rate in the sintering zone of 50C/h.

An example of heating-up scheme is shown in Fig. 19.
Fig. 19
Faster heat-up times are sometimes reported. Some Japanese suppliers heat-up new kilns
in 5 to 8 hours without any damage to lining and kiln system. Such procedures can,
however, not be recommended in general, i.e. without considering the entire situation of
lining quality and mechanical details of kiln design. Since expansion joints are dimensioned
to produce optimum lining stability under standard working conditions, rapid heat build-up
may cause spalling due to excessive mechanical stress of the brick surface because the
lining face heats-up more rapidly than the kiln shell.

5.1.2 Riding Rings (Floating Type)
From a riding ring point of view, the minimum heating-up period is determined by the time
required to stabilise the temperature difference between the kiln shell and riding rings. The
shell will heat-up more rapidly than the tyre. The critical temperature difference results in a
constriction of the kiln shell, which will cause permanent deformation if the yield strength of
the kiln shell material is exceeded. This in turn will cause excessive play in the riding ring
after the normal working temperature is regained, as well as increased ovality of the kiln
shell, a factor which may contribute to excessive refractory lining wear.
As a general rule, the riding ring creep, i.e. the relative motion between riding ring and kiln
shell should be monitored at regular intervals or, even better continuously, at least during the
heat-up cycle. Various measures may be taken to protect the riding ring section from
overheating.
With imminent danger of seizure, the heating-up process should be slowed down or
interrupted until a measurable amount of relative movement is again present. For this
reason, the tyre creep may become the limiting factor in determining the heating-up rate.
5.1.3 Exhaust Gas Control

Temperature Gradient in Kiln System
Already during the heating-up process a temperature gradient is sought which at the start of
the raw meal feeding to the kiln is equivalent or similar to the temperature gradient prevailing
under normal operating conditions. This condition can be approximately reached under ideal
heat-up conditions and with correct flame patterns, because in a dynamic heating process,
the cold kiln system is heated along the flow direction of the exit gas while the exit gases are
simultaneously cooled. With optimum fuel addition, this procedure is ideally continued with
material feed by preheating the raw meal in the counterflow. In practice, however, such a
smooth transition is not possible because the amount of material fed can, for mechanical
reasons, not be reduced to indefinitely small amounts and because the heat transfer to the
meal is governed by endothermic and exothermic reactions.
For correct temperature gradients, fuel combustion at lowest possible amount of excess air
is a necessity. In this manner, the temperature gradient between gas and lining and the
dwell time of the gases within the kiln system will result in optimum absorption of the
available heat by the refractory lining.
5.1.4 Shut-down of Kiln Systems

The procedure for shutting down a kiln depends mainly on whether it is a planned shut-down
or an emergency stop due to a breakdown.
To protect the lining, the cooling rate should, if possible, not exceed the maximum
permissible heating rate, i.e. approx. 50C/h measured in the sintering zone. Excessively
rapid cooling may cause damage to the refractory lining due to thermal shock.
After stopping raw meal and fuel feed, the exhaust gas damper is closed and the exhaust
gas fan is stopped. Closing the exhaust gas path prevents gas circulation within the kiln and
the temperature profile is more or less maintained. The cooling rate of 50C/h will initially be
reached by radiation alone. Only after a dull red heat is reached in the sintering zone cooling
should continue with a light draft until the end of the cooling phase is reached.

5.2 Kiln Shell Deformation
The refractory life and coating stability in the area of kiln tyres is significantly influenced by
kiln deformation. As a result of changes in the kiln shell radius, there is a movement
between the bricks which results in local surface pressure, leading to rapid wear and
possibly total lining collapse.
The measurement of kiln shell deformation can be made during kiln operation with the aid of
the Holcim Shell-Test Gauge.
The measuring principle is illustrated in Fig. 20.
Fig. 20

The yoke (A) of the Shell-test gauge covers a circular section with a chord of length of 1 m
and height h.
During kiln rotation the chord height h changes with the continuously changing curvature
radius r of the kiln shell.
This change is transferred via a plunger (D) and recorder (C) to produce a polar diagram
(shell-test diagram).
The closed line drawn on the diagram represents a geometrically similar figure to the
deformed kiln shell cross section.
The subsequent calculation is based on determining the difference between the longest and
the shortest radius of the shell-test diagram after which the ovality is calculated according to
the method of G. Rosenblad.
The mechanical condition of kiln shell and kiln tyres with respect to allowable ovality should
be regularly checked.
For kiln with loose tyre design, an indication on the mechanical condition of the tyre section
can be gained by continuous measurement of the difference in rotational speed of the kiln
tyre and kiln shell. This method permits continuous indication/recording in the control room
(Fig. 21).
Fig. 21

5.3 Fluctuating Process Parameters
Unstable burning conditions of cement kilns belong still to the most frequent reasons of
refractory lining failures. All types of used sintering zone refractories show structural
changes between hot and cold face, infiltrated circulating elements of faded brick sections
by overheating, do not immediately result in refractory failures. However, if kiln operating
conditions change, followed by changed thermal profile or changed coating conditions, the
bricks break off in large lumps since, due to its altered structure, their mechanical
characteristics have deteriorated.
Unstable burning conditions can be caused by
Unstable kiln feed due to insufficient material flow control
Fluctuating kiln feed composition due to insufficient homogenising of raw material
Fluctuating flame length due to insufficient fuel flow control or fluctuating fuel quality
Frequent kiln shut-down due to maintenance problems of auxiliary equipment, power
failures or refractory failures.
Without going into details it is evident, that these points need to be carefully watched and, if
required, improved, in order to establish the conditions which make the improvements in the
field of refractory lining selection, installation and operation effective.

6. CONCLUSIONS
Improvement of refractory lining lifetime is not a matter of one single criterion. There exist
several fields of influence which often are jointly responsible for the results obtained.
The following list contains suggestions concerning the most important and general
consideration for achieving improvement in lining life.
Since the major refractory problems occur in the sintering and transition zone, most of the
considerations will concentrate on these kiln areas.
Most important factors:
1) Optimisation of kiln feed (qualitatively and quantitatively)
by selective quarrying procedures
by selection of optimum raw material composition
by appropriate homogenising of raw materials
by improvement of kiln feeding equipment
2) Optimisation of flame shape and length
by evaluation of the influence of adjustable flames
by the most constant burning possible with the optimally established flame shape
3) Improvement of mechanical condition of kiln shell in the tyre area
by adjustment to minimum possible tyre clearance
by continuously monitoring tyre creep
by cooling of kiln tyre areas when danger of seizure exists
4) Optimisation of brick quality in the sintering zone
by methodical investigations (trials) with alternative brick qualities
5) Thermal load reduction
by introduction of secondary firing (if applicable)
6) Reduction of the number of kiln shut-downs
by improving the mechanical and electrical reliability of the system
by installation of emergency power supply
7) Optimisation of the bricking process
by refining existing or selecting better methods
by improvement of working conditions or methods
by introducing clear instructions, control (bonus systems)
by optimised bricking policy
8) Prevention of rapid heating or cooling
by observing the important factors on kiln light-up
by prevention of rapid cooling on shut-down
This list is not claimed to be complete and some of the suggestions might not be realisable
for a specific case on hand. As a check-list, however, it allows the identification of points of
weakness and the establishment of further procedures in order to be able to define the
actual problem and finally to solve it as effectively as possible.

7. TEST QUESTIONS
1) Please make a sketch of the various zones of a suspension pre-heater-, Lepol -and a
long wet kiln and indicate what type of refractories are used in each zone.
2) Explain the different methods you know of refractory brick installation in a rotary kiln
(sketches).
3) Where are the unshaped refractories mainly used in cement kiln system?
4) What are the main features of brick joints (jointing materials) for rotary kiln linings in the
case of alumina and basic bricks, considering dry and mortar lining?
5) Establishment of a heating-up program for cement kilns is mainly based on
considerations regarding three main factors. What are these considerations?

Chapter 4
Firing Systems
1. Firing Systems: Handling and Preparation of Noble Fuels ................................... 245
2. Safe Handling of Coal and other Combustible dusts ............................................. 269

Firing Systems: Handling and Preparation of Noble Fuels
Dirk Pauling
TPT 01/21062/E
(Replaces report no. PT 98/14353/E)
1. Introduction.................................................................................................................... 246
2. Coal Firing Systems ...................................................................................................... 247
2.1 Classification of Coal Firing Systems................................................................ 247
2.2 Quality of Coal Preparation ............................................................................... 251
2.3 Pulverised Coal Dosing..................................................................................... 252
2.3.1 Storage Silo for Pulverised Coal.................................................................... 253
2.3.2 Feed Bins for Pulverised Coal ....................................................................... 254
2.3.3 Accuracy of Weighing and Dosing for Pulverised Coal ................................. 254
2.3.4 Recommended Pulverised Coal Dosing Systems ......................................... 254
2.4 Pneumatic Transport of Pulverised Coal to the Burner..................................... 256
3. Oil Firing Systems ......................................................................................................... 258
3.1 Unloading unit: Transfer to the storage tanks ................................................... 258
3.2 Fuel Oil Storage ................................................................................................ 259
3.3 Fuel Oil Preparation Unit................................................................................... 259
3.4 Fuel Oil Burner Platform Valve Train ................................................................ 260
3.5 Quality of Fuel Oil Preparation .......................................................................... 261
3.6 Control of Fuel Oil Flow Rate ............................................................................ 264
4. Natural Gas Firing Systems.......................................................................................... 265
4.1 Gas Transfer Station ......................................................................................... 265
4.2 Gas Burner Valve Train..................................................................................... 266
4.3 Safety Precautions for Gas Firiung ................................................................... 267
4.3.1 Flexible Hoses Bursting................................................................................. 267
4.3.2 Leak Tests ..................................................................................................... 267
4.3.3 Explosions in the Kiln .................................................................................... 267

1. INTRODUCTION
Noble fuels are coal (including petcoke), fuel oil and natural gas. Handling and preparation
of those fuels has to fulfil certain requirements in order to produce similar combustion
conditions of these different fuels and avoid incomplete combustion (e.g. CO formation and
local reducing conditions).
For coal firing the main types of firing systems are described (direct, semi-direct and indirect
firing). For pulverised coal handling the important design criteria are outlined. The required
coal dust preparation quality (fineness) for a good combustion in the cement kiln is
described.
For fuel oil firing the most commonly used preparation and pumping system is described.
The required fuel oil preparation quality (pressure, viscosity and temperature) is given.
For natural gas firing the preparation and safety precautions are described.
Burners, fuel injection characteristics and flames are subject of a separate paper (HMC/TPT
report no. TPT 01/21063/E; "Burners and Flames").

2. COAL FIRING SYSTEMS
2.1 Classification of Coal Firing Systems

With reference to gas and material flow, the different coal firing systems can be classified
into four main groups, which in total sum up to six individual systems:
Direct firing systems without fine coal storage
Semi-direct firing without fine coal storage
Semi-indirect with storage of fine coal
Indirect firing with storage of fine coal
Figure 1: Classification of coal firing systems

System 1 - Direct firing
Represents the most simple case. The coal is ground in the mill, dried and blown into the
kiln together with the drying gases.
Basically describes the same solution with the exception of the mill working under positive
pressure. This solution is generally applied to protect the fan when processing abrasive coal.
In system 3, the fan is protected by separating the pulverised coal in a cyclone and feeding it
after the fan into the primary air stream.
System 3a - Direct firing with recirculation
Same as system 3, but with recirculating drying air. This arrangement allows reduced
primary air ratios.
System 4 - Semi-direct firing
Has little technical significance since the solution with intermediate storage of coal would
generally be given preference.
System 5a - Semi-indirect firing
With system 5a, the kiln can be operated independently of short mill shut downs since the
pulverised coal is stored in an intermediary storage bin. The exhaust air from the mill enters
the kiln as primary air.
System 5b - Semi-indirect firing with recirculation
Same as system 5a, but with recirculating drying air. This arrangement allows reduced
primary air ratios.
System 6 - Indirect firing
In system 6, the grinding installation is completely separated from the kiln. The pulverised
coal is stored in an intermediary storage bin and the exhaust air from the mill is released
through a filter into the atmosphere. By this way, the kiln operation is completely
independent from the combined drying and grinding operation.

Major Advantages / Disadvantages of the Different Coal Firing Systems:
Direct firing Semi-indirect firing Indirect firing

Conventional modified Conventional Modified
System 1 and 2 System 3a System 5a System 5b System 6
Advantages Simple design Lower No exhaust gas, Lower primary Simple flame
Low risk of primary air therefore no air ratios and control
explosions ratios and filter required, thus lower heat Low primary air
thus lower thus lower risk consumption ratio
Simple extinction heat of explosions compared to
of fire in the in the consumption than with conventional Water vapour from
grinding system by compared to indirect firing coal drying is not
stopping coal Independent introduced into the
conventional Short mill shut primary air fan
feed. No spread of kiln
fire into silos Independent down not = kiln
primary air shutdown Short mill shut
fan down not = kiln
Only one mill shutdown
required for
several kilns Only one mill
required for
Easy sampling several kilns
for fineness
control Easy sampling for
fineness control
Disadvantages Combined More More complex More complex More complex
operation with the complex installation installation installation
kiln, therefore installation requiring requiring requiring
often not optimal requiring additionally: additionally: additionally
operating additionally: cyclones , primary air fan, dedusting filter
conditions. primary air pulverised coal longer ducting Investment cost
Mill shutdown = fan, longer silos, pulverised higher than with a
kiln shut-down ducting coal feeders, direct firing (valid
measuring and for one kiln only)
Number of kilns = control system
number of mills More vulnerable to
thus reducing the High primary air fires and
advantage of ratio up to 30%, explosions in gas
lower investment during start up ducts and filter
cost if several and shut down
of grinding plant Risk of self-
kilns are installed ignition of the
disturbed kiln
High primary air operation pulverised coal in
ratio up to 30% storage silo
Investment cost
Slow control higher than with
loops, long dead a direct firing
time system (valid
Sampling for for one kiln
fineness control only)
difficult Risk of self-
ignition of the
pulverised coal
in the storage
silo

Impact of Firing System on Kiln Operation:
Direct firing systems tend to enhance coal fluctuations and therefore disturb
combustion.
If the mill vent air enters the kiln as primary air as it is the case with the direct- and
semi-direct firing, the primary air ratio is higher than required for optimum
combustion. At a given excess air factor, the primary air ratio has a direct influence
on the heat recuperation efficiency of the cooler and finally on the overall kiln heat
consumption. If the heat consumption can be reduced, the exhaust gas quantity is
also reduced, which offers the potential for a capacity increase.
Another important advantage of a lower exhaust gas quantity is a reduced gas
velocities in the kiln. This has the benefit of lower dust generation for wet kiln
systems.
With a direct firing system the water vapour from coal drying enters the kiln with the
primary air. The water vapour increases the exhaust gas quantity and reduces the
kiln capacity accordingly.
At the same time, the flame stability may decrease as a result of dilution of the
primary air.
The flame shape is strongly influenced by the type of firing system. An indirect
system will not only support a more stable flame, but also enable a shorter flame.
This results in smaller and more evenly distributed alite crystals with a higher
reactivity. Benefits are a better quality of the clinker and a lower energy demand for
the cement grinding.
Conclusion on the different firing systems:
For new projects direct firing will not be selected anymore because of the above mentioned
disadvantages. Today indirect firing systems are "State of Technology". Conversion projects
from direct to indirect firing of existing installations can not always be financially justified on
the basis of reduced thermal energy consumption. In countries with low coal costs, pay back
times of several years must be expected. However the positive effects on kiln operation and
product quality can make a conversion project interesting.

2.2 Quality of Coal Preparation
Before the coal is fired, it has to be prepared according to the required fineness and
moisture content.
Normal coal is dried to 0.5 - 1.5 % residual moisture. Whereas high volatile coal (lignite) can
only be dried to a residual moisture of 6 - 8%. Moisture in the coal means loss of calorific
value, as the water has to be evaporated and heated up in the kiln. Coal drying is done
simultaneously with the grinding.
Inadequate coal grinding (fineness) can result in both burn-out problems (CO formation) and
the presence of fuel in the material bed (local reducing conditions leading to an increased
volatility of sulphur).
The combustion time of coal depends on the content of volatile elements.
Figure 2 shows the principal requirements for coal fineness in function of the volatile content.
Figure 2: Grinding fineness of coal in function of volatile content
The aim is to comply with the following simple rule as an upper limit:
Residue 90 m sieve for coal (% volatile components)
Residue 200 m sieve for coal 2%
For petcoke the above mentioned rule has to be tightened:
Residue 90 m sieve for petrol coke 5 %
Residue 200 m sieve for petrol coke 1 %

Anthracite has to be ground even finer than petcoke.
The upper limit for the fineness of mixtures from coal and petcoke can be calculated
according to the following example:
Example: 60% coal with 20% volatile and 40% petcoke
R90m 0.6 * 0.5 * 20% + 0.4 * 5% 8%
R200m 1%
It has to be pointed out, that both values, the residues on 90 m and on 200 m are
important. The 90 m values influence flame length and CO formation. Excess residue on
the 200m sieve create reducing conditions on the material bed and thus increased
volatilisation of sulphur.
2.3 Pulverised Coal Dosing

For a coal firing system a careful design of the entire feed system (discharge from storage
silo, weighing and dosing as well as transport to the burner) is of utmost importance.

Figure 3: Pulverised coal dosing
2.3.1 Storage Silo for Pulverised Coal

Mass-flow type silo with mechanical or pneumatic activation of the outlet(s).
The mass-flow design for pulverised coal silos includes discharge hoppers in stainless steel
with a hopper half angle less than 15 and a minimum outlet diameter of 500 mm. Since the
silo outlet cone is crucial for the dosing, the design and supply should be done by the dosing
system supplier.
The silo should not be installed in front of the kiln to prevent exposure to heat radiation.

2.3.2 Feed Bins for Pulverised Coal
The storage silo should not be abused for feeding purposes because the poor feed rate
control out of a big storage silo (>100 t). The feeding should be done out of a small feed bin
between the silo and the dosing equipment. If the feed bin is not installed due to cost
reasons, the height for a later addition should always be foreseen.
Design recommendations for the feed bin:
Capacity of the feed bin sufficient for at least 15 minutes but not more than 60 minutes of
kiln or precalciner operation
Mechanical or pneumatic activation of the bin outlet
Load cells for an in-situ loss-in-weight calibration
2.3.3 Accuracy of Weighing and Dosing for Pulverised Coal

Proper weighing and dosing requires a uniform coal dust supply (feed bin discharge; see
above). It is necessary to distinguish between accuracy of weighing and short term
variations.
The dosing system should meet the following requirements:
Weighing accuracy: +/- 2% is normally accepted.
Short term variations (referring to 10 sec. measurements): < +/- 1% (short term
variations are responsible for CO peaks)
Long term variations (approx. 10 min. to 1 hour): < +/- 0.5%
Sensibility: < +/- 0.5%
(Example: A dosing system with a maximum capacity of 5 t/h has to be capable to
handle set point changes of +/- 25 kg/h).
Adjustment range: 1:20 (of the maximum capacity).
A good indicator for the accuracy of the dosing is the oxygen level at kiln inlet. Poor dosing
of coal dust leads to big fluctuation of the oxygen concentration.
2.3.4 Recommended Pulverised Coal Dosing Systems

The following two systems can be recommended:
Pfister rotor feed scale (figure 4)
Schenk coriolis scale (figure 5)
Only second choice is:
POLDOS (weight controlled airlift from Polysius)
Not recommended anymore is:
Loss-in-weight system (e.g. Schenk Simplex; complex set-up requiring skilled
maintenance)

Figure 4: Pfister rotor feed scale
Figure 5: Schenk coriolis type feed scale

2.4 Pneumatic Transport of Pulverised Coal to the Burner
The highest accuracy of the dosing and feeding system is useless if the pneumatic transport
to the burner is not designed correctly.
Criteria for pneumatic transport of pulverised coal to the burner:
Transport velocity: 28 35 m/s; optimum > 32 m/s
The pneumatic transport velocity is one of the most critical items for regular coal flow to
avoid pulsation caused by pocket formation in the line.
Maximum length < 80 m
For length > 80 m staged diameter of the transport pipe.
Length > 100 m must be avoided.
Only horizontal and vertical transport lines
(no inclining or declining sections)
Elbow design by diversion pots (figure 6):
- Maximum 5 pots
- No pot at the first turn after the scale / dosing
Long curves lead to segregation of the coal dust through centrifugal forces. The
diversion pots have proven to be a good solution because of:
1. Low wear
2. Low pressure drop
3. Re-mixing of coal dust with the transport air at every turn
The fuel load carried by the air is not a critical value. Normal values are 3 - 7 kgCoal/m3Air.
For burning of low volatile fuels such as petcoke, low values are preferred for the
combustion.
To assess the performance of the coal dust transport line, the pressure fluctuations in the
transport line near the burner can be measured. Criteria: < +/- 5 mbar.

Figure 6: Diverting pot for pulverised coal transport (units: [mm])

3. OIL FIRING SYSTEMS
The handling of fuel oil in a cement plant can be subdivided into the following steps:
1) Unloading unit: Transfer to the storage tanks (chapter 0)
2) Tank farm for storage (chapter 3.2)
3) Preparation unit (chapter 3.3)
4) Burner platform valve train (chapter 3.4)
5) Atomisation for combustion (HMC/TPT report no. TPT 01/21063/E "Burners and
Flames")
Figure 7: Fuel oil handling
3.1 Unloading unit: Transfer to the storage tanks

For easy handling (pumping), normal heavy oil must have a temperature of about 50 to
60C.
If the oil is delivered by truck, the cool down of the oil during transport is usually small
because of the short transportation time. In this case the discharge can be done
immediately. If the cool down is too high the temperature of the oil must be increased before
discharge by heating devices (e.g. unloading heating coil).
In case of delivery by train, the wagons are usually equipped with heating coils at the bottom
to increase the temperature of the oil for unloading if necessary. The heating up is usually
done with steam.

Through the discharge filters (for pump protection) the fuel oil is then pumped to the storage
tank.
3.2 Fuel Oil Storage

Typical the storage capacities allows a kiln operation of 2 to 10 weeks. The tanks are usually
designed as welded steel constructions. Due to the fuel oil forming an insulating layer on the
walls, any particular insulation efforts are unnecessary.
In the bottom of the tank a heating coil is used to maintain the temperature of the oil above a
certain limit to enable new deliveries to be pumped into the tank.
The discharge point of the tank is equipped with a suction heater to preheat the oil to a
certain temperature (usually 50-60C corresponding to approx. 380 cSt viscosity) to permit it
to flow to the preparation unit without excessive pressure loss.
The only local heating of the fuel oil also minimises the rate of deposit forming reactions in
the tank, which doubles with each 10C increase in oil temperature.
3.3 Fuel Oil Preparation Unit

In the preparation unit the fuel oil is treated in the following steps:
Suction filter for pump protection
Pumping and pressure increase to atomisation pressure (e.g. 40 bar)
Heating to reach the required viscosity for atomisation (e.g. 130C)
Fine filtering to protect the following equipment
Due to the foreign matter that all residual oils contain and the high rate of deposits that form
at elevated temperatures (resulting in frequent maintenance) all equipment of the
preparation unit is usually duplicated and installed in parallel.
The suction filters are always required for pump protection. The installation of two filters in
parallel allows cleaning without interrupting the flow. The filters can be equipped with
differential pressure monitors to indicate the need for cleaning.
The pump station usually consists of two parallel pumps to allow maintenance on one of the
pumps without interrupting the flow. The pumps should be designed with enough capacity
reserve so that the maximum required oil quantity can still be pumped even after a long
period of operation with associated wear. The pumps are usually equipped with circulating
flow pressure regulators.
In the fuel oil preheater the heavy oil is heated to reach the required viscosity for atomisation
(chapter 3.4). Thermal oil is the preferred heating medium today. Steam or electrical heating
is used less frequently. Available waste heat can also (or additionally) be used for
preheating (e.g. by heat exchangers using cooler exhaust air).
The essential advantages of the thermal oil as a heat transfer medium is:
Operation in a constantly liquid phase (compared to steam)
Low pressures even at operating temperatures of 250 to 300C
No freezing problems
Low flammability
Thermal oils are subjected to ageing. Their quality has therefore to be checked in regular
intervals of about one year. About every five years replacement by a new charge is required.
Before leaving the preparation unit the fuel oil has to pass a fine filter (typical mesh size 0.2
mm) to protect the following equipment (burner platform valve train and atomiser).

3.4 Fuel Oil Burner Platform Valve Train
The burner platform valve train mainly consists of the following instruments:
Quick shut off valves
Pressure control valves for primary and secondary oil flow
Flow rate measurements for primary and secondary oil flow
At the burner platform valve train, the fuel oil flow is usually divided the into a primary- and a
secondary oil feed going to the burner for control of the atomisation.
An additional feature which is frequently installed is the return flow start up possibility. By
using the secondary oil for reverse flow, a low oil flow into the kiln can be achieved without
dropping the pressure too low. The return flow is only used for start up of the kiln.
Figure 8: Fuel oil burner platform valve train

3.5 Quality of Fuel Oil Preparation
For optimum atomisation, heavy oil must be heated up to ensure that the droplet size
needed for good combustion can be achieved:
Required viscosity for optimum atomisation: 12 - 15 cSt
For a typical heavy oil (type S or ASTM #6) this corresponds to a temperature of approx.
130C. High viscosity fuel oil require a temperature of up to 200C of even higher.
In today's heavy oil market, particularly in the South American OPEC countries, heavy oil is
offered which has a significantly higher viscosity than the limit specified by DIN 51 603. It is
therefore essential to keep track of the relationship viscosity - temperature and adjust the oil
temperature accordingly.
Figure 9 shows the cinematic viscosity of different fuel oil types in function of the
temperature. The upper limits for atomisation and pumping are indicated.
Figure 10 shows a conversion table for the different viscosity units.
Furthermore it is important to keep the oil temperature constant within a very narrow range
to have a stable flame.

Figure 9: Cinematic viscosity of current fuel oils

Figure 10: Conversion of different viscosity units

3.6 Control of Fuel Oil Flow Rate
Between storage tank and the burner, there are usually the following control loops installed:
Constant oil temperature at the storage tank suction point
Constant oil temperature after the preheater
Constant oil pressure after the pump (supply pressure)
Control of pressure and consequently flow rate with the primary and secondary regulator
valves (at burner platform valve train)
Whenever a kiln stop occurs, the oil lance with the atomiser head has to be cleaned
automatically by steam or compressed air in order to avoid overheating and coking of the oil
inside the atomiser. Additionally the oil lance should be retracted for protection of the
atomiser.

4. NATURAL GAS FIRING SYSTEMS
The handling of gas in a cement plant can be subdivided into the following steps:
1) Transfer station (chapter 0)
2) Burner valve train (chapter 4.2)
Figure 11: Handling and preparation of natural gas in the cement plant
4.1 Gas Transfer Station

In the gas transfer station the following equipment is installed usually:
Filtering station:
For the separation of solid and liquid impurities in the gas.
Odorising equipment:
If not already done by the gas supplier of the gas, an odorising unit can be installed to
detect leaks more easily.
Pressure reduction and preheating:
The gas distribution by means of pipelines is usually done at a pressures of 10 to 80 bar.
At the consumer site the gas pressure must be reduced to the required operational
pressure (e.g. 3 - 10 bar).
The gas must be heated before pressure reduction to such an extent that the following
temperature drop due to expansion will not cause valve internal and external ice
formation (Joule - Thompson effect: -0.3 to -0.5C/bar expansion). The temperature after
expansion should be +5C to +15C. Heating is usually done with a gas fired warm water
boiler. Thermal oil, steam or electricity can also be used for heating.
Safety shut-off valves:
To protect the plant network against excessive high pressure.
Flow rate measuring:
To record the gas quantity supplied to the plant for billing the gas deliveries.

Calorimeter:
The heat value of the natural gas can be measured continuously by means of an on-line
calorimeter if the calorific value fluctuates. This is often worthwhile because in some
cases the heat value might vary in range of 300 kJ/Nm3 from day to day.
The filtering, preheating and pressure control station are usually installed double (in parallel)
to allow maintenance of the equipment during operation.
4.2 Gas Burner Valve Train

The valve train usually consists of the following equipment:
Pressure controller:
Setting and keeping constant of the required gas pressure because the flow to the
burner must be operated at a defined, constant inlet pressure.
Safety shut off valves:
For protection against excessive high pressure.
Flow rate control valves:
To adjust the required gas flow rate.
Flow rate measurement:
For measuring the gas consumption.
Splitting of the gas flow:
In cases where the burner uses two separate gas flows for flame shaping (e.g. axial and
radial gas) the flow is divided and controlled accordingly.
Quick shut off valves:
In the event of an emergency stop of the kiln, the quick safety shut off valve (usually two
in series) are actuated to stop any further flow of gas at once.
Figure 12: Burner valve train for gas

4.3 Safety Precautions for Gas Firiung
4.3.1 Flexible Hoses Bursting

Since the risk of the flexible gas hoses between the gas supply line and the burner bursting,
pressure monitors for the maximum and minimum pressure should be installed right before
the flexible hoses.
4.3.2 Leak Tests

To check the gas pipes and fittings for leakage, the following methods are used:
Normally a leak can be detected by the smell of the gas (odoriser)
When machines are switched off, the hissing sound of the escaping gas can be heard
A further possibility is to install gas detectors in critical places such as the gas
preparation station or the burner tunnel.
A somewhat riskier method is to run a naked flame along the gas pipe. This results in a
flaming torch being produced at the leak, which cannot be overlooked. There is no risk of
this flame striking back into the supply pipe (quenching distance, lack of oxygen), but
escaped gas could cause an explosion.
When constructing buildings which contain gas pipes, it is essential to allow for sufficient
ventilation.
4.3.3 Explosions in the Kiln

The most important requirement is that the fuel should not be allowed to enter the kiln
unintentionally or at an uncontrolled rate, as this is essential to prevent explosions occurring
in the kiln itself or in the systems following (preheater tower, EP, ..).
This means that the fuel input has to be stopped immediately in the event of the flame going
out.
The extinction of the flame in a hot kiln has not been observed so far, even during material
rushes.
During the start up of the cold kiln, lifting off and extinction of the flame can occur. For
example caused by partly blocked burner outlets which lead to increased injection speed of
the gas. A too high injection velocity of the gas can blown out the flame in a cold kiln.
Therefore careful observation of the flame during the whole start up period is of utmost
importance. In the case of the flame going out, the gas flow has to be cut off immediately to
prevent explosions.

References:
1) "Firing Systems"
VA 82/4898/E
2) "Flames and Burners"

VA 93/4056/E
3) "State of Technology of Rotary Kiln Burners"

Fritz Schneider, PT 96/14078/E
4) "Proportioning of Bulk Materials"

Franz Bucher, PT 96/14071/E
5) "A Review of Coal Firing Systems and their Influence on Heat Consumption,
Production and Kiln Operation"
Hansjrg Meier, PT 96/14210/E
6) Pillard Documentation Rotary Kiln burners

Safe Handling of Coal and other Combustible dusts
F. Schneider
(Original title: Basic safety theory of solid fuel preparation)
1. INTRODUCTION ............................................................................................................271
2. BASIC PRINCIPLES FOR HANDLING PULVERIZED FUELS .....................................272
2.1 Development of Dust Explosions and Fires .....................................................272
2.2 Possible Protective Measures against Dust Explosions and Fires ..................275
3. APPLICATION OF PROTECTIVE MEASURES IN THE INDUSTRIAL ENVIRONMENT
279
3.1 Preventive Safety Measures ............................................................................279
3.2 Explosion Protection Techniques.....................................................................280
4. LITERATURE .................................................................................................................284
4.1 Approximate Values for Explosion Limits and Ignition Temperatures..............285

SUMMARY
Dust explosions can only occur when - besides certain marginal conditions - the following
factors simultaneously are present:
Stirred-up, combustible dust
Oxygen
Source of ignition
A basic distinction is made between active explosion protection techniques (prevention of
the occurrence of explosions) and design related explosion protection (reduction of the
effects of explosions).
In practice, the following measures are applied:
1) Preventive Measures:
Their aim is:
to exclude possible ignition sources within the installation
to prevent the building up of coal dust deposits, wherever possible
to detect the source of a fire as soon as possible
to extinguish the fire with a minimum of danger
2) Explosion Protection Measures:
Either active protective measures involving inert gas operation. This is the case
when the oxygen concentration within the pulverizing plant is kept below the
critical limit - for solid fuels dust, as a rule, less than 12 to 14% - as long as
combustible dust is present in the system.
Or, alternatively, design related protective measures based on the use of
explosion resistant construction in accordance with VDI Guidelines No. 3673.

1. INTRODUCTION
The operational safety of solid fuel plants is an important decision factor when the choice of
the preparation system is being considered. For this reason, the three most important basic
systems will be briefly reviewed here, differentiated according to their different methods of
handling gas and coal dust (Fig. 1).
Fig. 1 Firing Systems
a) Direct firing
In this system, the combustible dust is conveyed into the kiln together with the exhaust
gases resulting from the drying-cum-grinding operation. This arrangement represents the
simplest design and is easily controllable from the safety point of view. However, there
are also considerable disadvantages involved with the use of this system in clinker
manufacture with increasing ballast content of the fuel.

b) Semi-direct firing
In this system the combustible dust is separated in an intermediate silo, while the mill
exhausts, possibly as recirculated air, are conveyed to the cement kiln as the primary air
supply. This results in the technical disadvantages of direct firing being reduced to a
certain extent - at the expense of a somewhat more complex installation - but all drying
gases are still conveyed to the kiln.
c) Indirect firing
This solution is surely the best possible version when the operation of a rotary kiln is
being considered. The pulverized fuel can be conveyed to the firing system from the silo
independently from the pulverizing plant operation. The firing system can be operated
with a primary air ratio designed for optimal flame generation, as the mill exhaust gases
are filtered. Against this we have increased risks with respect to safety due to the
operation of the filters and silos and higher control technique requirements.
Further discussion of the decision criteria for the selection of an optimal preparation system
is outside the scope of this lecture. However, it is certain that when factors such as
the growing size of installation
installations with several firing systems
the use of fuels rich in ballast
the use of fuels of widely differing quality characteristics
are considered, the decision will be influenced in favor of the indirect firing system which
needs far more advanced and sophisticated safety techniques than the simpler direct firing
system does.
For this reason the damage prevention possibilities discussed below refer basically to the
indirect firing system and must be adjusted accordingly if they are applied to other systems.
2. BASIC PRINCIPLES FOR HANDLING PULVERIZED FUELS
2.1 Development of Dust Explosions and Fires

In order to effectively ensure the safety of a solid fuel preparation plant, we must first be
aware of the sequence of the possible fuel reactions.
Dust explosions can only occur if the following three conditions are simultaneously fulfilled
(Fig. 2).
a) Stirred-up, combustible dust present in explosive concentration.
b) Air or oxygen above the critical concentration, for coal dust as a rule, above 14%, for
lignite above 12%.
c) An ignition source possessing energy above the minimal ignition energy (depending
upon the type of dust).

Fig. 2 Preliminary Conditions for an Explosion
After the ignition of an optimally explosive mixture in an enclosed space, the pressure
increases more or less rapidly until it reaches the maximal explosion pressure Pmax, and then
decreases more or less slowly to the original pressure, depending on the aerodynamic
conditions (Fig. 3). While the maximum explosion pressure is almost independent of the
containers form and size, and in case of coal and lignite dusts, amounts to approximately
dp
seven to nine times the initial pressure the maximum rate of pressure rise - which
dt max
is a measure of the explosion violence - is dependent on the container volume in accordance
with the cubic law:
1
dp
V 3 =cons tan t K st
dt max
Kst is a material coefficient that depends on the type of dust, the degree of turbulence of the
dust/air mixture at the moment of ignition, the grain size distribution, and the type of ignition
source. The method for determining KSt is given in the VDI Guidelines No. 3673.

Fig. 3 Pressure Development of an Explosion Over Time
The degree of explosion violence of dusts is subdivided in industrial praxis into explosion
classes, whereby the explosion class and KSt are related in the following manner:
Table 1
Dust Explosion Class KSt (bar . m . s-1)

St0 0
St1 > 0 to 200
St2 > 200 to 300
St3 > 300
All types of mineral coals as well as the majority of lignites belong to explosions class St1.
Table 2 illustrates arbitrarily selected comparative values for KSt characterizing different
types of dust.
Table 2
Dust Type KSt (bar . m . s-1)

Hard coal 85
Lignite 150
Organic pigments 300
Aluminium 550
This comparison shows, that hard coal dust develops a less violent explosion than
aluminium dust.
It must be noted, that the value KSt does not allow any conclusion as regards the risk
involved with that particular dust. The main significance of KSt is for the dimensioning of
design related protective measures.

Smoldering fires, characterized by slowly smoldering combustion, can occur wherever
combustible dust is stored for a longer period of time, whereby the ignition sources can be
spontaneous combustion, initiated by external heat sources, mechanical sparks, or electrical
sparks and arcs. Combustion propagation in smoldering fires is quite possible in very low
oxygen concentrations.
2.2 Possible Protective Measures against Dust Explosions and Fires

In dust explosion protection techniques a distinction is made between active protective
measures (prevention of the occurrence of explosions) and design related explosion
protection (explosion resistant construction).
2.2.1 Active Explosion Protection

The active explosion protective techniques aim to exclude at least one of the three
preliminary conditions necessary for an explosion, i.e.:
Stirring-up of combustible dust
Oxygen content above the critical concentration of generally 12% for lignite or 14% for
hard coal
Ignition source
2.2.1.1 Ignition Source

In a pulverizing plant, ignition sources cannot be excluded with absolute certainty. It is
always possible that mechanical sparks will be generated by the action of foreign bodies or
by friction between moving machine part or that the hot gas or coal feeding system will
supply smoldering fuel particles.
2.2.1.2 Combustible Dust

It is of course impossible to replace the combustible dust with a non-combustible material in
the preparation of fuel.
Thus, the only remaining possibility is the exclusion of air or oxygen respectively, or the
reduction of the oxygen content in the fuel preparation plant.
2.2.1.3 Air and Oxygen

Dust explosions can be effectively prevented through inertization, i.e. the replacement of the
oxygen in the air by a non-combustible gas, particularly CO2 or N2, if it can be ensured that
the inert gas atmosphere will be maintained as long as combustible dust is present in the
system.
The maximal O2 concentration, below which no explosive propagation reactions of mineral
coal dust are noted, is approx. 14%, the one for lignite approx. 12%. However, this
concentration can vary in accordance with the type of fuel processed. As a safety margin of
at least 2% O2 is required, the maximal permissible limit of 02 concentration for mineral coal
dust is therefore as a rule 12%, for lignite 10%.

2.2.2 Design Related Explosion Protection
Reduction of the effects of already proceeding explosions, and therewith the protection of
people and machines, can be achieved by:
Explosion pressure resistant construction
Explosion pressure venting measures
Explosion suppression techniques (Fig. 4)
Fig. 4 Passive Protection Measures
2.2.2.1 Explosion Pressure Resistant Construction

Explosion pressure resistant construction restrict any possible explosion to the dust
conveying installation, whereby a certain amount of minor damage to the installation
commensurate with the complexity of the facility is accepted.
All dust conveying installation parts as well as the adjacent equipment and sealing elements
must be designed to resist the maximal explosion pressure of 9 bar expected in the case of
coal or lignite dust. If deformation of the container is accepted, the maximum permissible
explosion pressure may be up to 50% above its design value (pressure shock resistant
design). A design for 6 bar static overpressure is required for an expected maximum
explosion pressure of 9 bar. Such construction methods are of course quite complex and
expensive. However, in the event of an accident the installation is again operational within a
short time.
2.2.2.2 Explosion Pressure Venting Measures

In a broader sense explosion venting means all measures that serve to open temporarily or
permanently the previously closed installation in a safe direction, at the beginning or after a
certain spreading of an explosion. The purpose of this is to prevent any overstressing of the
mechanical equipment beyond its pressure shock resistance. The strength of the equipment
does not have to be designed for Pmax, but only for the reduced explosion pressure Pred (Fig.
5). A deformation of the container may again be acceptable, but it must not burst.

Fig. 5 Pressure Response in Explosion-Pressure-Relief Techniques
The explosion pressure venting technique operates in the following manner: When the
dynamic response pressure of the pressure venting installation is reached, predetermined
breaking points, rip foils or doors open to vent the shock wave outdoors, mainly by means of
amply dimensioned discharge channels. Immediately after the pressure venting system
responds an increase in the temporal rate of pressure rise can often be observed which is
due to the higher turbulence caused during the venting of the shock wave. The pressure rise
then quickly stops at Pred. Guidelines concerning the design layout and dimensioning of the
explosion pressure venting installations are contained in VDI Guidelines No. 3673. If the
method of explosion venting is applied not only the inserts of the containers such as filter
cloths etc. must be considered but the expected recoil forces as well. With a pressure
venting area of 1 m2, a reduced explosion pressure of 2 bar, and under the assumption that
the shock wave escapes with the velocity of sound, a thrust of approx. 15 t acts upon the
housing to be protected. This must be properly supported or else the container may be torn
from its foundations.
2.2.2.3 Explosion Suppression

In the explosion suppression techniques, the shock wave preceding the combustion front or
the infrared radiation of the combustion area is detected by a device which quickly
distributes extinguishing agents under a propellant pressure of 60 to 120 bar by means of
detonator-operated valves. With a programmed dynamic response pressure threshold (Pdyn)
of the detectors, the maximal explosion pressure is again lowered to a reduced level (Pred).

2.2.2.4 Limitation: Explosions from Ducts into Containers
The described constructive protection techniques are effective under the condition that the
reaction takes place as described in the paragraph 2.1. The description is applicable to most
explosions that occur in pulverizing plants. However, if an explosion strikes from a duct into
a container, and in doing so the residual dust deposited there is stirred up with great
turbulence and ignited, the reaction within the duct and the adjacent container can develop
into a detonation of such dimensions that the resulting pressures can amount to 50 times the
original pressure, accompanied by a combustion front traveling at supersonic speed, so that
any relief or suppression installation is too sluggish in action. However, such events are,
fortunately, relatively rare in coal operations. As a limit for a spontaneous explosion
propagation, an explosion characteristic of 100 bar.m.s-1 was observed under particular
conditions in a 200 m long pipeline of 1800 mm diameter located at the experimental mining
research station in Dortmund, while the usual values for coal are generally lower (approx. 85
bar.m.s-1).
However, if the principles of design related explosion protection are to be consequently
pursued, every duct conveying combustible dust in an explosive concentration and whose
length exceeds five times its diameter must be safeguarded by an explosion vent placed
ahead of its inlet into a container (such as a filter) (Fig. 6). Through this any explosion
originating in the pipeline will be vented so that the protective measures taken with respect
to the adjacent container can be designed in accordance with the criteria of an explosion
starting in the container itself.
Fig. 6 Venting of a Duct in Front of a Precipitator

2.2.3 Prevention of Smoldering Fires
Smoldering fires in dust deposits are best prevented by preventing the possibility of greater
quantities of dust accumulating. This is achieved through the appropriate design and slope
of surfaces, pipelines and supports, as well as sufficiently high gas speeds within the
conveyor systems.
In silos where great quantities of combustible dust are stored for the plants own specific
purposes, any combustion that may occur must be detected as early as possible by carefully
monitoring of the dust temperature and the CO content of the silo atmosphere so that proper
countermeasures can be taken.
3. APPLICATION OF PROTECTIVE MEASURES IN THE INDUSTRIAL

ENVIRONMENT
The fire and explosion protection measures described above result for practical applications
on the one hand in a network of preventive safety measures that significantly reduce the risk
of an accident in the operation of combustible dust installation, and on the other hand in
actual explosion protection techniques that can prevent explosions, or at least shall hold the
explosions within acceptable limits.
3.1 Preventive Safety Measures

The primary aim of preventive safety measures is to exclude possible ignition sources as
causes of conflagration or explosion if at all possible. In addition, they are also intended to
prevent secondary damage caused by the expulsion or stirring up of vast quantities of dust
and their subsequent ignition.
These essentially preventive safety measures can be listed as follows:
Temperature measurement of
mill exhausts
stored dust, preferably in silo entry and exit
CO analysis of
the silo atmosphere in silos
mill exhaust after the filter
Prevention of local overheating caused by friction in conveyor belt systems, high speeds
of screw conveyors, bucket elevators, rotary valves, and bearing, and/or the detection of
increasing temperatures by measuring techniques. Relative velocities of moving parts <
1 m/s are considered safe, > 10 m/s are considered as potential ignition sources.
Spark separators in air heaters
Metal separator prior to the mill
Prevention of electrostatic discharges by conductive connections and grounding of all
installation parts
Prevention of arcing in electrofilters by appropriate voltage control measures
Prevention of dust accumulation possibilities:
All surfaces to have a slope of at least 70 to the horizontal plane, especially in filter
or silo cones
Regular disposal of dust deposits
Gas speeds in conduits of more than 22 m/s

Protection of the stored dust from the effects of external heat, for instance by spraying
the silo externally with cooling water
Provision of inert gas supplies (e.g. CO2) for inertization of the silos in the case of
smoldering fires
Cleanliness of operating rooms
Effective removal of the dust generated by means of proper dedusting installations
Safe elimination of dust deposits by means of suitable auxiliary material
From the point of view of safety a solid fuel pulverizing plant must be operated as
continuously as possible, as critical situations often arise when the plant is not in operation.
This fact must be considered when the capacity of the installation is being decided upon.
3.2 Explosion Protection Techniques
3.2.1 Inert Gas Operation

As described under 2.2.1, active explosion protection in solid fuel pulverizing is practically
limited to inert gas operation, i.e. operation with a maximum of 10 to 12% oxygen in the
pulverizing plant, depending on type of fuel, as ignition sources and the stirring-up of dust
can never be excluded with absolute certainty.
Active explosion protection can be applied if hot inert gases such as the kiln exhaust from
cement kilns or hot gases from a combustion chamber, combined with a corresponding
design for the mills recirculation gas are available. In the last case the dew point problem
becomes significant, therefore this solution is rarely applicable for very moist fuels, or special
measures will have to be taken for drying of the circulation gases.
If the inert gas atmosphere can be maintained with absolute certainty through appropriate
design and interlocking of the installation for as long as combustible dust is present in the
system, design related protection measures become in principle redundant. In those cases
where these conditions cannot be guaranteed, for example, because hot gases with higher
oxygen content are being used such as clinker cooler exhausts, or because of dew point
problems, design related explosion protection techniques must be rigorously applied.
3.2.2 Explosion Pressure Resistant Construction

Explosion pressure resistant construction, i.e. the dimensioning of the installation section to
resist maximal explosion pressure, are mainly applicable where pressure venting methods
cannot be used at all or only with difficulty, for geometrical reasons.
This is mostly the case in mills, and definitely in all conduit pipe systems where the length of
the system exceeds five times the tube diameter. As a rule such components are designed
to withstand a static overpressure of 10 bar.
3.2.3 Explosion Pressure Venting Measures

All combustible dust conveying components that are not in themselves designed to be
explosion pressure resistant, such as cyclone, filters, pulverized fuel silos, etc. are to be
provided with properly dimensioned devices for explosion pressure venting. Thereby
containers and all interconnected aggregates such as bin vent filters, etc. must be
dimensioned in pressure shock resistant design to withstand the reduced explosion
pressure.
Explosion venting openings within a particular building must be connected to properly
dimensioned exhaust channels leading into the open. In order to prevent an explosion
originating in the mill spreading into the filter via the conduit pipe, the conduit pipe must be
equipped with an explosion vent in front of its connection to the filter. This measure is not

required for pneumatic conveying systems as in this design the dust concentration is
normally above the explosion limit. In addition, the minimal ignition energy is significantly
higher under the operating conditions of pneumatic conveying than it is in the case of
stirring-up combustible dust in containers.
The area containing the vent opening for explosion pressure venting must not be accessible
to anyone when the installation in operation.
VDI Guidelines No. 3673 can serve as a basis for the design of such an explosion pressure
venting system. Naturally, the system must be inspected regularly.
3.2.3.1 Underpressure Protection

After venting an explosion in very large enclosures such as pulverized fuel silos through
explosion flaps considerable underpressure can develop inside the silo due to dynamic
effects and due to cooling down of the hot gases remaining in the silo after the explosion.
Typical examples for the size of underpressure valves are given in Table 3. Guidelines for
the individual design of underpressure valves can be taken from the relevant literature (10).
Table 3
Volume m3 100 1000

Diameter m 3.4 .5
Cylindrical length m 9.5 22.0
Plate thickness mm 6 8
Max. negative pressure mbar 100 25
Required aspiration area m2 0.1 1.0
3.2.4 Explosion Suppression

Techniques of explosion suppression can basically replace all the previously mentioned
methods. However, in practical experience it has been seen that in pulverizing plants, the
costs involved in the consequent application of explosion suppression techniques are
significantly higher than they are for explosion pressure venting techniques and explosion
pressure resistant construction methods, both with respect to procurement and maintenance
of the sensitive equipment.
Thus applicability of explosion suppression may be primarily limited to existing, insufficiently
protected pulverizing plants whose retrofitting in accordance with alternative protection
techniques would be entirely uneconomical.
3.2.5 Fire Extinguishing Measures

If an accumulation of considerable quantities of combustible dust can be prevented inside
the actual pulverizing plant (except in pulverized fuel silos), any fires that may arise following
an explosion will not be able to grow to any significant size. The installation of a fire
extinguishing system can nevertheless still be recommended for cloth filters and electofilters.
In the case of smoldering fires in pulverized fuel silos, all further fuel supply must be stopped
immediately. Following this, the silo exit must be made airtight and the silo atmosphere
flooded with CO2. Sufficient time must now pass until the temperature conditions have
normalized. An underpressure valve is required in order to avoid collapsing of the silo due to
the vacuum produced during cooling down. The above procedures can take several days,
depending on the size of the smolder location. An alternative technique is to deliver the fuel
as quickly as possible to the burner system via the dosing and conveyor systems. Of course
this method is possible only when the dosing and conveying systems are heat-resistant,

dustproof and explosion resistant. In addition, under no circumstances is glowing fuel to be
returned to the silo, as for instance via overflow feeders.
Fig. 7 shows the practical preventive safety measures for pulverized fuel silos. In Fig. 8 the
application of design related protective measures for solid fuel preparation is illustrated.
Fig. 7 Preventive and Safety Measures for Coal Dust Silos

Fig. 8 Example to Show the Application of Design Related Explosion
Protection

4. LITERATURE
1) VDI Guidelines 2263
Verhtung von Staubbrnden und Staubexplosionen
2) VDI Guidelines 3673
Druckentlastung von Staubexplosionen
3) VDI Report No. 304
Sichere Handhabung brennbarer Stube
4) Arbeitskreis der chemischen Industrie, October 1, 1973
Sicherheitsmassnahmen gegen Staubbrnde und Staubexplosionen
Published by BASF, Bayer, Ciba-Geigy and Farbwerke Hoechst
5) Dr. W. Bartknecht
Explosions, Course Prevention Protection
Springer Verlag, Berlin, Heidelberg, New York, 1981
6) K.N. Palmer
Dust Explosions and Fires
London, Chapman and Hall, 1973
7) H. Wibbelhoff
Der Umgang mit Kohlenstaub in der Zementindustrie
Steine und Erden 2/1978
8) F. Schneider
Kohlenaufbereitung und Kohlenfeuerung fr Zementdrehfen
Zement, Kalk, Gips, No. 7/1976
9) E.W. Scholl, D. Reeh, W. Wiemann, M. Faber, G. Khnen, H. Beck, N. Glienke
Brenn- und Explosionskenngrssen von Stuben
STF-Report No. 2 - 79 (as well as BVS-Report) *
* see paragraph 4.1
10) W. Wiemann, R. Bauer, F. Mller
Unterdruck-Sicherung von Silos nach Staubexplosionen bei Anwendung von
Explosionsklappen
VDI Report No. 701, 1988, Volume II

4.1 Approximate Values for Explosion Limits and Ignition Temperatures
The numerical values of the following data are depending on the test procedure applied and
can vary within certain limits according to the origin and geological age of the coals. The
following values refer to the Literature (9).
Explosion Limits
1) Dust concentration:
lower explosion limits 40 to 130 g/m3
upper explosion limits 2000 to 6000 g/m3
2) Oxygen concentration:
hard coal 14%
lignite 12%
3) Concentration of non-combustible parts (ash):
hard coal (-medium volatile bituminous) 65%
Ignition Temperature
Cloud C Layer C
Lignite 380 to 450 225 to 300
Hard coal 590 to 710 245 to 380
Petrol coke 690 280

Chapter 5
Burners and Flame

Burners and Flames
Dirk Pauling
TPT 01/21063/E
Replaces report no. PT 98/14354/E
1. Terminology ..................................................................................................................289
2. Burners..........................................................................................................................290
2.1 Monochannel Burner........................................................................................290
2.1.1 Burner Design...............................................................................................290
2.1.2 Key Figures for Mono-channel Burner..........................................................291
2.2 Multi-channel Burner ........................................................................................292
2.2.1 Multi-channel Burner Design ........................................................................292
2.2.2 Pillard Rotaflam Burner ................................................................................293
2.2.3 FLS Duoflex Burner ......................................................................................295
2.2.4 KHD Pyrojet Burner ......................................................................................296
2.2.5 Unitherm M.A.S. Burner ...............................................................................297
2.2.6 Greco Burner ................................................................................................299
2.2.7 Key Figures for Multi-channel Burner Design and Operation .......................301
2.3 Fuel Oil Atomizers............................................................................................302
2.3.1 Mechanical Atomisers with Fixed Orifice and Variable Pressure .................302
2.3.2 Mechanical Atomizers with Variable Orifice and Constant Pressure............305
2.3.3 Nozzles with Assisted Atomization through Steam or Compressed Air .......306
2.4 Natural Gas Burners ........................................................................................307
2.4.1 Pillard Rotaflam KGD (gas/coal/oil burner)...................................................307
2.4.2 KHD Gas Burner...........................................................................................308
2.4.3 FLS Duoflex..................................................................................................309
3. Flames ...........................................................................................................................310
3.1 Effects of Poor Burning ....................................................................................310
3.2 The Ideal Flame.............................................................................................310
3.3 Flame Characteristics of the Different Burner Systems ...................................311
3.3.1 Mono Channel Burner ..................................................................................311
3.3.2 Multi Channel Burner....................................................................................312
3.4 Factors Influencing the Flame..........................................................................314
3.4.1 Primary Air Settings......................................................................................314
3.4.2 Position of the Burner in the Kiln ..................................................................314
3.4.3 Alignment of the Burner in the Kiln ...............................................................314

3.4.4 Secondary Air Temperature .........................................................................314
3.4.5 Excess Air.....................................................................................................314
3.4.6 Interaction Flame - Material Bed ..................................................................315
3.4.7 Pulverized Coal Characteristics....................................................................315
3.4.8 Fuel Oil Flame Adjustments .........................................................................315
3.4.9 Natural Gas Flame Adjustments...................................................................315
3.4.10 Burning of low reactive fuels.........................................................................316
3.4.11 Oxygen Enrichment ......................................................................................316
3.5 Combustion Indicators .....................................................................................317
3.5.1 Clinker Quality ..............................................................................................317
3.5.2 Sintering Zone Temperature.........................................................................318
3.5.3 Coating Formation ........................................................................................318
3.5.4 Exhaust Gas Analysis...................................................................................318
3.5.5 Kiln Inlet Temperature ..................................................................................319
3.5.6 Volatilisation of SO3 .....................................................................................319
3.6 NOx Formation.................................................................................................321
3.7 Flame Adjustment Procedure...........................................................................321
4. Combustion in Secondary Firing and Precalciner ....................................................322

Introduction
The function of the burner is to introduce the fuel into the burning zone.
The propagation of the combustion process depends on how fast the combustible comes
into contact with oxygen. It is therefore the essential function of the burner to regulate this
mixing process adequately in order to achieve a correct flame shape.
This process must take place in such a fashion that the heat is released at exactly the right
place without producing any damaging effects and without producing excessive pollutant
elements such as NOx, SOx and CO. Consequently, any optimization of the burning process
must start with the correct adjustment of the flame.
This paper describes how the flame can be adjusted, what burner types are available and
under what conditions they work best.
1. TERMINOLOGY
Primary air + secondary air + false air = combustion air

Stoichiometric combustion air + excess air = combustion air
Figure 1: Terminology of combustion air

2. BURNERS
2.1 Monochannel Burner
2.1.1 Burner Design

The monochannel burner is the most simple burner design. With this burner type, coal dust
and all the primary air is injected together through a single tube. Usually this type is used for
long kilns, equipped with direct firing.
Characteristics of the monochannel burner:
High amount of primary air
(In combination with direct firing systems for coal, up to 50% primary air)
Very limited flame shaping possibility
High NOx formation
Long and stable flame
Mono channel burners can also be used for fuel oil firing or for a combination of coal and oil
firing (additional channel for the oil nozzle in the centre).
Conical burner tips can be used to increase the injection velocity.
Figure 2: Mono-channel burner

2.1.2 Key Figures for Mono-channel Burner
Recommended range of specific axial impulsion (Gax) for mono channel burners:
M(transport air + fuel) Vtransport air N

Gax = = 67
Q fuel MW
M: Massflow Transport Air + Fuel (kg/s)

V: Injection Velocity (m/s)
Q: MJ kg
Fuel Input (calorific value fuel massflow) = [MW ]
kg s
This corresponds with the old rule of thumb, which states that the kinetic energy of the
primary air jet of a mono channel burner should be kept constant within certain limits:
(Velocity of Primary Air)2 * (% Primary Air) = 65'000 - 75'000
Even if this formula will not give optimal values in each case, it enables a rough estimate of
the dimension of the burner, if presupposed as a second condition:
Primary air injection velocity = 50 - 100 m/sec (valid for straight burners without swirl)

2.2 Multi-channel Burner
2.2.1 Multi-channel Burner Design

The most simple design of a burner is the mono channel burner. However, for optimum
flame shaping when considering changing coal quality and different requirements from the
point of view of raw mix burnability, burners with adjustable flame are to be preferred. In
such burners, the primary air is usually divided into an axial and a radial component with the
coal also introduced via a concentric ring tube.
These burners are called multi channel burners and are usually suitable for alternate or
combined firing of coal, oil or even gas. The axial air is injected in the direction of the kiln
axis (similar to a mono channel burner where all the air is injected in axial direction).
The radial air (or swirl air) is injected with a direction towards the kiln wall.
The swirl component of the radial air creates a rotating air flow along the kiln axis (similar to
the threat of a screw) which is also pushing towards the outside, in direction of the kiln wall.
Since the radial air channel is located inside the axial air channel (both are concentric ring
channels), the radial air is opening up / widening the flow of the axial air.
An increase of radial air versus axial air therefore creates a shorter flame. An increase of
axial air versus radial air create a longer flame.
Besides flame shaping, the primary air (especially the axial air) also has to cool the burner
pipe.
A typical example of the first generation of multi channel burners is the Pillard 3-Channel
Burner (Fig. 3). This burner has the coal channel in between the axial- and the radial air
channel. A problem recognised with these burner types is that a shortening of the flame
tends to produce a too wide flame (flame impeigements on the kiln wall).
Furthermore coarse cool particles (residue on 200 m sieve) can be thrown out of the
primary air jet by the radial air. These particles can cause reducing condition on the clinker
bed and high NOx formation.

Figure 3: Conventional Pillard 3-channel burner
A new generation of multi channel burners has therefore been developed. With special
arrangements and constructions of the primary air channels the above mentioned negative
effects can be avoided through the creation of a longer and more homogeneous internal
recirculation zone in the flame (chapter 3.3.2).
2.2.2 Pillard Rotaflam Burner

The Rotaflam burner is characterised by the location of the coal channel inside the axial and
radial air channels, as well as the flame holder / flame stabiliser (bluff-body) in the enlarged
centre cross section.
Features of the Rotaflam:
Primary air fan with 250 mbar
Separate adjustment of burner tip cross section for axial and radial air
Conclusion on the Rotaflam burner:
Very flexible burner (adjustment possibility of the tip cross sections)
Extensive and good experience in the Holcim Group
Required amount of primary air (12%) must be specified when purchasing (chapter 2.2.7)
The Rotaflam can be recommended. It is installed in the majority of the new projects.

Figure 4: Pillard Rotaflam burner

2.2.3 FLS Duoflex Burner
The Duoflex burner is characterised by the the joining of the axial and radial air inside the
burner right before the tip as well as the enlarged centre cross section (bluff-body / flame
stabiliser).
Features of the Duoflex:
Primary air fan with 250 mbar
Adjustment possibility of the burner tip cross section (only the sum of axial and radial air)
Conclusion on the Duoflex burner:
Flexible burner (adjustment possibility of the tip cross sections)
No experience so far in Holcim Group (will be installed in Ramos Arizpe 2)
FLS is a known and experienced supplier
Required amount of primary air (12%) must be specified when purchasing (chapter 2.2.7)
The Duoflex burner can be recommended for new projects
Figure 5: FLS Duoflex burner

2.2.4 KHD Pyrojet Burner
The Pyrojet burner is characterised by the jet air. The jet air is being introduced at nearly
sonic velocity.
Features of the Pyrojet:
Jet air with approx. 1000 mbar axial air pressure (blower needed)
No adjustment possibility of tip cross sections
Conclusion on the Pyrojet burner:
The concept does not foresee the recommended (chapter 2.2.7) amount of primary air
(big blower needed - high operating costs)
Mechanical problems of burner tip (tightness)
The Pyrojet is not recommended because of too little momentum, primary air rate and
the lack of adjustability especially for burning AFR.

Figure 6: KHD Pyrojet burner
2.2.5 Unitherm M.A.S. Burner

The Unitherm burner is characterised by only one primary air channel with adjustable swirl.
Features of the M.A.S. burner:
New concept with one single, adjustable primary air channel
The mechanical adjustment mechanism in the burner tip is protected by refractory steel
and cooled with all the primary air (according to the supplier non of the burners in
operation was damaged so far)
Conclusion on the M.A.S. burner:
No application so far in the Holcim Group

Special deal for the first burner in the Holcim Group is offered with the right of return after
one year
The required amount of primary air (12%) and the primary air fan pressure must be
specified when purchasing (the standard design values are much lower)
The M.A.S. can be an interesting alternative for an existing (mid-size) kiln
Figure 7: Unitherm M.A.S. burner

2.2.6 Greco Burner
The Greco burner is characterised by the high pressure of the primary air and the injection of
the coal between axial and radial air.
Features of the Greco burner:
The burner hardware is not supplied by GRECO. Only the drawings are given and
manufacturing has to be done in a local workshop or by plant.
No adjustment possibility of tip cross sections
Conclusion on the Greco burner:
The majority of the Greco burners in the Holcim Group are not in operation anymore
Concept of self-made burner normally does not work
Poor support outside South America
The Greco burner is not recommended at the moment because of too many negative
examples. Positive experience only in the MERCOSUR countries.

Figure 8: Greco burner

2.2.7 Key Figures for Multi-channel Burner Design and Operation
1. Primary Air Ratio: PA = 10 - 12%

(based on stoichiometric combustion; without transport air)
2. Specific axial momentum*: Gax = 7 10 N/MW
(including fuel and transport air)
To fulfil this criteria a minimum of 200 mbar primary air fan pressure is required
(better 250 mbar).
3. Swirl number*: S = 0.1 0.25 (max. 0.4)
4. Injection velocity of the coal at the burner tip: v = 25 30 m/s
Seen from a heat saving point of view the primary air ratio should be as low as possible in
order to recuperate as much hot secondary air as possible but on the other hand, the kinetic
energy of the primary air must be sufficiently high to provide a good mixture with the
secondary air to ensure rapid burning.
Experience with low primary air burners has shown, that primary air ratios of 6 - 8% are on
the technical limit below which it is no longer possible to guarantee stable combustion
conditions. With primary air ratios of 6 - 8%, any disturbance of the burning process tends to
shift combustion to the kiln back end, producing high kiln inlet temperatures and poor clinker
quality. Especially with the use of difficult fuels such as petcoke and AFR, enough primary
air and momentum is a prerequisite.
When purchasing a new burner, additional channels in the burner centre for the future use of
AFR should be foreseen because they can usually not be retrofitted into an existing burner.
* The formulas for the calculation of the burner momentum and swirl number are outlined in
the annex.

2.3 Fuel Oil Atomizers
Once properly prepared in terms of filtering, heating up and delived to the burner with
constant pressure and viscosity, the fuel oil must be atomised for effective mixing with the
combustion air. Therefore fuel oil atomising nozzles are used. These nozzles are located in
the centre of the burner, surrounded by the injection of the primary air. The oil nozzle is held
in place by a jacked tube which is a fixed part of the burner. Thus the atomising nozzle is
retractable.
The oil gun with the atomiser head should be taken out of the burner whenever it is not
needed to prevent overheating and coking of the remaining oil.
For fuel oil atomisation different principles are employed:
Mechanical atomisation with fixed orifice and variable pressure
Mechanical atomisation with variable orifice and constant pressure
Assisted atomisation with steam or compressed air
2.3.1 Mechanical Atomisers with Fixed Orifice and Variable Pressure

This type of atomisers are the most common for fuel oil.
Hereby the oil throughput is governed by the pressure of the fuel oil (within the range given
by the selected discharge opening/orifice plate).
With these atomisers the fuel oil flow in the atomiser head is usually subdivided into a
tangential flow (also called primary circuit) and an axial flow (also called secondary circuit).
By adjusting the pressure and thus the ratio of these components, it is possible to alter the
spray angle of the fuel jet. In general, an increase of the tangential oil pressure leads to
intensified swirling of the oil which has the tendency to shorten the flame. Typically the
differential pressure between the two flows is in the range of 0,5 - 2 bar (pressure of
tangential is higher). The overall pressure is usually approx. 40 bar. However flame shape
control is not only a result of atomiser adjustments, but also a function of primary air control.
Fig. 9 and 10 show two current atomisers (Pillard and Unitherm) with tangential and axial
flow and alternatively return-flow for start-up operation. For return-flow, the axial oil flow is
used to return a portion of the tangential oil flow back to the storage tank, in order to have a
high flow velocity and oil pressure in the nozzle head (swirl chamber) despite the small
amount of oil injected in the kiln (start up phase). Thus the turndown ratio can be increased,
still with a good atomisation.
The practical turndown ratio for this type of atomiser is 1:5 (without changing the orifice
plate).
As an additional feature, the length of the swirl chamber in the Unitherm atomiser is
adjustable.
For the operation of the atomiser it is of utmost importance that the pieces in the atomiser
head are assembled correctly and that the flexible hoses for primary and secondary oil
between the burner platform valve train and the atomiser gun are connected correctly.
Fig. 11 (Coen Tri-Tip Nozzle) shows a mechanical atomiser with fixed orifice without radial
and axial oil flow.

Figure 9: Pillard MYS atomiser
Figure 10: Unitherm atomiser

Figure 11: Coen Tri-Tip atomiser

2.3.2 Mechanical Atomizers with Variable Orifice and Constant Pressure
This type of atomiser employs the adjustable needle valve principle for throughput control.
By moving the needle back and forth, the open cross section of the orifice can be adjusted.
Needle value atomisers are mainly used by FLS. Atomising pressures is kept constant at 25
bar. The turndown ratio is limited to 1:5 1:6.
Figure 12: FLS atomiser type TFSM (needle valve principle)

1 = tangential slots; 2 = swirl chamber

2.3.3 Nozzles with Assisted Atomization through Steam or Compressed Air
This type of atomiser uses compressed air or steam instead of tangential oil to create an
intense swirl in front of the orifice plate. The advantage of these atomisers is the higher
turndown ratio because even a small amount of oil can be atomised effectively with
compressed air or steam. Another advantage is that fuel oil with some solid impurities can
also be atomised. The disadvantage of these atomisers is the need for a significant amount
of compressed air or steam (typical value: 0.1 0.12 kgAir/Steam/kgFuel).
Figure 13: Pillard atomiser with assisted atomisation

2.4 Natural Gas Burners
A gas burner does not necessarily need primary air because flame shaping can also be
accomplished by splitting up the gas flow into an axial and radial component (example KHD;
chapter 2.4.2).
However most burners use primary air for flame shaping. These burners are often multi fuel
burners which can also be used for coal or oil firing.
2.4.1 Pillard Rotaflam KGD (gas/coal/oil burner)

The Rotaflam multipurpose burner is designed for combined or separate firing of gas, coal
and oil. Flame shaping is achieved with separate axial and radial air.
Arrangement of the channels (from outside to the centre):
Axial air
Radial / swirl air
Gas channel
Pulverised coal channel
Central air / flame stabiliser
Jacket tube for oil atomiser nozzle in the centre
Figure 14: Pillard KGD gas / coal / oil burner

2.4.2 KHD Gas Burner
Primary air is not used for this burner. It requires a rather high supply pressure (3 - 5 bar)
because the hole burner momentum has to come from the gas flow.
The burner has been used in various kilns since a long time.
Arrangement of the channels (from outside to centre):
Axial gas channel
Radial gas in the centre
Figure 15: KHD gas burner

2.4.3 FLS Duoflex
The Duoflex gas burner is based on the normal Duoflex burner for coal and oil. Instead of
the central air plate, a gas nozzle is used.
Arrangement of the channels (from outside to centre):
Primary air (sum of axial and radial air)
Central air
Radial gas
Axial gas with needle valve control cone in the centre
Figure 16: Duoflex gas burner

3. FLAMES
3.1 Effects of Poor Burning

Effects of poor burning:
Unstable coating behaviour, particularly in the transition zone, reduces the lining life.
Incomplete burning and a local reducing atmosphere increase sulphur volatilisation and
build-up of coating in the preheater and in the kiln inlet area. Thus a significantly higher
dust cycle is created which shifts the entire temperature profile toward the kiln inlet.
With high CO-formation, secondary combustion forms at the back of the kiln which leads
to ring formation.
As a result, the kiln cannot operate at maximum output, the specific heat consumption
increases and the efficiency drops.
Low availability, rate and thus OEE
Therefore an optimisation of the burning process has to start with the correct adjustment of
the flame.
3.2 The Ideal Flame

The "ideal" flame can prevent, or at least keep within limits, the operating problems
described above.
The flame is stable over the entire burn-out distance:
By continually mixing hot secondary air into the burning zone. Therefore combustion can
take place in a controlled manner over the entire flame length.
No local temperature peaks are formed.
No local reducing conditions develop over the clinker bed.
Burn-out is complete at the end of the sinter zone.
In addition this "ideal flame" has to be achieved with the lowest possible formation of NOx.

3.3 Flame Characteristics of the Different Burner Systems
3.3.1 Mono Channel Burner

Mono channel burners tend to produce an axial flame without internal recirculation. The
heating up of the fuel jet to ignition temperature is predominated by external recirculation of
the hot combustion gases.
Figure 17: Flame shape mono channel burner
Effects:
Long sinter zone
Long retention time of the kiln charge in the hot zone and thus high volatilisation of
alkalis and sulphur (very suitable for the production of low-alkali clinker)
High NOx formation
With a constant primary air ratio, the length of the flame reaches a minimum for a given
primary air velocity. If the velocity is further increased, the primary air jet develops an
excessive suction effect which results in a reverse flow of flue gases. The recirculating flue
gas thins the secondary air so much that the flame becomes longer again.
For the recommended range of the specific axial impulsion see chapter 2.1.
However, for optimum flame shaping in response to changing production requirements,
burners with adjustable flame (multi channel burners) are to be preferred.

3.3.2 Multi Channel Burner
Multi channel burners can produce a divergent flame with internal and external recirculation
zones. The ability to change the relationship between axial and radial air provides an
important control mechanism for influencing the flame shape.
A hollow cone flame shape is produced, which can be modified by adjusting the pressure,
flow rate and burner tip cross section of the radial and axial air (Fig. 18).
The first generation of multi channel burners (e.g. Pillard 3-Channel) has some negative
effects on the flame shape, if there is a high content of radial air used. Two different flame
zones can appear:
In the first zone with internal recirculation there is intense combustion. Depending on the
arrangement of the swirling flow, in this zone coarse fuel particles are spun out of the
flame and then burn quickly in the oxygen-rich atmosphere of the hot secondary air.
In the second, long and instable zone, dominated by external recirculation, burn-out is
completed.
Effects:
Peak temperatures in the internal recirculation zone.
With very divergent flames, there are problems with the lining.
CO formation above the clinker bed caused by incomplete burn-out of the extended fuel
jet.
Unstable coating formation in a long transition zone, caused by an enlarged unsteady
burn-out zone.
Increased NOx formation because of the long retention time of the gasses at high
temperatures.
High sulphur volatilisation because of the reducing zone above the clinker bed and the
long retention time at relatively high temperatures.
The modern multi channel burners have been optimised so that these effects are largely
avoided. The special arrangement and construction of the primary air channels make the
internal recirculation zone (IRZ, Fig. 18) longer and more homogenous. This reduces the
length of the burn-out zone with external recirculation. For a faster mixing of the primary air
with the fuel, these burners have an enlarged flame stabiliser in form of a bluff body in the
centre.

Figure 18: Flame shape of modern multi channel burner
Effects:
Homogeneous temperature distribution, no excessive temperature peaks.
Low volatilisation rate of alkalis and sulphur.
Homogeneous recirculation zone, and therefore less NOx formation.

3.4 Factors Influencing the Flame
In most cases the most favourable operation is achieved with a rather short, hot and stable
flame, giving a high heat transfer rate to the material bed and a short and stable burning
zone.
The flame shape may be optimised during operation by adjusting the following parameters:
3.4.1 Primary Air Settings

A shortening of the flame can usually be achieved by:
Increasing the injection velocity (and thus momentum) of the primary air
Increasing the amount of radial air
Increasing the primary air rate
3.4.2 Position of the Burner in the Kiln

One of the most pronounced influence on flame length is the position of the burner tip:
Shifting the burner further into the kiln increases the flame length significantly and vice
versa.
This is because the turbulence field of the in-flowing secondary air significantly intensifies
the mixing of the secondary air with the fuel and primary air jet.
In planetary cooler kilns this effect is less noticeable as the position of the burner tip is
defined by the kiln's internal cooling zone.
Recommendations for burner tip position (except for planetary coolers):
SP kiln: Distance kiln outlet burner tip 1 m
(for petcoke 0 m or even outside rotary kiln)
Long wet and dry kiln: Distance kiln outlet burner tip approx. 1 m or a little more
Sometimes if the burner tip is too close to the kiln outlet, overheating of the nose ring and
snowman formation in the cooler can occur.
3.4.3 Alignment of the Burner in the Kiln

Basically the burner should be aligned parallel to the kiln axis. In the cold kiln the burner
should even be pointed slightly upwards, (especially for long burners in kilns with planetary
coolers), to compensate for the bending downwards in the hot kiln. If the burner is aligned
horizontally (the kiln axis has an angle of approx. 3 to the horizontal) as is often seen, the
flame tends to reach the material bed. A local reducing atmosphere is created resulting in
high sulphur volatilisation.
3.4.4 Secondary Air Temperature

The secondary air temperature defines, firstly, the ignition behaviour of the flame (black
plume) and, secondly, the possible flame temperature. Insufficient secondary air
temperature has to be compensated by fuel, and this means an increase in the combustion
gas quantity and a lengthening of the temperature profile. In point of fact, the clinker cooler
operation is one of the main factors influencing the flame.
3.4.5 Excess Air

Some excess air is required for complete combustion. Operating with a too low excess air
factor increases the combustion time and hence the flame length. This can create a reducing
atmosphere which increases sulphur volatility thus leading to clogging problems in the
preheater. If the excess air is significantly higher than the optimum value, the temperature

profile is extended again because of a too long flame. This results in an insufficient
temperature gradient towards the material bed and a longer sinter zone.
3.4.6 Interaction Flame - Material Bed

As the heat transfer from the flame to the material bed in the sinter zone is almost entirely
through radiation, the key factors affecting the heat transfer are the temperature and the
emissivity of the flame. If radiation to the clinker bed is reduced due to a dusty kiln
atmosphere, a long drawn-out temperature profile and thus a long sintering zone is
produced. In this situation, the dust is overheated in the flame and often deposited in the
transition zone or even further down the kiln and can cause ring formation. Ways to counter
this effect include all those measures which serve to improve clinker granulation (e.g.
adjustments in raw mix design).
3.4.7 Pulverized Coal Characteristics

Volatile content:
The combustion time of pulverised coal increases as the volatile content decreases,
therefore low volatile coal has a longer burning time and ignition distance than
bituminous coal.
Grinding fineness:
The burning time of a coal dust grain increases approx. with the square of its diameter.
The combustion time of a grain of coal increases as its volatile content decreases. Thus,
low volatile coal must be ground finer in order to burn within the desired time, e.g. in
order to produce the desired flame length.
Recommendations for optimum fineness: "Firing Systems - Handling and Preparation of
Noble Fuels"; HMC/TPT report no. TPT 01/21062/E
Ash content:
A high content of ballast material (ash) has a retarding effect on the burning time caused
by the reduced coal dust concentration and the lower flame temperatures as a result of
the heat absorption of the ballast material.
Rate of swelling:
The higher the expansion of the coal grain during heating in the flame, the shorter the
burning time. Nevertheless, the swelling index has no practical significance for the
cement kiln.
3.4.8 Fuel Oil Flame Adjustments

A faster burn out of the fuel oil can be achieved by:
Lowering the oil viscosity, thus increasing the oil temperature (recommendations for
optimum oil temperature: "Firing Systems - Handling and Preparation of Noble Fuels";
HMC/TPT report no. TPT 01/21062/E)
Atomiser adjustments (e.g. increasing primary oil pressure; chapter 2.3)
3.4.9 Natural Gas Flame Adjustments

The main requirement with natural gas burners is the possibility of producing a reverse flow
zone in the centre of the flame in order to achieve locally, inside the flame, a reducing
atmosphere where hydrocarbon molecules agglomerate to larger chains. This is necessary
to increase the emissivity of the gas flame, a prerequisite for heat transfer in the sintering
zone.

Adjusting the shape of the flame visually is almost impossible, because clearly defined flame
contours are hardly recognisable. Optimisation of the flame shape should be done following
the combustion indicators (see chapter 3.5).
3.4.10 Burning of low reactive fuels

To improve the combustion of low reactive fuels such as petcoke or anthracite, a small
amount of liquid fuel can be used to accelerate the ignition and thus burn out (called pilot
flame).
3.4.11 Oxygen Enrichment

By adding pure oxygen to the combustion air, the flame temperature can be increased. At
the same time the specific exhaust gas quantity is lowered. This decreases the energy
losses of the exhaust gas and allows to increase the production capacity of the kiln.
The disadvantages are the high costs of oxygen, higher NOx formation and lower cooler
efficiency.

3.5 Combustion Indicators
One of the main problems in the evaluation of the flame is that, in the rotary kiln, it is only
possible to observe the flame visually to a limited extent. On closer consideration, however,
there are a number of indicators which can provide much more information about the quality
of the flame than can be obtained from simple visual observation.
In the following, the most important operation indicators (combustion indicators) with direct
relation to firing parameters are discussed:
Clinker quality:
- Free lime
- Clinker microscopy
- Magotteaux-test
- SO3
Burning zone temperature:
- Pyrometer
- Amps
- NOx
- Clinker bed inclination
Coating formation in rotary kiln:
- Kiln shell temperature profile
- Length of fixed coating
- Stability of coating
- Ring formation
Exhaust gas composition:
- CO
- O2
Kiln inlet temperature
Volatilisation of sulphur:
- SO3 in hot meal
- SO3 volatility
- Encrustation at kiln inlet and preheater
3.5.1 Clinker Quality

The most important indicator for assessing the quality of the combustion is the clinker
quality:
Free lime:
Free lime is the most important quality parameter of the clinker to assess the burning
degree. In the past, especially with Lepol kilns and long wet kilns, the liter weight was
used frequently instead of free lime because of the simpler determination. For modern
suspension preheater kilns the liter weight is a poor indicator for the burning degree. To
assess the usefulness of liter weight measurement for kiln control, the liter weight must
have a correlation to the free lime. If there is no clear correlation, the liter weight can not
be used for kiln control.
Clinker microscopy:
The microscopy can reveal important information on the burning conditions such as
reducing conditions in the kiln.
Magotteaux test:
The Magotteaux is a simple and fast test to detect reducing conditions during the burning

process. Ground clinker is mixed with a special liquid which indicates reducing conditions
by changing its colour.
SO3 in clinker:
A decrease in the concentration of SO3 in clinker (at constant sulphur input) indicates
over-burning or reducing conditions.
3.5.2 Sintering Zone Temperature

Information about the sintering zone temperature can be obtained by:
Pyrometer:
Measuring the clinker bed temperature under the flame using a radiation pyrometer.
Amps:
Changes in the kiln drive power consumption (Amps or kW) are an indicator for the
temperature of the sintering zone and thus the burning degree. An increase of the kiln
drive power indicates a hotter sintering zone.
NOx:
The higher the NOx concentration in the kiln exhaust gas the hotter is the flame and thus
the sintering zone. In case of secondary firing or precalciner firing the NOx concentration
should be measured at kiln inlet because the concentration after preheater is strongly
influenced by the secondary firing or calciner firing.
Kiln charge inclination:
The kiln charge inclination in the kiln (angle of the material bed) is an indicator for the
temperature of the sintering zone and thus the burning degree. The hotter the clinker
bed, the more sticky is the material and the higher is the inclination.
It has to be noted, that all the above mentioned measurements do not supply absolute but
rather relative temperature indications and that the NOx-level is also highly depending on the
flame characteristics (see chapter 3.6).
3.5.3 Coating Formation

Kiln shell temperature profile and length of fixed coating:
The length of the sinter zone is an indicator for the length and temperature profile of the
flame. The coating formation can be determined indirectly, by measuring the temperature
profile of the kiln shell or directly be measuring the length of the fixed coating inside the
kiln during a stop. The influence of burner adjustments on coating formation can be
checked by recording the kiln shell temperature profile before and after any change to
burner settings.
Guideline for the length of the fixed coating: 5-6 x DKiln = good; 7-8 x DKiln = long flame
Stability of coating:
The stability of the coating especially in the transition zone is an indicator for the stability
of the flame. An instable flame length will lead to a continued falling and forming of
coating which is dangerous to the refractories.
Ring formation:
The formation of a ring in the kiln can be an indicator for poor combustion or a too long
flame.
3.5.4 Exhaust Gas Analysis

The gas analysis at kiln inlet supplies important information on the completeness of the
combustion.

Set point O2 kiln inlet:
Calciner kiln: 3-4%
Preheater kiln: 2% (if secondary firing is used, the O2 set point must be high enough to
supply the oxygen needed for the combustion of the secondary fuel)
CO:
Good burn-out: COKiln Inlet < 0.05% (500ppm)
Too high CO levels do not only cause energy losses but do also increase sulphur
volatilisation and thus build-up problems.
3.5.5 Kiln Inlet Temperature

In the case of cyclone preheater kilns, the kiln inlet temperature gives information on flame
length and retarded combustion. The target is to have the kiln inlet temperature as low as
possible.
Kiln inlet temperatures in excess of 1100C can be an indicator for a too long flame.
In short kilns (L/D < 12) higher temperatures can be normal.
3.5.6 Volatilisation of SO3

The flame has an important influence on the behaviour of SO3 in the kiln. This is governed
by factors such as retention time of the material in the hot zone (flame length), local reducing
conditions and excess air factor.
Encrustation at kiln inlet and preheater:
Encrustation and cyclone blockages can be an indicator for a combustion problem.
Assessment of the cause is done by hot meal analysis and determination of SO3
volatility.
SO3 hot meal concentration:
The enrichment of sulphur in the hot meal is an important indicator for the combustion.
To assess the limits, it is important to consider also the chlorine concentration in the hot
meal.
Figure 19: Combination of SO3 and Cl in hot meal
% Cl
2.0 Fre
que
n t bl
Zon ock
1.4 e of i a ges
n cre
asin
1.0 g enc
0.8 rus
ta tion
No e pro
n crus ble
ta tion ms
s
% SO3
0 1.0 2.0 3.0 4.0 5.0
2.5 3.75

SO3 volatility:
SO3 > 0.7 indicates a combustion problem ( = 1 cFeed/cHM; c = concentration)
Figure 20: Influence of temperature and O2 on sulphur volatility

3.6 NOx Formation
NOx formation is dominated as well by peak temperatures as by the amount of air entrained
into the primary fuel jet at ignition. Primary NOx reduction measures (main burner) are
deduced essentially from the following facts:
Flame front near the burner (short ignition distance)
Flame shaping with the aim to avoid high peak temperatures with at the same time
shorter flame
Lower burning temperatures (increase of free lime target, smooth kiln operation /
LINKman, no overburning and changes of the raw mix)
Low primary air rate
Note: A too low primary air rate has clearly negative consequences on kiln operation.
The minimum technically achievable NOx emission with measures related to the rotary kiln
burner are in the order of magnitude of 800 to 1'000 mg/Nm3. Further reduction of NOx
requires additional secondary measures (e.g. SNCR technology).
3.7 Flame Adjustment Procedure

1. Follow the operating instruction of the supplier for a medium flame setting
2. Wait until the kiln is stable before undertaking any adjustment
3. Progressively adjust one parameter at the time (e.g. axial- / radial air, oil pressure) and
monitor the corresponding combustion indicators. The kiln reacts slowly to any change. It
may take up to a few days to reach the new stable condition. It is therefore useless to try
to adjust a flame within one shift !
Caution:
The axial air channel also serves to cool the burner pipe. Always keep a minimum of 1%
primary air for axial air.
4. It is not recommended to operate the kiln with the shortest possible flame. A safety
margin for adjustment in both directions should be maintained for control of burning zone
disturbance.

4. COMBUSTION IN SECONDARY FIRING AND PRECALCINER
The burning conditions for secondary firing or in the precalciner are quite different from the
main firing.
In most cases the combustion takes place in an exhaust gas and air mixture instead of
pure air (except in-line calciner and precombustion chamber)
Combustion takes place in a very dusty atmosphere (raw meal)
The temperature range is only 1000C instead of 2000C
Due to these less favourable burning conditions, incomplete combustion can be the
consequence. Beside CO, incomplete combustion produces also carbon skeletons and CH4,
which both cannot be traced by CO measuring equipment. Further signs for incomplete
combustion in the preheater are:
Higher gas temperature at bottom cyclone outlet than at precalciner outlet
Only moderate temperature drop over the two lowest cyclone stages
Both indicate an after-burning within the preheater. This results in increased exhaust gas
temperature and heat consumption as well as coating problems.
Improvement measures:
Providing of enough gas retention time in the precalciner
Grinding of the coal to the required fineness
Avoiding fluctuations of the fuel feed

REFERENCES
1) "Firing Systems"
VA 82/4898/E
2) "Flames and Burners"
VA 93/4056/E
3) "State of Technology of Rotary Kiln Burners"
F. Schneider, PT 96/14078/E
4) W.L. van de Kamp / J.P. Smart
IFRF Research Report CEMFLAM1
"The effect of burner design and operation and fuel type of cement kiln flames"

ANNEX
Formulas and Definitions for the Calculation of Burner Momentum and Swirl Number
Primary Air / Combustion Air:
Kiln heat consumption Q [MJ/kgcli]

Min. combustion air Amin. 0.26 x q 3 Good approximation for
[Nm /kgcli]
conventional fuels
Total combustion air A n x Amin. 3 n = excess air factor, n>1
[Nm /kgcli]
Excess combustion air Amin. x (n-1) 3
[Nm /kgcli]
Primary air ratio Expressed in % [%A min.]
Amin.
Note:
In order to get lower primary air ratio figures, burner suppliers usually relate primary air ratio
to total combustion air.
Burner Geometry:
Following burner geometry calculations are based on the list of symbols and units stated
below:
ri Burner channel radius of channel I [m]

reg. i Equivalent channel radius of channel I [m]
Gx Axial momentum [N]
Gx.i Axial momentum of channel I [N]
Gax Specific axial momentum [N/MW]
Gt Tangential momentum [N]
Mi Mass flow through channel I [kg/s]
Qfuel Fuel heat input [MW]
Sb Burner swirl number [-]
vi.ax Axial velocity in channel I [m/s]
vsw.tan Tangential velocity on swirling channel [m/s]

Figure A: Typical Burner Geometry
Mono Channel Burner:

Total specific axial momentum through burner [N/MW]:
M v
tr

+ c tr N
Gax = MW
Q
fuel

Multi Channel Burner:
Total specific axial momentum through burner [N/MW]:
Gax =
(M sw v sw ,ax + M (tr +c ) v tr + M ax v ax ,ax ) N
Qfuel MW

Burner Swirl Number:
Swirl number =
[ ] [ ] []
Tangential Momentum N Characteristical Swirl Radius m
(Axial Momentum[N ] Characteristical Channel Radius[m ])
Gt [N ] r eq .sw [m ]
Sb = []
G xi [N ] req.i [m]
A common method for the calculation of the characteristical or equivalent radius is to
determine the radius of gyration for each individual channel cross-section as follows (Mathur
and Maccallun - 1967):
=
(
2 r2 r1
3
[m]
3
)
req .i
(
2
3 r2 r1
2
)
For a typical multi channel coal burner with axial-, transport- and swirl air, the burner swirl
number can be calculated according to the following formula:
Msw vsw,tan req.sw
Sb = []
M ax vax req.ax + M (tr +c ) vtr req.tr + M sw vsw,ax req.sw

Chapter 6
Combustion Engineering

Combustion, Gasflows and Gas Composition
A. Obrist
VA 89/5584/E
1. INTRODUCTION ............................................................................................................330
2. COMBUSTION ...............................................................................................................330
2.1 Basic Relations and Definitions........................................................................330
2.2 Calorific Value ..................................................................................................333
2.3 Combustion Calculations .................................................................................337
3. GAS FROM RAW MATERIAL .......................................................................................342
3.1 Gases from Dry Raw Meal ...............................................................................342
3.2 Water from Wet Raw Meal or from Raw Slurry ................................................342
4. KILN GAS ......................................................................................................................343
4.1 Measurement of Exhaust Gas Composition.....................................................343
4.2 Immediate Calculations from Gas Composition ...............................................353
4.3 Calculation of Exhaust Gas Quantities.............................................................357
4.4 Calculation of the CO2 Content of Cement Kiln Exhaust Gas ..........................361
5. FALSE AIR INVESTIGATIONS .....................................................................................363
5.1 Introduction ......................................................................................................363
5.2 Evaluation ........................................................................................................364
5.3 Example of an Investigation .............................................................................365

SUMMARY
In context with gas measurements on a cement kiln there are important numerical relations
which must be understood.
Such numerical relations involve the following subjects:
Fuel properties, calorific value
Raw meal properties
Combustion calculations
Gas composition
Gas properties
Gas quantities
False air
Calculations can be done with complete, exact formulas but sometimes also by using
approximations. Approximations are never 100% precise but often sufficient for practical
purposes. Important, basic approximations given in this chapter are e.g.:
Min. combustion air = Amin 0.26 Nm3/MJ
Min. combustion gas = Vmin 0.28 Nm3/MJ

The exhaust gas of a cement kiln consists of:
Gas from raw meal
Gas from combustion
Excess air and false air
Water from slurry or raw meal
By considering the above contributions the exhaust gas quantity of a kiln can be calculated.
This method of calculation and many other relations are given in this chapter.

1. INTRODUCTION
Once, in the Greenfield cement factory, they wanted to do a few measurements on their kiln
system, which seemed to run at its capacity limit. The specialist who was in charge of these
measurements wanted to know the actual gas flows in the kiln system. Unfortunately, a few
years ago, the designers of the Greenfield kiln had to fight with some difficult space
problems when they had to plan the layout of the gas ducts in the narrow space. They never
had considered that somebody would have to take flow measurements in this ductwork. The
shape of the gas ducts was far away from the ideal, long straight tube for a precise flow
measurement.
After a hard job within the hot areas of the gas ducts the specialist realized that his results
from the pitot tube measurement was still not as precise as he had imagined. He therefore
started to think about his problem. Maybe, there was another method to come to a result?
Obviously gas flow has something to do with the fuel combustion and also raw meal
produces some gas. So, why not calculate the gas from other parameters such as fuel
quantity and raw meal? And wouldnt it be possible to calculate also the gas composition?
But where do I find the necessary relations and formulas?
With this kind of thoughts in mind the specialist in the Greenfield plant was about to use the
paper on hand, and finally to do a more interesting and effective job.
2. COMBUSTION
2.1 Basic Relations and Definitions
2.1.1 Combustion
Combustion of fuels is a chemical reaction of fuel with oxygen (O2) according to the general
scheme:
FUEL + OXYGEN COMBUSTION PRODUCTS

(CO2, H2O, SO2, Ash)
Combustion reactions usually go through intermediate steps where some intermediate

products may occur. However after complete combustion (total oxidation) the resulting
combustion products are of extremely simple nature because complete combustion always
ends at only 3(!) simple gas molecules, namely CO2, H2O and SO2.
The overall combustion reactions can therefore be characterized by three very simple
combustion equations:
C + O2 CO2
2H + O2 H2O
S + O2 SO2

2.1.2 Air
Oxygen for combustion of fuels is taken from the air. For the considerations within this
chapter the composition of the dry air can be simplified (neglecting trace gases) as follows:
OXYGEN (O2) 21.0 Vol%

NITROGEN (N2) 79.0 Vol%
AIR 100.0 Vol%
According to the local climate air contains some water vapor, e.g. 2 Vol% at 20C and 80%
relative humidity.
2.1.3 Normal Conditions

Normal conditions for gases are defined as:
Pressure = 1.0133 bar = 760 Torr
Temperature = 0C = 273.16K
By the above conditions, the Nm3 (Normal-cubicmeter) is defined which is used as unit for
gas quantities.
Note:
1) The normal conditions refer to the average atmospheric pressure at sea level (Fig. 1 at
altitude = 0 m) and at the zero point of the Celsius temperature scale (0C).
2) There is a similar definition of standard conditions (USA) which refers to the 60F point
of the Fahrenheit temperature scale (30 inch Hg = 1.016 bar, 60F = 15.6C).
Unfortunately these conditions are not equal to the normal conditions in the metric
system.
Fig. 1 Average Barometric Pressure in Function of Altitude

2.1.4 Kmol and Ideal Gas
To characterize the quantity of substances in context with chemical reactions the unit kmol is
used.
One kmol means a certain number of molecules (Avogadros constant = 6.02 x 1026/kmol).
If the molecular weight of a compound is taken in kg it equals to 1 kmol of this substance.
For example: The molecular weight of CO2 is 44. Therefore 44 kg of CO2 are equal to 1 kmol
(which contains 6.02 x 1026 molecules) of CO2.
One of the basic relations of chemistry and thermodynamics says that a certain number of 1
kmol of any gas molecules takes always the same volume (at constant pressure and
temperature). The formula of the gas molecules does not have any influence on their volume
in gaseous state.
To say this more precisely it must be added that this applies only for ideal gases. Gases at
low partial pressures and at temperatures above the critical point can be considered as ideal
gases. All gases occurring in context with this chapter can be considered as ideal gases with
sufficient accuracy (approx. 0.1%).
This fact can be used by the following quantitative relationship:
1 kmol of an ideal gas takes a volume of 22.4 m3 at normal conditions (1.0133 bar, 0C)
Or in short form:
1 kmol of gas = 22.4 Nm3
Note that even H2O and CO2 behave nearly like ideal gases as long as they occur in gas
mixtures at low partial pressures.
2.1.5 Conversion of the Volume of Ideal Gases

Ideal gases or mixtures of ideal gases behave according to the fundamental equation:
pv = RT
(p = absolute pressure, V = volume, R = gas-constant, T = absolute temperature)
Therefore, volume conversions from condition 1 to condition 2 can be done easily by means
of the ratios of absolute temperature [K] and absolute pressure as follows:
P T
V2 =V1 1 2
P2 T1
Example: Convert V1 = 1 m3 at 350C (623.16 K) and 0.9 bar to normal conditions:
0.9 bar 273.16 K
V2 =1m 3 =0.389Nm
3
1.0133 bar 623.16 K

This type of calculation is frequently used for practical gas flow calculations. However, for
the considerations within this paper it will not be required in the following.

2.1.6 Minimum Air and Air Factor
To perform a complete combustion a theoretical minimum amount of oxygen O2 min) is
required, depending on the type of fuel. The corresponding minimum quantity of air is called
Amin:
O min
Amin = 2 [Nm 3 ]
0.21
Practically a combustion requires always slightly more air than the theoretical minimum Amin
in order to avoid local lack of O2 and unburnt products. The air factor n is defined as the
following ratio:
A
n= [ ]
Amin
A is the effective air quantity, n must be always higher than 1 in order to maintain complete
combustion.
2.2 Calorific Value
2.2.1 Gross and Net Calorific Value (CV)

The calorific value of a fuel sample is measured in a bomb calorimeter. The fuel sample and
a surplus of oxygen are ignited in the bomb and after combustion the amount of heat is
measured. The bomb is thereby cooled to room temperature level.
By cooling the calorimeter the water vapor is condensed and therefore the heat of water
condensation (2450 kJ/kg H2O at 20C) is included in the resulting heat.
The direct result (amount of heat) obtained from the calorimeter is therefore always the so
called gross calorific value CVgross (in German: oberer Heizwert, Brennwert, Ho) of the fuel.
The gross value, however, is not very significant for common technical applications because
the effect of water condensation does usually not occur. Therefore the net calorific value (in
German: unterer Heizwert, Hu) value is calculated by subtracting the heat of condensation,
as follows:
CVnet = CVgross - (water in combustion products) x 2450 [kJ/kg fuel]
The water in combustion products is calculated as follows:
H2O = H2Ofuel + 9 x Hfuel [kg/kg fuel]

hydrogen in fuel [kg/kg fuel]
Important Note: Within this chapter and also within the cement course chapter on Heat
Balances the net calorific value is used as reference. Unless otherwise noted, fuel energy or
fuel heat always refers to net calorific value.

2.2.2 Calculation of CV
2.2.2.1 Gaseous Fuels

If the gas components are known by analysis the calorific value can be calculated exactly by
adding the contributions of the pure gas components.
The gas analysis is given as volume % (which is the same as mole %) and the calorific value
is usually referred to one Nm3 (not kg of gas).
Gas Component Formula CV net

[MJ/Nm3]
Methane CH4 35.8
Ethane C2H6 63.8
Propane C3H8 91.3
Butane (gas) C4H10 118.8
Pentane (gas) C5H12 146.1
Ethylene C2H4 59.1
Propylene C3H6 86.1
Acetylene C2H2 56.1
Carbon monoxide CO 12.6
Hydrogen H2 10.8
Hydrogen sulfide H2S 23.2
Example:
Natural gas: CH4 = 90.5%
C2H6 = 2.0%
C3H8 = 0.5%
CO2 = 0.4%
N2 = 6.6%
100.0%
CV = 0.905 x 35.8 + 0.02 x 63.8 + 0.005 x 91.3 = 34.1 MJ/Nm3

2.2.2.2 Liquid and Solid Fuels
Usual liquid and solid fuels (fossil fuel oil and coal) consist of complex organic compounds.
Usually the ultimate analysis or the elemental analysis (content of C, H, S, N, O) is available.
Strictly speaking it is impossible to calculate the exact calorific value of a fuel only from its
elemental composition without knowing the kind of its organic compounds. However, as long
as fossil fuel oils and coals are considered, the following approximation produces fairly
accurate results:
CVnet 34.8 x C + 93.9 x H + 10.5 x S 6.2 x N - 10.8 x 0 - 2.5 x W [MJ/kg]
C, H, S, N, O, W are the weight fractions [kg/kg fuel] of carbon, hydrogen, sulfur, nitrogen,
oxygen and water.
The above approximation produces also good results for other organic materials such as
wood, paper and peat. But is should not be used for extreme cases such as e.g. pure carbon
or pure sulfur.
Note that for exact determination of the CV of fossil fuel oils and coals only the calorimeter
method - and not the above approximate calculation - can give the correct result.

2.2.3 Examples of Calorific Values
(Including also alternative fuels)
CV [MJ/kg] (net)
Pure polyethylene 46
Light oil 42
Heavy oil 40
Pure polystyrene 40
Pure rubber (without inert material) 36
Anthracite 34 High grade
Waste oils, various refinery wastes 30 to 40
Petcoke 33
Waste tires (with steel and inert material) 28 to 32
Bituminous coal (low ash) 29
Bituminous (high ash) 24
Acid sludge, acid tar (from oil refining) 16 to 22
Lignite (10% moisture) 16 to 21
Pot liners (from aluminium smelter) 20
PVC 19
Palm nut shells (10% moisture) 19 Medium
grade
Pressed olive cake 18
Dried peat (10% moisture) 18
Fuller earth (from oil refining) 13 to 18
Dried wood, bark, saw dust (10% moisture) 16
Rice husks (10% moisture) 16
Shredder wastes 15
RDF (from domestic refuse, 10% moisture) 15
Cardboard, paper (air dry) 15 Low grade
Dried sewage sludge (10% moisture) 10
Domestic refuse (30% moisture) 8.5
Pure iron (heat of oxidation!, occurs e.g. in waste tires) 7.5

2.3 Combustion Calculations
2.3.1 Exact Calculations

The following calculations are based on the combustion equations (see 2.1.1) and some
basic relations.
To demonstrate the calculation method, the simple example of combustion of pure carbon is
given as example.
2.3.1.1 Combustion of Pure Carbon (C)

The combustion of 1 kg of pure carbon (C) without any excess air (n = 1) is considered.
The combustion equation and the corresponding weights and volumes are as follows:
C + 02 C02
= 12 kg/kmol = 32 kg/kmol = 44 kg/kmol
= 22.4 Nm3/kmol = 22.4 Nm3/kmol
The above equation refers to 1 kmol or 12 kg of C. The final results are wanted per 1 kg of C
i.e. per 1 kg of fuel.
The minimum oxygen is:
O2min = 22.4 Nm3/12 kg = 1.87 Nm3/kg C
Air contains 21% O2, therefore:
Amin = 1.87 Nm3/kg / 0.21 = 8.89 Nm3/kg C
The combustion products are only CO2 and N2 (coming from Amin):
CO2 = 22.4 Nm3/12 kg = 1.87 Nm3/kg C
N2 = Amin x 0.79 = 7.02 Nm3/kg C
Total combustion gas = Vmin = 8.89 Nm3/kg C
Calorific value of pure C = CV (from handbook) = 32.8 MJ/kg C
If Amin and Vmin are referred to the CV the following results are obtained:
Amin/CV = 0.271 Nm3/MJ
Vmin/CV = 0.271 Nm3/MJ
From this simple example it would appear that Amin = Vmin, but in general the Vmin will be a
few percent higher than Amin. Nevertheless the example of pure C gives already a fairly
representative impression of a typical combustion calculation.
2.3.1.2 General Case

In general a fuel is given either by its elemental composition or by its volume composition in
case of gases. Therefore two types of formulas are required (2.3.1.2.1 and 2.3.1.2.2) which
are based either on weight composition or on volume composition of the fuel.
Amin = Minimum air
Vmin = Minimum combustion air

2.3.1.2.1 Calculation based on Elemental Analysis / Weight Composition
C, H, S, N, O, H2O are the weight fractions (kg/kg fuel) of carbon, hydrogen, sulfur, nitrogen,
oxygen and water in the fuel.
Amin = 8.89 x C + 26.5 x H + 3.3 x S - 3.3 x 0 [Nm3/kg fuel]
Vmin = 0.79 x Amin + 0.8 x N + 1.87 x C

+ 0.7 x S + 11.2 x H + 1.24 x H2O
+ Amin x H2O air [Nm3/kg fuel]
The terms in the Vmin-formula mean:

0.79 x Amin + 0.8 x N N2
1.87 x C CO2 dry gas
0.7 x S SO2
11.2 x H + 1.24 x H2O wet
H2O+
+ Amin x H20 air
Example:
Coal: C = 71.0% Moisture of air = O
H = 4.0%
O = 6.0%
N = 1.5%
S = 1.0%
H2O = 0.5%
Ash = 16.0%
100.0%
Amin = 8.89 x 0.71 + 26.5 x 0.04 + 3.3 x 0.01- 3.3 x 0.06 = 7.207 Nm3/kg coal
Vmin = 0.79 x Amin + 0.8 x 0.015 + 1.87 x 0.71+ 0.7 x 0.01

+ 11.2 x 0.04 + 1.24 X 0.005 = 7.494 Nm3/kg coal

2.3.1.2.2 Calculation based on Volume (or Mole) Composition
(only for gaseous fuels)
The general formula of any compound (also inert gases) in the fuel gas is defined as:
Cc Hh Ss Oo Nn
Examples: CH4 c=1 h=4

C2H6 c=2 h=6
CO c=1 o=1
H2S h=2 s=1
N2 n=2
A fuel gas is a mixture of various compounds, each having a certain volume fraction of vf
(Nm3/Nm3). The combustion calculations need a summation of all compounds in the mixture,
therefore the (sigma) sign appears in the equations.
1 h o [Nm 3 / Nm 3 fuel ]
Amin =
0.21
vf c + + s
4 2
n h [Nm 3 / Nm 3 fuel ]
Vmin =0.79 Amin + vf + c + s + + Amin H 2Oair
2 2
The above two formulas apply for any type of gaseous compounds (combustible or even
inert) and their mixtures.
The terms in the Vmin-formula mean:
0.79 x Amin + vf x (n/2) N2
vf x (c) CO2 dry gas
vf x (s) SO2
vf x (h/2) + Amin x H2Oair H2O } wet
Example: 90% CH4 vf = 0.9 c=1 h=4
4% C2H6 vf = 0.04 c=2 h =6
3% CO vf = 0.03 c=1 o=1
2% N2 vf = 0.02 n=2
1% O2 vf = 0.01 o=2
1 4 6 1 2
Amin = 0.9 1 + + 0.04 2 + + 0.03 1 + 0.01 =9.262Nm 3 / Nm 3
0.21 4 4 2 2
4 6 2
Vmin =0.79 Amin + 0.9 1 + + 0.04 2 + + 0.03 (1) + 0.02 =10.267Nm 3 / Nm 3
2 2 2

Composition of Vmin
The composition of Vmin (N2, CO2, SO2, H2O) can be easily calculated by identification of the
terms of the formula of Vmin.
The best way to show this is on the previous example of coal where a Vmin of 7.494 Nm3/kg
coal was obtained.
Comp. Calculation Term Nm3/kg coal % in Vmin

N2 0.79 x 7.207 + 0.8 x 0.015 5.706 76.1
(Amin)
CO2 1.87 x 0.71 1.327 17.7
SO2 0.7 x 0.01 0.007 0.1
H2O 11.2 x 0.04 + 1.24 x 0.005 0.454 6.1
Total 7.494 100%
Note: Due to the absorption of SO2 with cement raw meal the above calculation of SO2 will
not produce the true amount of SO2 in the gas of a cement kiln.
Fig. 2 shows typical compositions of Vmin for 6 common types of fuels.
Fig. 2 Typical Compositions of Vmin (for 6 common fuels)

2.3.2 Approximate Calculations
In the previous paragraph 2.3.1 the exact calculation procedures for Amin and Vmin are shown.
In many practical cases it is, however, not necessary to do exact calculations if
approximations of sufficient accuracy can be given.
Such approximates have been found by investigating a large number of different types of
fuels (see HMC report VA 81/4849/D).
For all fuels of practical use in the cement industry such as coal, fuel oil, natural gas and
most of the alternative fuels the following approximations for Amin and Vmin can be used:
Minimum air: Amin 0.26 x CV

Minimum combustion gas (wet): Vmin 0.28 x CV
where Amin resp. Vmin = [Nm3] and CV = [MJ]

Or expressed in words:
Each MJ (Megajoule) of combustion energy requires 0.26 Nm3 minimum air and produces
0.28 Nm3 minimum combustion gas.
By using the above approximations, a very quick determination of gas quantities is possible.
Note that no analysis of the fuel is required.
Furthermore, the above approximations can also be used if the CV of the fuel is not given
expressively. For example, if the specific heat consumption of a cement kiln is 3.5 MJ/kg
clinker the specific Amin and Vmin per kg clinker can be calculated directly:
Amin = 3.5 x 0.26 = 0.91 Nm3/kg cli
Vmin = 3.5 x 0.28 = 0.98 Nm3/kg cli
If the type of fuel is known the actual factors can be taken from the following table in order to
obtain somewhat more accurate results.
Type of Fuel Amin-Factor Vmin-Factor

(Nm3/MJ) (Nm3/MJ)
Anthracite, coke at 1% moisture 0.26 0.27
Bituminous coal at 1...2% moisture 0.26 0.28
(low to medium volatile)
Lignite (high volatile) at 10% moisture 0.26 0.29
Wood / Peat at 10...15% moisture 0.26 0.30
Light oil 0.26 0.29
Heavy oil (bunker oil) 0.26 0.28
Tar 0.26 0.28
Waste tires (rubber) 0.26 0.27
Natural gas (high CH4 content) 0.27 0.30
Natural gas (approx. 24% C2H6) 0.26 0.30

3. GAS FROM RAW MATERIAL
3.1 Gases from Dry Raw Meal

If cement raw meal is heated up, hydrate water and CO2 are released.
The true quantity of dry raw meal (not kiln feed) required to produce 1 kg of clinker is called
R. Normally R equals to approx. 1.55 kg/kg cli.
R must be calculated from the loss on ignition:
1 [kg/kg cli]
R=
1 L.O.I

loss on ignition of raw
meal in [kg/kg meal]
The above formula applies if the kiln dust is completely returned into the kiln. If some dust is
discarded (without return to kiln) the factor R increases accordingly.
The quantity of hydrate water released from dry raw meal is calculated as follows:
H2Ohyd = R x hydrate content x 1.24 [Nm3/kg cli]

(kg hydrate water/kg raw meal)
The quantity of CO2 released from dry raw meal is calculated as follows:
C02 = [R x (1 - hydrate content) - 1] x 0.509 [Nm3/kg cli]

(kg hydrate water /kg raw meal)
Normal dry cement raw meals release approximately the following gas quantities per one kg
of clinker:
H2Ohyd 0.01 Nm3/kg cli

CO2 0.27 Nm3/kg cli
Total 0.28 Nm3/kg cli
In most cases it is sufficient to use the above approximation. If dust is discarded the above
value must be increased accordingly.
3.2 Water from Wet Raw Meal or from Raw Slurry

If w (kg/kg) means the moisture or slurry water content of the kiln feed, the amount of water
vapor is calculated as:
w [Nm3/kg cli]
H 2O= R 1.24
1 w
In case of a dry kiln the above quantity is usually negligible, but it is essential on a wet kiln.
At a slurry water content of e.g. 35% it amounts to 1.03 Nm3/kg clinker.

4. KILN GAS
4.1 Measurement of Exhaust Gas Composition

Within this section O2, CO2, CO and N2 are considered. Trace gases such as NOx and SO2
and hydrocarbons belong to the scope of emission measurements and are usually in the
order of less than 0.1% and are therefore not significant with regard to the main components
(O2, CO2, CO, N2).
4.1.1 Gas Sampling

Although gas sampling may appear as one of the easiest things, it is in fact the source of
more than 50% of all measuring errors! Sampling can be done either continuously or by
extracting a gas sample into a rubber bladder.
The aim of sampling is to obtain a representative gas sample from a gas duct into the
analyzer. The possible problems and solutions in general are as follows:
Problem Solution
Gas is not homogeneous Sample the average of the total cross section (instead of
within the cross section of one sampling point).
the gas duct, because
gases do not mix well Avoid sampling points where poor upstream mixing seems
(sample not obvious.
representative)
Prefer sampling points after fans (mixing effect of fan
blades).
False air is entering into Check tightness of sampling system, especially when
gas sampling system(too sample is extracted from high underpressure.
much O2 in gas sample)
Avoid small sampling tubing which may become clogged by
dust (and produce high suction pressure at low sample gas
flow).
In case of spot sampling with rubber bladder: rinse bladder

at least once with sample gas.
O2 may react with steel Use hot extraction tubes made of sicromal, ceramics or
tube at > 400C (too little quartz.
O2 in sample)
CO2 may be absorbed by Use a filter directly at the hot extraction point.
dust and condensate (too
little CO2 in sample) Keep sampling system free of dust and condensate(e.g.
rubber bladder must be clean inside).
To a certain extent reliable sampling for CO2 will always be

a problem as long as gas cooling is applied.
Long storage time in Analyze rubber bladder after 30 min at the latest.
rubber bladder (too little
CO2 and too much O2 due
to diffusion)

4.1.2 The Orsat Apparatus
An Orsat apparatus is shown in Fig. 3. A gas sample is filled into a burette volume of 100%.
Step by step, CO2, O2 and CO are absorbed by contacting the gas sample several times with
the absorbing agent. The absorbed gas volume is measured after each step in the burette.
Fig. 3 The Orsat Apparatus

It is important to maintain the following, correct sequence of absorption:
1) CO2 KOH solution
2) O2 pyrogallic acid / KOH
3) CO Cu2 Cl2 solution
In order to avoid measuring errors the following hints are useful:
1) Check tightness of Orsat apparatus first.
2) Use fresh absorption liquids.
3) While taking a reading always hold liquid level in burette and expansion tank at same
height (constant pressure).
4) Check O2 absorption by measuring ambient air (21% O2).
Advantages of the Orsat:
Measuring principle is very clear and simple
The Orsat is available (or at least known) in every cement factory
Disadvantages of the Orsat:
Due to is principle a continuous measurement is impossible
Analyzing is time consuming
Although the Orsat can measure CO it is impossible to detect small traces of CO
0.1%. Therefore the CO result is only useful in cases of bad combustion producing
about 0.3 to 1% CO.
4.1.3 Other Gas Analyzers

In most cases continuous gas analyzers which are permanently installed are used for
process control in a cement kiln (see chapter Exhaust Gas Analysis). Such permanent
analyzers measure only at one location and are not flexible enough for an investigation of
the complete kiln system (involving a large number of measuring points).
More and more portable gas analyzers are available on the market. They can usually
measure O2, CO and combustibles continuously. At fairly low costs (less than $4000) a
quite handy and useful equipment can be purchased. For an investigation on a cement kiln
this is virtually the ideal tool to perform extensive and quick measurements at various
locations in the plant.
In the following a few important measuring principles are compared.

Frequent Measuring Principles for O2
Paramagnetic effect of O2
Either thermomagnetical or magneto-mechanical principle is used.
Thermomagnetical principle is cross-sensitive to gases having other heat
conductivities than O2 e.g. CO2.
Magneto-mechanical principle with cell containing movable dumb-bell is not cross-
sensitive to CO2 (see Fig. 4A).
Analyzers based on the paramagnetic effect are the most common type of
permanent installation in the cement industry.

Fig. 4A Frequent Measuring Principles for Gases -
Paramagnetic O2 Sensor

Electrochemical voltage effect of hot zirconium oxide (Fig. 4B)
Working temperatures above 600C.
Measuring the voltage of an electrochemical cell, according to the equation:
RT PO2 [V]
E= ln
nF PO2Re f
(R = Gas constant, T = [K], n = 4, F = Faraday constant, PO2 = partial pressure of oxygen)

Logarithmic output signal, depending on PO2.
Principle is seldom used.

Fig. 4B Frequent Measuring Principles of Gases -
Zirconiumoxide Cell

Self consuming electromechanical sensors (Fig. 4C)
Basic design of such a measuring cell is similar to a normal battery. Cell consists of
anode, electrolyte and air cathode (see Fig. 4C).
O2 enters via diffusion barrier to cathode.
At the air-cathode O2 is reduced to OH. Thereby an electrical current is produced
which is proportional to the O2 content and is used as output signal.
Above reaction causes an irreversible consumption of the anode (lead). Therefore
lifetime of the cell is limited.
Practical lifetime of such a cell is about 1 year (even if cell is not in use a certain
aging occurs).
Reliable measurements without major cross-sensitivities.
Calibration only with air (zero is self adjusting).
Application: Portable low cost analyzers.
Similar applications: Other electrochemical sensors have been developed for CO,
SO2 and NOx working on similar principles. By special design and compensation the
effect of self-consumption could be avoided. These analyzers are widely used for
portable applications but not for permanent emission control.

Fig. 4C Frequent Measuring Principles for Gases -
Electrochemical Cell
4.1.4 Reference to Dry or Wet

The most frequent type of gas sampling includes a cooling of the gas to approx. 0C (or at
least to room temperature). By this cooling the water vapor is condensed and the
measurement in the gas analyzer refers to dry gas.
The reference to dry gas will be considered as normal case in this paper (unless otherwise
noted).
The classic analyzer - the Orsat apparatus - refers always to dry gas. For this particular
analyzer even some residual water content in the gas sample will not affect the final result,
the reading means exactly the dry composition (in this case it would even be wrong to make
any correction for residual water vapor in the gas sample!).
Note that there are exceptional cases e.g. where gas samples are not condensed but
introduced directly in a hot zirconium oxide cell. In such a case the result will refer to wet
gas. Furthermore sampling systems without gas coolers but with dilution of the hot gas
instead (thus avoiding any water condensation) exist, but they are hardly used for O2, CO,
CO2.

Summarizing, sampling with gas cooling and reference to dry gas is considered as normal,
i.e. is used for about 99% of all practical measurements.
4.1.5 Determination of Water Content

Since all measuring systems described above are not able to measure water, other methods
have to be used if water really has to be measured. Important methods are:
Two temperatures method:
Measuring of wet bulb and dry bulb temperature.
If the gas temperature is < 150C wet and dry bulb temperature can be measured
directly in the gas duct. If the gas temperature is > 150C an extraction of the gas
and cooling down to < 150C is required.
Evaluation of gas moisture according to calculation sheet 32599e (see measuring
techniques part of cement course).
Method is preferred due to its simplicity, however accuracy at high moisture contents
may not be always sufficient.
Condensation method:
A gas sample is extracted for a certain time and cooled by ice water, thus water is
condensed.
The remaining dry gas extracted is measured by a volume counter.
By measuring the weight of the condensate the moisture content of the gas can be
calculated.

4.2 Immediate Calculations from Gas Composition
4.2.1 Dew Point

From the water content the dew point can be calculated.
Vf [Nm3/Nm3w] is the volume fraction of water vapor in the wet gas
Ptot [bar] is the total pressure (usually assumed as 1.0133 bar)
By expressing the vapor pressure in form of a numerical equation the following dew point
formula was developed for precise calculations:
336.48 [C]
= 179
5.3362 17.045 + ln(Vf Ptot )
Example:
Vf = 0.10 Nm3/Nm3w Ptot = 1.0133 bar
336.48
= 179
5.3362 17.045 + ln(Vf Ptot )
=46.09 o C
The corresponding inverse function giving the saturation pressure of water (PW) is as
follows:
3591 113216
PW =EXP 11.4297 + [bar]
t + 179 (t + 179 )2
t = 0...120C, max. deviation 0.03%.

A section of the above function is shown in Fig. 5 which can be used for graphical evaluation
of the dew point.

Fig. 5 Determination of Dew Point

4.2.2 Air Factor n
The air factor n can be calculated precisely from the gas composition by considering the N2,
according to:
N2eff N2 eff
n= =
N2 min N2 eff N2 excess
By using the O2 and N2 content of the gas the following important formula can be derived:
N2 1
n= =
0.79 O
N2 O2 1 3.76 2
0.21 N2
If CO is present the following, general formula applies:
1
n=
O 0.5CO
1 3.76 2
N2
Note that e.g. the CO2 from raw meal will not affect the n although it does affect the gas
composition (the ratio O2/N2 is constant when CO2 is added to the gas).
Therefore this formula is very useful and applicable for any case. On the other hand O2, CO
and N2 are required as inputs which needs a complete Orsat analysis (N2 is the rest
composition after absorption of O2, CO and CO2).
Example:
O2 = 4% N2 = 68% (CO = O)
1
n= =1.28
4
1 3.76
68
4.2.3 Combustion Efficiency
Unburned gases are a sign of incomplete combustion. The most important unburned gas is
CO, whereas hydrocarbons are usually < 0.1% on a cement kiln and can be neglected with
regard to energy losses.
If only CO is known the loss due to incomplete combustion is:
Q= (dry gas quantity) x CO x 12.6 [MJ]
3 3 3
[Nm ] [Nm /Nm ]
Example:
On a cement kiln the fuel combustion is 3.3 MJ/kg clinker, the dry exhaust gas quantity is 1.3
Nm3/kg clinker and the CO content = 0.2%.
q = 1.3 x 0.002 x 12.6 = 0.033 MJ/kg cli
0.033MJ
Relative loss = =0.01=1%
3.3MJ

Some gas analyzers have also a combustible sensor based on a catalytic combustion of the
sample gas. Such a sensor responses to all types of combustible gases (CO, H2, CH4, C2H6
etc.). Although the main combustible component is not CH4 (methane) but rather CO it is
common to express the result as CH4 equivalent.
If the total of combustibles is given as their CH4 equivalent the loss is:
Q= (dry gas quantity) x CH4 x 35.8 [MJ]

[Nm3] [Nm3/Nm3]
4.2.4 Gas Density

The gas density at normal conditions can be calculated by a simple mix calculation by using
the following densities:
Gas M Density*
[kg/kmol] [kg/Nm3]
O2 32.0 1.429
CO2 44.0 1.964
CO 28.0 1.250
N2** 28.15 1.257
H2O 18.0 0.804
* Density in a gas mixture (considered) as ideal gas

** Not as pure N2 but as rest in the air (including also Ar and trace gases)
Note: For the calculation of the true gas density all concentrations (O2, CO2, CO, N2,
H2O) are to be referred to wet gas here.
Example:
Exhaust gas density of a suspension preheater kiln with coal firing:
Gas Composition Composition
component of dry gas of wet gas
[Vol%] [Vol%]
O2 4% 0.9 x 4 = 3.6%
CO2 31% 0.9 x 31 = 27.9% 100 - 10 = 90%
CO 0%
N2 65% 0.9 x 65 = 58.5%
H2O -- 10 %
Total 100% 100 %
Density = 0.036 x 1.429 + 0.279 x 1.964

+ 0.585 x 1.257 + 0.10 x 0.804 = 1.415 kg/Nm3
(at normal conditions)

4.3 Calculation of Exhaust Gas Quantities
4.3.1 Firing / Hot Gas Generator

The following calculation applies for any type of firing where gases are coming only from fuel
combustion and excess air (no gases from raw meal).
Problem: The fuel consumption Q = [MJ/h] of the firing is known and the complete
exhaust gas analysis (CO2, O2, CO, N2) is given. By means of this information the gas flow
[Nm3/h] shall be calculated.
Solution: Calculate first the n according to 4.2.2:
1
n=
O 0.5CO
1 3.76 2
N2
The gas flow V [Nm3/h] is then:
V= Q [Vmin + (n - 1) Amin] [Nm3/h]
Vmin and Amin are in [Nm3/MJ] and can either be calculated exactly according to 2.3.1 or just
simply by introducing the approximations according to 2.3.2, namely:
Vmin 0.28 Nm3/MJ
Amin 0.26 Nm3/MJ
If the complete gas analysis is not known but only the O2 concentration (dry) a further
approximation for the dry amount of Vmin, namely 0.25 Nm3/MJ, has to be introduced which
results in the following approximation formula:
%O2 [Nm3/h]
V =Q 0.28 + 0.25
21 %O2
[MJ/h] O2 ref. to dry
Note: If the O2 was based on wet gas the above formula would become even more simple,
namely:
21
V Q 0.28
21 %O2
4.3.2 Cement Kiln

On a cement kiln it is common practice to work with specific quantities referred to 1 kg of
clinker, therefore the fuel consumption is expressed in q [MJ/kg cli] and V means here
[Nm3/kg cli].
The general calculation formula is:
V= q (Vmin + (n -1) x Amin) + (gas from RM) [Nm3/kg cli]

[MJ/kg cli] [Nm3/kg cli]

n is calculated from the complete gas composition according to 4.2.2.
If only O2 concentrations are available the following approximations can be used for dry
kilns:
- from combustion 0.28 * q [Nm3/kg cli]

- from raw meal 0.28 [Nm3/kg cli]
- excess air plus false air %O2 [Nm3/kg cli]
(0.28 + 0.25 q )
21 %O2
%O2 [Nm3/kg cli]

Total = V = 0.28 (q+1) + (0.28+0.25xq) x
21 %O2

[MJ/kg cli]
In case of wet raw meal or for wet kilns additional water has to be added (see 3.2) to the
above result.

The above approximation can be used in form of diagram (Fig. 6)
Fig. 6 Quick Determination of Kiln Exhaust Gas Quantity

Example:
Dry kiln, q = 3.3 MJ/kg cli, O2 = 3.5%
3 .5
Total gas = 0.28 (3.3+1) + (0.28+0.25x3.3) x = 1.425 Nm3/kg cli
21 3.5
Three typical results for cement kilns are shown in Fig. 7.
Fig. 7 Typical Exhaust Gas Quantities for Three Cement Kiln Systems

4.4 Calculation of the CO2 Content of Cement Kiln Exhaust Gas
4.4.1 Introduction
If no measured CO2 concentration is available it is possible to complete the gas analysis with
a calculated CO2 concentration. This can be useful e.g. for calculation of density or specific
heat.
It is possible to calculate the CO2 content at any point in the exhaust gas system
(suspension preheater, cooling tower, raw mill, filter) of a cement kiln if the corresponding O2
concentration is known.
The calculation is based on two main facts:
1) The CO2 balance applies (CO2 comes from combustion and from raw meal)
2) Starting from the O2 concentration the corresponding dilution with air can be quantified.
In addition a few other preconditions are to be observed here:
Reference is made to dry gas composition.
The assumption is made that all CO2 from raw meal is in the kiln gas (i.e. after complete
calcination).
No gas is lost or extracted from the gas stream under consideration (e.g.: an upstream
extraction of a bypass gas is excluded here).
For numerical calculations the following approximations are used:
CO2 from raw meal 0.27 Nm3/kg cli
N2 from combustion at n = 1
N2 = 0.79 x Amin + N2 from fuel
= 0.79 x 0.26 + 0.0015 0.207 Nm3/MJ
(this approximation is very accurate for all types of fuels)
4.4.2 Maximum CO2 Content

In a first step the influence of dilution (or excess air) is assumed as zero (0% O2). Therefore
a theoretical, maximum CO2 content can be calculated from the CO2 balance.
The dry gas contains only CO2 and N2 which can be calculated as follows:
CO2 = 0.27 + cf x q [Nm3/kg cli]

N2 = 0.207 x q [Nm3/kg cli]
q [MJ/kg cli] = specific heat consumption

cf [Nm3/MJ] = CO2 from fuel, referred to CV
cf depends on the type of fuel and can be calculated according to paragraph 2.3.1.2 (by
considering only the CO2 term in the Vmin formula).

Typical values for cf are:
Coal = 0.050 Nm3 CO2/MJ

Fuel oil = 0.039 Nm3 CO2/MJ
Natural gas = 0.028 Nm3 CO2/MJ
The concentration CO2max is calculated from the above quantities of CO2 and N2:
CO2 [Nm3/Nm3]
CO2 max =
CO2 + N2
4.4.3 True CO2 Content

Due to dilution of the gas with air and due to possible formation of some CO the true CO2
content is:
O CO 0.79 [Nm3/Nm3]
CO2 =CO2 max 1 2 + CO
0.21 2 0.21
CO2, O2, CO = [Nm3/Nm3]

Fig. 8 shows a graphical mode of evaluation of a gas composition based on the formulas
given in this paragraph.
Fig. 8 Relation Heat Consumption/Exhaust Gas Composition for Cement Kilns

(CO2 = 0.27 Nm3/kg cli)

Notes:
1) Although it is theoretically possible to calculate also the specific heat consumption from a
given gas composition (backward calculation) this procedure is not recommended. In
most of the practical cases this will be quite inaccurate because of the lack of a precise
gas composition (CO2 readings are often too low due to systematic sampling error).
2) It is not recommended to introduce calculated CO2 values in the n-formula according to
4.2.2, because calculation with calculated values may finally end in a somewhat
doubtful result. The air factor n should therefore only be calculated from a measured gas
composition (or alternative calculations based on O2 only should be used).
5. FALSE AIR INVESTIGATIONS
5.1 Introduction
Underpressures at various locations in the gas system may vary from 0 to -100 mbar.
Leakages of the equipment can therefore cause considerable amounts of false air which
increase the gas flow.
In addition to the obvious increase of gas flow, false air can also be detrimental for the heat
consumption.
If the exhaust gas fan (or the mill fan) is running at full capacity any additional false air
causes a reduction of the kiln production capacity. Shortage of production may become an
essential financial loss for the cement factory. Therefore measures against false air must be
taken.
The first step is always to locate the major leakage points. For this purpose it is necessary to
measure a complete O2 profile of the kiln system. This may e.g. include:
suspension preheater
cooling tower
kiln fan
raw mill
E.P.
mill fan
As already mentioned in 4.1.3 such a profile can be measured easily with a portable O2
analyzer. A large number of sampling points and repeated checks can be realized.
Proper sampling is essential for such an investigation! During such a measurement the
operating conditions of the system should be constant in order to obtain a consistent O2
profile. Proceed quickly from one sampling point to the next and finally re-check all O2
concentrations if they are really stable.

5.2 Evaluation
The measured O2 profile may already give a qualitative impression of the tightness of the
gas system. For a quantitative information in form of real gas flows (Nm3/h) the
approximation formula from 4.3.2 is used here (because it is based only on O2).
If m denotes the clinker production in [kg/h] the complete formula is:
%O2 [Nm3/h]
V = m [0.28 x (q + 1) + (0.25 X q + 0.28) ]
21 %O2
[kg/h] [MJ/kg cli]
V is the gas flow at the measuring point. The amount of false air is easily calculated as the
increase of V between two measuring points.
Note:
V does not include water evaporation e.g. from raw slurry or from a cooling tower. It is also
obvious from the formula that V will not be influenced by any water evaporation because O2
is based on dry gas composition.
For the purpose of false air calculation it is not required to add these additional water
quantities. It would be even wrong to add such quantities to V before calculating the false air
as difference between two V-valves.

5.3 Example of an Investigation
On a dry suspension preheater kiln with cooling tower and roller mill in combined operation
the O2 profile was checked by means of 5 sampling points.
The clinker production is 100 t/h with a specific consumption of 3400 kJ/kg cli.
Fig. 9 shows the evaluation of the results by calculating the V [Nm3/h] for every sampling
point.
Fig. 9 False Air Investigation on a Kiln System
V equals to the true wet exhaust gas quantity at the sampling points 1 and 2. For the points
3, 4 and 5 the amount of water vapor from cooling tower and the amount of water
evaporated in the raw mill is not included in V.
The false air flows are simply calculated as the increase of V.
The final result gives the amounts of false air inleak (Nm3/h) for each section. These values
are needed in order to evaluate the possible potential for optimization.
Conclusion:
For the example according to Fig. 9 the exhaust gas quantity after preheater is 145800
Nm3/h which is finally mixed with up to 101700 Nm3/h or 70% of false air when it reaches
the chimney. Therefore clear optimization potentials exist in the area of the cooling tower,
the raw mill and the electrofilter.

Chapter 7
Alternative Fuels
1. Use of Alternative Fuels and Raw Materials ........................................................... 367
2. Low Grade Fuel - Study ............................................................................................. 415
3. PREPARATION AND HANDLING OF ALTERNATIVE FUELS ................................. 483

Use of Alternative Fuels and Raw Materials
Christian Engmann/ Albert Obrist

TPT 03/21206/E (replaces PT 00/14674/E)
1. Introduction...................................................................................................................369
2. Types of AFR ................................................................................................................371
2.1 Definition (from HGRS/CIE) .............................................................................371
2.2 AF Classification ..............................................................................................371
2.3 AR Classification ..............................................................................................371
2.4 Properties of AF ...............................................................................................372
2.5 Properties of AR...............................................................................................373
2.6 AFR Codification List........................................................................................374
3. Utilisation ......................................................................................................................375
3.1 Graphs of TSR/ TEE and AR usage in Holcim plants worldwide .....................375
3.2 Feed points for AF............................................................................................378
3.3 Selection of feedpoint.......................................................................................379
3.4 Feedpoints for AR ............................................................................................381
3.5 Supply and Inlet Control...................................................................................381
3.6 Quality Control .................................................................................................382
3.7 Preparation of AFR ..........................................................................................385
4. Impacts ..........................................................................................................................386
4.1 Circulation Phenomena....................................................................................386
4.2 Potential loss of production capacity................................................................387
4.3 Compatibility with raw mix (quality) ..................................................................390
4.4 Emission...........................................................................................................390
5. Technical AFR Assessment ........................................................................................394
5.1 Basic Data Collection (self assessment)..........................................................395
5.2 Data Evaluation................................................................................................396
5.3 Optimal Fuel Mix & alternative raw material usage..........................................396
5.4 Example of a plant ...........................................................................................397
6. Application ....................................................................................................................402
6.1 Tires .................................................................................................................403
6.2 Whole Tires ......................................................................................................404
6.3 Shredded Tires.................................................................................................407

6.4 Ground Tires at the main burner ......................................................................408
6.5 Pyrolisis/ Gasifier .............................................................................................408
6.6 Domestic Refuse (usually not recommended) .................................................409
6.7 Solid AF < 5 mm: Imp. Sawdust in EC .............................................................411
6.8 Liquids (Ras el Ma) ..........................................................................................411
7. References ....................................................................................................................412
ANNEX 1 AFR Codification list

Summary
The main target of using AFR is cost saving. To succeed we need a profound understanding
of essential key elements. This paper is focused on the technical aspects, how to use
alternative fuels and raw materials in kilns. It is essential to make use of available sources of
experience and be in control of the burning process in the kiln.
1. INTRODUCTION
AFR reduces Fuel Cost

The main advantage of alternative fuels are their low purchase prices or the
possibility to attain revenue for burning these materials, however, there are some
drawbacks which have to be kept in mind. Not every waste is suitable for a cement
kiln and there are influences on the kiln process.
Holcim Goal
The EXCO of Holcim decided the following goal for the whole group:
Zero costs for thermal energy by the year 2006
Several plants have achieved this, however not yet sustainable on a long term
basis.
100% AFR substitution is possible

The Eclpens (Holcim Schweiz) plant ran in the early part of 2003 one month
without any noble fuel. The reason for going back to use noble fuels was the
insufficient supply of certain AF. It was a constant learning process until they
reached this goal. One of the biggest stumbling blocks was to be able to control
the kiln process and secondly to attain sufficient amounts of AF at the right price.
During this period they were burning the following fuel mix:
8%1 Tires at kiln inlet
7% Solvents
55% Animal Fat
16% Animal Meal
5% Plastics
8% Impregnated Sawdust
1% Waste Water
Definitions
Two parameters are used for evaluation AFR performance. They are:
Thermal Substitution Rate (TSR)
1
: % are percent of the total heat consumption

The Thermal Substitution Rate describes the percentage of the total heat
consumption of a kiln system which is substituted by alternative fuels.
Thermal Economic Equivalent (TEE)

The Thermal Economic Equivalent describes the total economic benefit
from using alternative fuels (AF) by the relationship between the actual
fuel costs and the theoretical fuel costs if no AF were used. It is
calculated based on preparation costs and material costs as well as on
thermal energy consumption.
At zero fuel costs the TEE would be 100%
Other sources of information

There are other resources beside this paper focusing on different aspects
concerning AFR:
Low Grade Fuel Study [LGF-2001]

Study on the impacts of alternative fuels on the kiln process.
AFR Quality Control Manual [QCM-2000]

How to select Waste suppliers and different AFR specifications.
Laboratory design and equipment needed. Sampling equipment and
procedures as well as Health& Safety Aspects. Complete with training
requirements.
Preparation and Handling of Alternative Fuels [PHAF-2003]

The process of preparing waste into a alternative fuel to be used in a
cement kiln is the subject of this paper. It covers both description of so
called AFR Platforms (AFR preparation facilities) and handling systems
for AFR in the cement plant.
Holspace:
There are Communities of Practice as well as other applications taking
care of this subject:
CoP AFR Application on Kiln [CopAoK]
The technical side of AFR
CoP AFR [CoPAFR]
Overall aspects (mostly non technical)
- Pyramid [PI]
An interactive tool for AFR Business development
CoP Petcoke[CoPPet]
Petcoke is not considered as AFR and therefore not treated here, but it
can be a great opportunity to save fuel costs.

2. TYPES OF AFR
2.1 Definition (from HGRS/CIE)

Traditional fuels are defined by international bodies, traditional raw material are the
pure materials gained from quarries or bought from primary material supplying
industries. Traditional fuels and raw materials are therefore better defined and
consequently the Holcim definition for AFR is the following: All fuels and raw
materials, which are not traditional, are alternative.
Traditional fuels:
Liquids: Fuel oils (heavy & light), other petroleum products
Solids: Coal, Petcoke, lignite, etc.
Gaseous fuels: natural gas
2.2 AF Classification
Fuel Family Examples

Liquid AF high grade > 21 e.g.: solvents, waste oil
MJ/kg
Liquid AF medium grade e.g.: emulsions
Liquid AF low grade < 7 MJ/kg e.g.: waste water
Whole tires
Lump > 50 mm e.g.: tire chips, toner cartridges or filter cakes
Solids < 50 mm e.g.: residue from car shredding, tire chips
Solids < 5 mm e.g.: saw dust or coarse animal meal
Solids < 1.5 mm e.g.: saw dust or dried sewage sludge
Solids < 0.5 mm e.g.: fly ash or ground AF
Table 1 Fuel Families
2.3 AR Classification
Current raw material list (ATR 02) and their classification
Raw material classification AR/N2 Examples

Limestone N CaCO3 > 75%
Marl N CaCO3 = 35% - 75%
Clay N CaCO3 < 35%
Blast furnace slag AR e.g. Httensand, Laine de laitier
Other slag AR e.g.: Copper slag, Iron slag
Fly ash AR e.g.: fly ash from thermal, power station
2
: AR = Alternative Raw material, N = natural raw material

Bottom ash AR e.g.: bottom ash from power plant
Calcareous corrective AR/N e.g.: Granule calcique, Espuma de
Azucarera, carbide sludge
Siliceous corrective AR/N e.g.: Sables de fonderie, Sand
Aluminous corrective AR/N e.g.: Waste Catalysts, Oxyde alumine
Ferrous corrective AR/N e.g.: Pyrite ash, Mill Scale
Other raw mix correctives AR/N e.g.: contaminated soil, drilling cut base
Gypsum AR/N
Table 2 Raw material Classification (ATR 2002)
2.4 Properties of AF
Handling and application of alternative fuels depends on physical properties. E.g. it
may be very simple to use waste oil which has been purified by the supplier. On
the other hand it is impossible to use e.g. raw domestic waste directly as solid fuel,
because of its poor quality and very inhomogeneous state. The only practical way
to burn it in a cement kiln is a sophisticated pre-treatment to produce RDF (refuse
derived fuel), which is very cost intensive and therefore seldom used.
Table 3 shows a list of fuels sorted by their calorific value. Traditional (or noble)
fuels are marked as such. The calorific value (CV) alone does not directly indicate
the suitability as cost effective AFR. E.g. waste tires are as good as coal from the
viewpoint of CV but require expensive handling and tend to cause negative
impacts on the kiln process, so an adequate compensation must be included in the
price (disposal fee).

Table 3 Examples of Alternative Fuels (shaded = noble fuel)
Material CV net [MJ/kg]
3
Pure polyethylene 46
Light oil 42
Heavy oil 40
Tar (by-product) 38
Animal Fat 37
Pure rubber (without inert material) 36
Anthracite 34
4
Aluminium metal 31
Waste oils, various refinery wastes 30 to 40
Petcoke 33
Waste tires 28 to 32
Bituminous coal (low ash) 29
Bituminous coal (high ash) 24
5
Liquid mix (CSL from SCORIBEL or SYNFUEL of Safety Kleen) 20 to 30
Landfill gas 16 to 20 (MJ/Nm3)
Acid sludge, acid tar (from oil reprocessing) 16 to 22
Lignite (10% moisture) 16 to 21
Pot liners (from aluminum smelter) 20
6
PVC 19
Palm nut shells (10% moisture) 19
Pressed olive cake 18
Dried wood, bark, saw dust (10% moisture) 16
Rice husks (10% moisture) 16
Car shredder wastes 15
RDF (from domestic refuse, 10% moisture) 15
Animal meal 15
Cardboard, paper (air dry) 15
Impregnated saw dust (25% moisture) 10 to 137
Dried sewage sludge (10% moisture) 10
Fullers earth (from oil purification, LD actual) 10
Domestic refuse (30% moisture) 8.5
Dried sewage sludge (30% moisture) 7.5
Pure iron 8 7.5
2.5 Properties of AR
Alternative Raw materials are distinguished into classes according to their moduli.
For each class two examples are given.
3
: Usually not in pure form, but contained in mixed plastics
4
: Al metal may occur e.g. in composite packaging wastes and is oxidized to Al2O3
5
: The properties of such mixes have to meet acceptable specifications
6
: Usually not in pure form, but contained in mixed plastics
7
Such mixes can be prepared according to acceptable specifications
8
Fe metal occurs e.g. in waste tires and is oxidized to Fe2O3, contributes slightly to CV

To distinguish whether a material is classified as a certain corrective is not always
clear. Certain materials would be applicable in two categories e.g. Silica and
Alumina corrective.
Potentials of AR usage depend on the chemical compatibility with the raw mix.
CLASS DESIGNATION SIO2 AL2O3 FE2O3 CAO H2O SM9 AM10 LSF11
Blast Httensand 35.59 14.79 0.26 34.5 10 2.36 56.88 29.42
furnace slag Laine de laitier 35.56 11.86 3.3 34.23 33 2.35 3.59 29.58
Copper slag 32.97 4.99 55.49 1.44 1.28 0.55 0.09 1.07
Other slags
Iron slag 26.47 4.29 63.22 1.38 11.58 0.39 0.07 1.15
Fly ash 48.66 24.06 4.11 1.52 0.5 1.73 5.85 0.91
Fly ash
Fly ash 46.18 31.64 6.19 4.25 0.75 1.22 5.11 2.49
Bottom ash 40.81 22.13 8.9 3.92 16.47 1.32 2.49 2.68
Bottom ash
Bottom ash 46.94 29.55 10.05 8.17 21.75 1.19 2.94 4.73
Granule calcique 4.95 0.92 0.59 52.07 3.28 1.56 339.7
Calcium
correctives Espuma de 1.35 0.39 0.16 47.34 29.78 2.45 2.44 1089
Azucarera
Foundry sands 32.62 1.89 6.97 0.49 2.4 3.68 0.27 0.50
Silica
corrective
Sand 89.61 2.67 3.86 1.53 2.34 13.72 0.69 0.60
Alumina Waste Catalysts 57.68 37.37 0.94 0.12 0.3 1.51 39.76 0.06
corrective Oxyde alumine 0.6 54.2 0.5 0.5 62 0.01 108.4 0.76
Iron Pyrite ash 1.83 2.34 80.93 0.56 12 0.02 0.03 0.93
corrective Mill Scale 3.58 0.96 91.09 1 3.21 0.04 0.01 1.42
Other Contaminated soil 44.14 6.03 3.23 25.2 20.5 4.77 1.87 18.97
corrective Drilling cut base 42.71 10.67 3.61 7.06 22 2.99 2.96 5.25
Phospho gypse 3.56 0.35 0.12 34.08 18.47 7.57 2.92 325.8
Gypsum
Rsidus Pltre 14.75 0.88 2.58 27.97 14.5 4.26 0.34 63.55
Table 4 Examples of alternative raw materials
2.6 AFR Codification List

[AFRCOD-2000] The actual version is attached in Annex 1
It lists all categories of alternative fuels and raw materials used in plants.
9
Silica module
10
Alumina Module
11
Lime Saturation Factor

3. UTILISATION
3.1 Graphs of TSR/ TEE and AR usage in Holcim plants worldwide

The graphs below show the values from ATR12 2002 from all plants reporting
values in the corresponding category (AF or AR)
80
AF solid AF liquid
70
Europe Americas Africa ASEAN
60
50
Substitution [% TSR]
40
30
20
10
MY
MN
MS
ME
MC
SM
MI
TE
TD
TA
HT
ML
TH
CT
TC
TN
OB
KO
PR
DN
HE
RF
HV
GE
CP
RN
CS
GA
LO
BN
EC
OZ
SG
UV
AD
AR
CV
DS
DU
HH
PD
CA
CD
PV
BO
PO
NO
CG
ER
CB
SE
AC
AP
RA
CU
SB
RS
GS
NR
AL
PL
BL
S2
S3
LU
LD
LA
JE
JO
LN
WP
Picture 1 Thermal substitution rate per plant (ATR 2002)
90
80
70
60
[% TEE]
50
40
30
20
10
0
MY
MN
MS
ME
MI
MC
ML
SM
OB
KO
PR
DN
HE
HV
GE
TE
CP
RN
CS
GA
TD
BN
EC
SG
TA
UV
HT
AD
AR
CV
DS
DU
HH
PD
TH
CA
CD
BO
CT
PO
NO
CG
ER
CB
AC
OZ
RA
TC
CU
RS
TN
GS
NR
HN
AL
LU
RF
LD
LA
JE
LO
JO
PV
PL
SE
AP
SB
LN
BL
S2
S3
WP
12
: Annual Technical Report

Picture 2 Thermo economical equivalent per plant (ATR 2002)
26.77
10
8
[% AR based on wet consumption]
0
ME
MI
MC
MP
OB
TE
HT
CT
KO
DN
RN
DU
CD
PO
NO
CG
GD
GS
BE
DF
H2
H1
PR
HE
RF
LU
HV
SH
BN
EC
UV
CK
AP
OZ
JO
AR
CV
DS
AN
LN
DV
LG
AL
BL
S3
LA
JE
Picture 3 Alternative Raw material usage per plant (ATR 2002)
16
14
12
10
[% TSR]
0
1997 1998 1999 2000 2001 2002
Picture 4 Thermal substitution rate development of the consolidated group plants

20
15
[% TEE]
10
0
2000 2001 2002
Picture 5 Thermo economical equivalent development (consolidated plants)

3.2 Feed points for AF
The selection of the feed point for an AF depends mostly on its physical properties.
liquids all feed points
solid < 5mm3 all feed points

alternative fuel
long kiln1 main firing
solids < 50
2nd firing
short kiln2
Precalciner
long kiln1 mid kiln

lump fuel
short kiln2 kiln inlet
via kiln feed 1: long wet/ dry kiln 2: Suspension& grate preheater kiln 3: and foils < 50
Picture 6 Feed point selection for alternative fuels
Comments for the selection of feed points:
Solid fuels of large size tend to produce more combustion problems, especially when
pushed to high substitution rates. So the practical substitution rates are limited. One
way to increase the substitution rates is better preparation, which means size
reduction if it is economically feasible.
In exceptional cases solid fuels can be transferred into a combustible gas by means
of a process integrated gasifier. The gasifier is then the "ultimate preparation" which
allows a comparatively easy burning. Since such solutions are expensive they are
reserved to special applications. For more details about a gasifier solution see
chapter 6.5.
The option kiln feed is no introduction point for alternative fuels because introduction
of organic content at the top of the results in emission.

3.3 Selection of feedpoint
SP-KILN Raw meal Exhaust gas
possible feed points fr AF's
Main firing
Clinker
Raw meal Water

GRATE PREHEATER KILN
Exhaust
Nodules Gas
Granulating
table
Greate
Preheater
Clinker
LONG WET KILN mid kiln Exhaust

Gas
Chain zone
Slurry
Clinker
LONG DRY KILN mid kiln Exhaust

Gas
Chain zone
Raw meal
Clinker
Picture 7 Possible feed points depend on the kiln type

3.3.1 Temperature/Gas Residence Time
The high temperature zone at the main firing ensures good features for a safe
destruction of stable toxic organic compounds in AF.
3.3.1.1 Typical temperatures and gas residence times at the main firing
Flame temperature 1800 2000C
Total gas residence time in rotary kiln depends on kiln system:

short kiln (2 supports) ~ 3s
normal SP/PC kiln > 5s
long wet or dry kiln > 10 s
So typically a gas residence time of 3 - 5 sec. above 1200C can be expected.
3.3.1.2 Secondary Firing and Precalciner

Secondary firing (no PC) 13 1s > 820C
Precalciner with tertiary air 14 2-4 s > 860C
For optimum combustion and safe destruction of stable organics only the main
firing can be used. Example: waste oil that is contaminated with traces of PCB15.
The other feed points are reserved for less critical AFs or if they are used for
critical substances tests are necessary. To avoid extensive tests and ensure
destruction of harmful organics, it is often easier to just select the main firing.
Finally the temperatures mentioned above are not valid for start up or upset
conditions, so critical AFs have to be used only in normal operating conditions.
13
with unextended riser duct
14
In case of hot spot design peaks up to 1200C
15
Poly Chlorinated Biphenyls

3.4 Feedpoints for AR
The most important prerequisite for every AR is the compatibility with the raw mix
(see chapter 4.3). Normally the feed point at the raw mill is the option that is
considered first. However if AR contains organics introduction via kiln feed
requires detailed investigations, because emissions can become a critical point.
(chapter 4.4.2).
Alternative raw material
Suitable for raw mix?
Yes Preparation
Critical emission ? No
(C Content1, Expulsion test2)
cold end
Yes Yes
end of pipe
emission control system3
No
hot zone
Preparation
1: Organic C- Content as first indicator 3: Examples:
2: by HGRS/ MT can predict emissions Thermal Oxidizer @ Dundee removes organics
Activated carbon filter @ Siggenthal
removes organics > 5 C- atoms
Wet scrubber do not remove organics
Picture 8 Feed point selection for alternative raw materials
3.5 Supply and Inlet Control
Supply of AFR to the cement plant can be done basically in two ways:
Get AFR directly from a waste source.
Get AFR through a specialized company that prepares AFR in a form suitable for use
in a cement kiln.
Both ways are being used. The first possibility is adequate for certain materials
that can be used with minimum or no pretreatment. Example: whole tires.
The other possibility is a more sophisticated approach, which involves integration
into the whole waste market. A company is assigned to collect the waste material
and process it to adequate AFR for the cement plant. Specialized waste
companies have emerged in this field.

Classical examples are:
SCORI (France)
SCORIBEL (Belgium/ Obourg)
SYSTECH (USA/ Lafarge)
The organizations were controlled by the cement industry (preferred approach),

whereas independent companies are also been in operation.
The newer companies that have emerged are mostly orientated on the
SCORIBEL/Obourg model. From a technical viewpoint of a cement plant these
companies provide the following functions:
Allocating adequate waste categories to the cement plant

Control of properties
Preparation/pretreatment
(in particular blending/ homogenizing)
The preparation of waste into a AF useable in cement kilns is done externally. This
is the preferred approach.
3.6 Quality Control

There is a paper available "AFR Quality Control Manual" [QCM-2000] describing
the whole process, starting from setting the specifications to the suppliers ending
with standard operating procedures. Examples of AF specifications from different
group companies, as well as laboratory equipment and representative sampling
methods can be found in this paper. It is available on the group E- library in
Environet.

Example of a checklist for properties of AFR
Type of waste Name, trade name origin

Physical state:
Solid Size, form, grindability
viscosity at ...C,
Liquid impurities
Gaseous mixing proportions
solid/liquid
Density kg/m3
Calorific value (net) MJ/kg
Proximate analysis Moisture, ash, volatiles, Cfix
Ultimate analysis C, H, O, N, S
Halogens Cl , Br, F
Ash composition CaO, SiO2, Al2O3, Fe2O3, K2O, Na2O,
P2O5, etc
Heavy metals Hg, Cd, Tl , Be, As, Co, Cr , Pb, Zu, V,
etc.
Flashpoint C
Explosivity non-explosive
Toxicity Toxic /non toxic, safety precautions,
warnings
Legal restrictions containing
transport and storage
Storage Chemical or natural degradation,
putrefaction phenomena,
segregations, precipitations,
Corrosivity Construction materials required
Mixing possibilities Mixing with oil, water, solvents
Quantities to be used min, max, average (now, in future)
Fluctuations in quality Quality specification
Table 5 Checklist for AFR
Delivery control is very important. In case of hazardous wastes this is mandatory

(by the permit). What is less obvious is that even harmless or non-hazardous AFR
need to be checked when delivered to the plant. The main problem is that AFR
can be contaminated with undesirable impurities. Whether this happens
intentionally or not, it needs to be excluded.

Examples:
Possible contamination:
Waste Oil:
The original motor oil would not be critical from its properties, but the waste
product that is finally delivered to the cement plant may be contaminated, e.g. with
Solvents: a small quantity of solvents decreases the flash point drastically and thus
safety can become a issue.
PCB contaminated oils have a high disposal fee and the cement kiln could be abused
to get rid of them cheap.
Waste tires:
When accepting waste tires it is mandatory to have a visual delivery control and
the receiving area must not allow uncontrolled access for various suppliers. Some
plants that believed they could do it without any control finally paid a high price
because they were abused as a dumping area for non usable sizes, rims and other
materials (for which they had to pay for the disposal).
Representative samples
"Never accept quality data until you have checked it."
The diagram below shows a comparison, of analysis of different waste oil
deliveries. One done by the supplier of the waste oil, the other is the crosscheck of
the plant laboratory. Assuming that both values are analyzed correctly, the
difference is caused only by the "difference of sampling".
In this case it is very likely that the supplier of the waste oil was not interested in
taking representative samples since he would not always meet the specifications
of the plant.

Without the samples from the plant a constant heat input in the kiln would not be
possible, and the use of this "high quality AF" results in problems with the kiln
process.
46000
Analysis plant
Analysis supplier
44000
42000
40000
CV gross [kJ/kg]
38000
36000
34000
32000
30000
PICTURE 9: Different analysis of waste oil
3.7 Preparation of AFR

Certain wastes cannot be used directly in the kiln. The properties of the AFR and
the feed point to the kiln system determine how intense the preparation needs to
be. Everything in the range between no preparation (whole tires at kiln inlet) and a
impregnation platform of Scoribel type is possible. The preparation costs have a
large impact in defining the cost-effectiveness of the proposed AFR.
Preparation costs are strongly dependant on the granulometry required. Therefore
the preparation of solid AFR to be used at the main burner is in most of the cases
not the most cost effective solution. Use at kiln inlet or precalciner is generally
preferred, for such materials.
Detailed information on equipment used can be found in the paper "Preparation
and handling of alternative fuels" [PHAF-2003] Another source for equipment used
for feeding of AF, are the so called "Euremi" files on the CoP AFR Application on
Kiln [CoPAoK]

4. IMPACTS
4.1 Circulation Phenomena

Kiln systems with suspension preheater and without bypass are most sensitive to
circulating phenomena. If the following criteria are not respected, the use of AF
can cause severe problems in kiln operation or will even fail.
Total chlorine input below 300 g/t clinker

(from all fuels and raw materials). If this limit is permanently but only
slightly exceeded, the removal of main filter dust (during direct
operation) from the process is necessary to have stable kiln operation.
But if the input is constantly > 350 - 400 g/t a bypass may be required.
With excessive Cl input and without bypass the preheater will just plug
and continuous operation is not possible.
Keep sulfur cycles under control! The most critical factor is not the possible sulfur
input by AF but the impact of poor combustion (lump fuel) on sulfur volatilization. This
will promote a high sulfur cycle and sulfur plugs in the preheater. Remedies: improve
combustion, higher O2 at kiln inlet, enhanced preheater cleaning.
Kiln systems with grate preheater (LEPOL) are of similar sensitivity to circulating
phenomena as cyclone preheater kilns, instead of cyclone plugs the nodule bed
becomes less permeable.
Long dry kilns or long wet kilns are more forgiving in terms of circulating
phenomena as they have no critical narrow cross sections. However, rings and
build-ups in the rotary kiln also occur, but they take longer till they grow to a critical
size. If the kiln system works with 100% dust reintroduction the sustainable
chlorine limit is almost the same as on a suspension preheater kiln (300 g/t
clinker).
The difference to the suspension preheater kiln is that most of the long kilns are
operating with dust discarding, which opens an easy valve for the Cl cycles. Within
certain quality limits such dust can be added to the the cement. Landfilling of such
dust will no more be accepted.

4.2 Potential loss of production capacity
In this paper some of the highlights of the Low Grade Fuel Study [LGF-2001] will
be presented.
If low grade fuels are used to substitute high grade conventional fuels (coal, oil,
gas) the kiln will react with certain effects which increase the thermal consumption
and decrease the maximum kiln capacity.
100%
Induced
thermal losses
Conventional of kiln system
fuel utilisation
Consequence
higher Nm3/kg cli
AF
Consequence
reduced kiln capacity
0%
PICTURE 10: Conventional and alternative Fuels
Both phenomena's are related to each other. If the energy costs for AFs are low or
even negative one may think the increase of heat consumption is not that
negative because the additional consumption can be covered with low cost AF.
This is only half of the truth. Production capacity loss is the logical consequence of
increased heat consumption. Roughly every one % increase in heat consumption
also means one % loss in potential kiln capacity. And if kiln availability is lower due
to AFR the OEE can decrease even further.

4.2.1 Factors for increased heat consumption
Why can AFs increase the thermal losses on a cement kiln system and thus
create induced losses?
There is a number of reasons that contribute to the effect as follows:
1 Water
2 Ash
Reduced
3 combustion
properties
Fluctuating
4
AF feed
Cold air
5
introduction
Circulation
6
phenomena
PICTURE 11: Factors for elevated heat consumption when using AFR
1. High water content in AF

A high water content increases both exhaust gas quantity and exhaust gas
temperature.
Consequence: increased heat loss in the exhaust gas that needs to be
compensated by more fuel.

2. High ash content in AF
A high ash content reduces the amount of kiln feed that passes through the
preheating zone and decreases the heat recovery by kiln feed. The exhaust
gas temperature increases.
Consequence: increased exhaust gas loss that needs to be compensated by
more fuel.
Note: the same effect happens if cold raw material is added directly in the
precalciner.
3. Reduced combustion properties

Certain AFs have poor combustion properties because of too coarse
granulometry. Depending on the control strategy this can mean either
increased CO losses or increased O2 level to compensate this effect.
Example: whole tires at kiln inlet
Consequence: No matter what the strategy, the final result is always a higher
energy loss in the exhaust gas, which has to be compensated by more fuel.
4. Fluctuating AF feed (at good combustion properties)

Fluctuations in AF energy input can result from
Inhomogeneous AF properties (CV)
Fluctuations of the dosing rate due to more difficult handling properties
Both items have the same impact as for item 3. What happens if a temporary
excess of energy input occurs? Either a CO loss is generated or the O2 level
has to be adjusted until the CO level stabilizes on a low level.
Consequence: same as in case 3.
(The difference to case 3 is that this can happen even with fuels with good
combustion properties.)
Example: poorly homogenized liquids or poor performance of dosing system.
5. Cold air introduction

Solid coarse AF usually need a high amount of air for pneumatic injection or an
air leakage can occur at a poorly sealed feed chute for AF. This has the same
effect as an increase of the primary air on a burner.
Consequence: Inleak of additional cold air (due to AF) causes higher fuel
consumption.

6. More circulation phenomena due to AF
Circulation phenomena do not only occur because AF. Even if the AF contain
circulating elements, most of the times, problems are the result of poor
combustion (local CO formation). Although the evaporation and condensation
of circulating elements induces a heat transport from the hot zone to the colder
zones, the direct effect on balance heat consumption is not that critical. What
really disturbs is the unstable kiln operation, which results in reduced
availability and higher average fuel consumption.
Consequence: reduced availability due to unstable operation/stoppages and
increased average fuel consumption.
The majority of the above factors (1, 2, 4 and 5) can be quantified (see [LGF-
2001]) the other factors are based on experience.
4.2.2 Conclusion
The potential fuel substitution value of an AF (USD/GJ) cannot be calculated by
just using its net calorific value. Normally only 70 - 100% of the net CV can used.
Example: 1 t tires @ 28 MJ/kg
1 t coal @ 28 MJ/kg
If the kiln operator adds 1 t/h of tires, he cannot reduce the coal feed
rate by 1 t/h, only 0.8 - 0.9 t/h are possible to cut back.
When reporting the heat consumption of a kiln, we have to accept higher

consumption when using low grade AF.
The potential capacity loss when using AFs is a consequence of the increase in
heat consumption. Increased heat consumption and possibly also reduced kiln
availability can have an important impact on OEE.
4.3 Compatibility with raw mix (quality)

The chemical composition of the AR and the available natural raw materials give
the maximum potential for a given AR in terms of AR substitution rates. A high
CaCO3 source from the quarry usually offers more flexibility to integrate
Silica/Alumina rich materials, be it AR or ashes from AF.
As composition of AR differ much from clinker composition (typically common AR
are low in CaO) substitution rates are usually less than 10%. This is a main
difference to Alternative Fuels where substitution rates (TSR) may range up to
100%.
Compatibility with the mix is a prerequisite but what finally matters is the financial
benefit (min cost of raw mix). Prices and tonnages enter into a raw mix calculation.
The goal of the optimisation is to have the lowest possible cost per ton of kiln feed.
The best benefits can be achieved by either small quantities of high price materials
(iron corrective) or high quantities of moderate price materials.
4.4 Emission
Emissions occur in any cement kiln, but they are hardly caused by AFR, provided
that they are used correctly. Emissions result mainly from the raw material and
from the high temperature process (NOx). The fuels have only a limited influence.

e.g. the SO2 emission on a suspension preheater kiln does not depend on the
sulfur in the fuel. A major challenge of using AFR or applying for a permit to use
them can be how to handle the normal emission, if burning of AFR attracts public
interest and implies more stringent emission rules.
4.4.1 General Features of Cement Kiln Systems

"Barriers" which prevent toxic substances from being emitted or becoming harmful
to the environment (see figure 12)
clinker flame preheating system raw mill filter
5. Incorporation 1. High 2. Intensive contact with 3. Low 4. Efficient

in clinker temp. raw meal temp. edusting
PICTURE 12: Emission barriers of a kiln system
1. High incineration temperature

In the sintering zone flame temperatures of about 2000C is required for
process reasons. Even very stable organic compounds (e.g. PCB) are
destroyed completely. This argument does not apply for secondary firings.
2. Contact with fine, dispersed raw meal

Intensive contact of gas and raw meal is required for process reasons (heat
transfer). It also purifies the gas through absorption of toxic compounds while
contact occurs in counter current flow. Excellent retention of acid gases (e.g.
HCl, SO2) and also of most of the heavy metals is achieved in SP-preheater
and raw mills. The key is the contact of gas with fine suspended particles.
This does not apply for the bypass gas extraction, which must be considered
e.g. in an emission estimate.

3. Low final gas temperature (favorable equilibrium)
Condensation or absorption on surface active raw meal reduces the
concentrations of toxic elements according to the physical/chemical
equilibrium. This effect strongly depends on gas temperature.
The lower the stack temperature the lower the equilibrium concentrations of
the vapors of toxic compounds. Examples for very low stack temperatures
are:
Kiln gas after passing the raw mill (during combined operation) ~ 100C
It is therefore possible to keep emission levels low, while the gases are
passing the raw mill.
4. Efficient dedusting equipment

The high absorption capacity of the kiln system avoids emissions but on the
other hand can cause enrichments of the filter dust with certain elements that
can reach the outer dust cycle (e.g. Tl). Also no EP shut offs are acceptable,
bag filters are preferred. Moreover, excessive dust emissions have an over
proportional negative psychological impact as all neighbors can see it and you
may also risk to lose your permit for AFR burning.
5. Safe disposal of trace elements

Trace elements or heavy metals cannot be destroyed nor can they disappear.
If they are fed into the kiln process and are not emitted they need to have an
outlet. Unlike other incineration systems, producing concentrated and often
toxic by-products, a cement kiln, with complete dust re-usage, offers the
unique possibility to incorporate trace elements in the clinker production in
diluted and immobile form. These trace elements occur in concentrations
which are usually not different from clinker, produced without alternative fuels.
None of them are leachable (exception: hexavalent Cr during make up with
water).
Caution: If the dust is not completely re-used in the kiln or the cement the
above argument - in its simple form - is no longer true.
4.4.2 Emission behavior of alternative raw materials

Alternative fuels are introduced into the high temperature zone of the kiln (main
firing/ secondary firing or precalciner). Therefore organic emissions are not a
special concern for AF as they are destroyed anyway. However alternative
materials are different. As they may contain organics (or volatile inorganics) AR
need special consideration. This type of problem is also known from natural raw
materials containing organics.
The following factors matter:
Shall the AR be introduced via the cold end (kiln feed) or into the hot zone of the kiln?
If hot zone is considered: no potential problem, no investigation
If cold end is considered: case must be investigated

Cold end case :
Does the AR have a potential to develop critical emissions? (usually organic emissions,
but also volatile inorganics like SO2 and NH3 can be produced)
Possibilities to practically assess AR's :

Organic C content gives a first indication
Expulsion test (by HGRS/MT, predicts amount of organic emissions and their nature)
If any potential for critical emissions exists :

Consider also impact of kiln system and emission control system.
As a general rule cold end feed is always critical, but there are exceptions.
Not all kiln systems (including emission control device within the system boundary)
are identical in terms of their emission behavior, differences are as follows:
Long wet kilns offer slightly better burnout than SP kilns due to different temperature
profiles (gas temperature versus material) in the preheating zone. If it is necessary
the burnout can be enhanced by running at higher O2 levels/ higher exit temperatures
(example VOC HH).
Certain "end of pipe" emission control systems allow reduction of volatilized organics.
Examples:
Thermal oxidizer at Dundee removes all organics.
Activated carbon filter at Siggenthal does not eliminate all organics but absorbs
organics with > 5 C-atoms. Contaminated soil can be used via kiln feed.
Wet scrubbers offer only small reduction of organics (just the effect from
temperature reduction)
Exceptional kiln systems (former Fort Collins kiln) where the kiln feed enters
directly into the hot zone without preheating can have advantages. This system
avoids organic emissions completely.
4.4.3 Special rules regarding emission behavior on cement kilns
The reality with emission is usually too complex for a safe and accurate prediction.
However, from an engineering point of view it is better to have some ideas or rules
about the emission behavior to roughly identify what could be critical or not.
Low emission of Cl, F < 0.1% of balance input.
Low volatile heavy metals are not critical. Emission usually < 0.1% of balance inputs.
Medium or high volatile heavy metals can reach the outer dust cycle (Cd, Ti) or even
escape in form of vaporous compounds (Hg). Whereas Cd and Ti are still below <
1% emission, Hg can be emitted almost totally (again this depends on process
conditions).
Any Br Input has to be avoided
SO2 emissions are not depending on fuel sulfur in case of a SP kiln. Wet kilns
however show a moderate influence by S on emissions.
AF burning in the secondary firing usually decreases the NOx (0 30%).
CO as discussed previously is often increased when using the secondary firing.
Fluctuating energy input may also cause CO peaks.
Virtually no organics result from AF burning (even in cases of CO formation it does
not necessarily correlate with organics). So firing of AF is safe, but the storage and
handling needs to be appropriate, in order to avoid additional emission due to AFR.

Dioxine/Furane precursor substances(BTEX16 and especially chlorinated organics
e.g. chlorbenzene) should be avoided in AR and in secondary firing, when a bypass
is installed.
Dioxine/ furan emissions on SP kilns are not critical in view of a limit of 0.1 ng TE/m3.
Generally there is no correlation with AFR use, but in the case when using AF in the
secondary firing and having a bypass installed, there might be the possibility of
Dioxine/Furane generation in the bypass gases.
5. TECHNICAL AFR ASSESSMENT

I got an offer for 30'000 t/a of animal meal, can I burn this?
I'm preparing a business plan for zero fuel cost, are the scenarios
technically feasible?
Determination of technical limits for the kiln is the aim of a technical AFR
Assessment. What are the limiting factors regarding AFR substitution rates of a
specific kiln, is it the combustion limit or are other criteria (e.g. ID Fan at it's
maximum) the limiting points? Which actions, measures have to be taken in order
to fully exploit the technical AFR potential.
These are some questions which can be clarified during such an assessment.
For a determination of the technical limits a close look at the kiln, and the process
parameters is required.
Objectives of the assessment:

Determination of technical limits imposed by the kiln system & the raw material
situation (based on the measuring campaign).
Evaluation of technical AFR potential of each kiln with the potential substitution rate
for each AF family.
Scenarios for zero fuel cost respecting all technical constraints.
Actions, measures.
16
: benzene, toluene, xylene and ethylbenzene

In short terms: Know your process first before using AFR
Assessment coached by HGRS Follow-up
Kick Preparation Workshop Support by

off on site HGRS
Self assessment package
Zero Fuel Cost scenarios
Action Plan / AFR Projects

List of actual and future
Technical AFR Potential
with Fuel Mix Optimizer

Elaboration/Update
Company
Conclusion Meeting
AFR types, quantities, cost/fees of Business Plan
Sending of
per Plant
Progress Report
Self Assessment Realization of

Plant Action Plan &
Measuring Campaign AFR Projects
3 months 1d 1d 1d
Participation of CIE/AFR-BD
PICTURE 12: steps of the technical AFR Assessment
5.1 Basic Data Collection (self assessment)

The first step of the determination of AFR limits is a measurement campaign to
assess the following parameter:
Process
Fluctuations in the composition of material flows, as well as stability of
the kiln are important prerequisites, as well as a good equipped central
control room. To assess the situation a medium term measuring
campaign on the material streams and gathering of process data is
required.
Burner
One of the key elements, a high momentum burner with an adequate amount
of additional channels is very important for a successful introduction of AFR
at the main firing. Detailed investigations on the burner, measuring the
amount of primary air and injection velocities of solid fuels are the most
important values. But not only the burner is important, also the preparation of
the coal (fluctuations in coal fineness) as well as pressure fluctuations in the
coal transport duct and fluctuations in the composition of AF and dosage of
the fuels are important to define technical limits.
Balances
An investigation not only on a mass flow, but on different chemical
components basis is another part of a the assessment of the process. Key
elements in this balances are the sulfur and the chloride balances.

Paretos
A very important tool to find possible problems when introducing AFR are the
detailed failure reports. To find the sources of the actual problems, which
already create downtime, is important, since the introduction of AFR makes
the situation worse. e.g. circulating elements.
Installations
Are there any installations available and how are they performing? This
information can also influence the choice of the most cost effective fuel mix.
A self assessment package is available through HGRS/AFR-Processes.
5.2 Data Evaluation

At the end of the measuring campaign, the data gathered in the previous step is
now compiled into a summary to prepare the workshop afterwards.
Summary
Out of this summary, which lists key parameters of the measurement
campaign potential, bottlenecks can be identified. Also some of the strengths
and weaknesses, regarding AFR usage, of the plant can be identified.
However, some open questions have to be clarified on site during the
workshop.
Technical limits, bottlenecks and limitations of existing AF equipment

In discussion with the plant personnel the bottlenecks and strengths for AF
use can be identified. Also the existing AFR installations can be rated
whether they are sufficient or if they need to be improved.
Specific guidelines for AFR substitution rates per AF family

Out of a list of the general guidelines, together with the results of the
previous steps, the specific limiting factors can be identified. Based on the
current situation guidelines for substitutions rates for every AF family can be
worked out.
Zero fuel cost scenario

Based on the current market information, elaborated in the same time as the
measuring campaign was carried out, different fuel mix scenarios are
elaborated. If the scenarios do not reach the target a gap analysis will be
carried out, to identify the bottlenecks.
Actions & measures

One further step is to define corrective actions to improve the current
situation. For an optimized situation new guidelines for substitution rates can
be elaborated.
5.3 Optimal Fuel Mix & alternative raw material usage

Before being able to calculate an optimal fuel mix more constraints need to be
investigated:

Market research:
Determination of AFR types, properties, quantities and costs for the next
years.
Permits:
The legal situation is different in every country, and changing. The situation
of this permits is also a very important constraint on the way for an optimal
fuel mix (PI-Assessment covers this aspects [PI]).
All the factors shown above result in a optimal fuel mix combined with an
optimized raw mix scenario. Two fields need to be matched, the technical limits
need to be matched with the market constraints. In this field where both criteria are
fulfilled, the optimal fuel mix, and also the lowest possible costs of raw material
can be found to have the lowest possible fuel17 costs per ton of clinker.
Technical AFR Market

constraints
+ constraints
Optimal Fuel & AR Mix

(currently, not static)
PICTURE 13: Finding the optimal AFR Mix
Once an optimal mix is found, it will not stay untouched for years. The market for
wastes, and as a consequence the AFR market, is very dynamic regarding
tonnages available and disposal fees (prices). Furthermore the legal situation can
change quite quickly and have a huge impact on the whole waste business. Every
new opportunity for other AF can change the fuel mix.
5.4 Example of a plant

Different plants have already gone through the process described above. As an
example this chapter shows only the technical steps, since the more long term
actions, like permitting and business plan issues are not yet completed.
All Substitution figures given in this chapter are only valid for this example and are
not intended for general usage
5.4.1 Basic data collection: Results

The following table shows the findings of the measurement campaign, compared
to guideline values from HGRS for this type of kiln system. The example is based
on a suspension preheater kiln with in line calciner (without tertiary air).
17
: Accordng to HARP benefits from AR are transferred to the fuel costs

Current Current
Min Min Min Max Max Max
Value status
1 Material Preparation (based on 5-days campaign)
1.1 Kiln feed
LSF fluctuations 1.12 0 0 1 1.3 sdv
dosing fluctuations 0.49 0 0 0.5 2 %
Coefficient of variation R90m 1.08 0 0 5 10 %
1.2 Coal / petcoke (at main burner) 0 0 0 0
Fineness at R90 m 3.00 0 0 3 4 %
Finesse R90 m fluctuations 0.70 0 0 1 1.5 sdv
Fineness at R200 m 0.11 0 0 0 1 %
dosing fluctuations 0.80 0 0 1 2 %
Pressure fluctuation at the burner +/-5mbar 100.00 0 90 0 %
Pressure fluctuation at the burner +/-5-7.5mbar 0.00 0 0 5 0 %
2 KILN OPERATION
2.1 Combustion (based on 5-days campaign)
O2 at kiln inlet 3.00 3 4 0 0 %O2
CO at kiln inlet 0.05 0 0 0.05 0.1 %CO
Temperature at kiln inlet 950 0 0 1050 1100 C
Temperature at exit lowest cyclone 839 0 0 860 880 C
2.2 Hot meal (based on 5 days-campaign) 0 0 0
% apparent calcination 86.00 0 0 85 90 %
SO3 6.30 0 0 5 7 %SO3
Cl 0.40 0 0 1.2 1.5 %Cl
K2O 1.30 0 0 1.5 1.6 %K2O
Sulfur Volatility 0.85 0 0 0.7 0.9 -
2.3 Main Firing 0 0 0 0
Heat distribution to main firing 75.40 70 75 85 90 % total fuel
Ash input at burning zone 0.80 0 0 3 4 % in cli
Length of fixed coating 5.90 0 0 4 6 x kiln
Fine AFR powder in main flame (diam.<0.5mm) 0.00 0 0 0 30 % total fuel
Solid AFR diam.<1.5 mm in main flame 0.00 0 0 20 % total fuel
Solid AFR diam<5 mm or foils < 50 mm 0.00 0 0 10 % total fuel
0 0 0 0 % total fuel
comment: all substitution refer to total sub. of the fuel and can not be accumulated
2.4 Precalcination, secondary firing or MKF 0 0 0 0
Heat distribution to precalcination 24.60 10 15 25 30 % total fuel
Whole tires 12.00 0 0 12 % total fuel
Lump fuel >50mm 0.00 0 0 15 % total fuel
Solid AFR < 50 mm 0.00 0 0 20 % total fuel
3 MAIN BURNER
% Primary Air( radial/axial) 12.00 10 12 15 20 %
Axial Momentum 7.60 7 10 11 0 N/MW
coal injection velocity 26.40 0 25 30 0 m/s
4 CHEMISTRY
4.1 Clinker Current Min Min Min Max Max Max Current status
Liquid phase 1450C of main clinker 23.29 0 23 25 0 % in cli
CaOfree 0.78 0 0.8 1.5 0 % in cli
% off-spec clinker 3.30 0 0 3 5 % in cli
P2O5 in Clinker 0.12 0 0 0.45 0.5 % in cli
0 0 0 0
4.2 Alkali & Sulfur (based on A/S balance) 0 0 0 0
molar A-Cl/S - ratio (input) 0.56 0.75 0.8 1.2 1.25 -
Sulfur input (process point of view) 1.80 0 0 1.5 2 % in cli
Alkali input (process point of view) 0.87 0 0 0 1.5 % in cli
0 0 0 0
4.3 Chlorine & Dust 0 0 0 0
Chlorine input from raw materials & fuels 360.00 0 0 200 300 g/t cli
Chlorine output through all cement 170.00 0 0 800 1000 g/t cem
Filler in cement 0.00 0 0 5 6 % in cem
5 KILN PERFORMANCE
Current Min Min Min Max Max Max Current status
OEE net 88.30 75 85 0 0 %
Rate 96.50 90 95 0 0 %
Availability 91.40 85 90 0 0 %
MTBF 444.00 100 240 0 0 hours
Number of kiln stops 12.00 0 0 40 80 stop/a
Stops due to cyclones plug-ups 5.00 0 0 5 10 stop/a
Refractory consumption (4 years rollings av.) 609.23 0 0 400 600 g/tcli
Emissions (below legal limits: "Yes" or "No") Yes 0 0 0 0 -
PICTURE 14: Summary of an AFR Assessment

5.4.2 Potentials& Bottlenecks:
Bottlenecks:
Chlorine input
AFR solids (low grade) with highly fluctuating chlorine contents
enrichment by dust recirculation from baghouse
Alkali/ sulfur Ratio

Raw materials with low content of alkalis
High sulfur input by petcoke
Refractory consumption
High castable consumption:
2000 Riser duct reparation
2001 & 2002 satellites modification
2001 - High bricks consumption. The high and low transition zones were changed
twice due to initiate using petcoke & AF
2002 Between 500 & 600 g/t due to AF liquid (high grade) dosing at main burner
without an adequate preparation.
Burner Momentum
Primary air fan at 100% of the capacity but limited to 180 mbar pressure (7.6 N/MW
axial momentum)
LSF Fluctuations
Fluctuations are < 1.5 stdv LSF (plant objective: <1.2)
O2@ kiln inlet

Kiln inlet probe does not work due to high temperature reached in that zone
Measurements at the preheater tower do not have a correlation with gas
measurement at kiln inlet due to AFR feeding at secondary firing.
Potentials:
Potential for more AFR to be used in the main burner
Limits of the AFR installations:

AFR solid, tires & bags
Only tires or bags are fed at one time
Fuel substitution not more than 15%
Solids size larger than 500 mm are fed
Temperature control needs 5% of heavy oil
Roller conveyor causes obstruction of bags

Hot sludges at main burner
Variable humidity from 10 to 30%
Poor atomization. Needs appropriate nozzle and compressed air
Drill cuttings
Belt conveyor makes operation dependant on AF density and fluidity
Feeder equipment needs increased capacity, presently 3.0 t/h
5.4.3 Potential substitution rates per AF family and kiln
Main Firing Kiln inlet

Liquid AF (high grade) < 85% < 20%
Liquid AF low grade) separate separate
injection injection
Whole tires - < 12% *
Bags (20kg) - < 8% *
Lump fuel < 200mm - < 15% *
Drill cuttings (7 MJ/kg) - < 4% *
Solid AF < 50mm - < 20%
Solid AF < 5mm (foils < 10% not interesting
< 50mm)
Spherical solid AF < 20% not interesting
<1.5mm
Spherical solid AF < < 30% not interesting
0.5mm
* only in combination with min 5% fuel oil
Substitution rates must not be added up!

These rates are specific for this kiln, no general usage!
Table 6 Specific substitution rates (maxima, combustion limits)
5.4.4 Actions
1 Verify the 1.68% O2 at the kiln inlet without AFR 31-03-03

2 Repair kiln inlet gas probe and install special operating 31-03-03
mode (15 minutes in; 45 minutes out)
3 Install new atomizing nozzle for lodos de fondo de 15-10-03
tanque
4 Increase sulfur in the clinker by: 15-08-03
inclined burner well above kiln axis
clear excess air in the kiln (3-4% O2)
highest possible fineness of petcoke

5 Study installation of new primary air fan to increase 31-06-03
momentum to > 10 N/MW (HGRS to supply offer of 350
mbar fan)
6 Realize tests with burner inclination well above kiln axis 18-08-03
7 Assure consistent fineness of petcoke < 3% R90m 31-03-03
8 Shorten the flame by adjusting the existing burner 30-04-03

settings
9 Check maximum possible percentage of filler in the 30-04-03
cement
10 Separate liquid AFR of low quality (<7000 kJ/kg) from 30-12-03
the main burner by separate injection tube
11 Increase free lime from 0.8 to 1.2 % 30-04-03
12 Compare chlorine content of hot meal (X ray vs. wet 30-04-03
chemical analysis)
13 Realize each month a chlorine balance to monitor 30-04-03
excess chlorine input
14 Repeat tests with raw materials of higher alkali content 30-06-03
to facilitate evacuation of SO3 by clinker
15 Realize project to extract the dust during direct operation 31-04-03
and to put it into the cement mill; maximize periods of
compound operation
16 Make tests with reduced fineness of raw mix to get a 31-04-03
more stable kiln operation and, thus, to increase the
amount of AFR (during a period of at least 2 weeks )
17 Control the bottom cyclone exit temperature by the 30-04-03
calciner fuel
18 Study the application of petcoke in the calciner 30-08-03
20 Study modification of the solid AFR feed points at 30-08-03
preheater
Table 7 Actions of the AFR Workshop
5.4.5 Zero fuel cost scenarios

Two scenarios:
Scenario 1: is the budget for 2003,
Scenario 2: minimal fuel cost based on current market data

Scenario 1 K3 Scenario 2 K3
Gap Analysis: Scenarios
Specific fuel cost at burner tip USD/t cli 2.64 0.45
General Lost capacity due to AFR %OEE 4.9% 7.1%
Dandelion is limited by Cl and S input Specific heat consumption kJ/kg cli 3,700 3,752
- stabilize operation to increase S in clinker Chlorine input g/t cli 361 448
(see actions) 1.67
Sulfur in clinker % SO3 1.64
- Reduce chlorine by dust extraction
A/S - 0.41 0.41
Scenario 1 (budget 2003): 3.3
Solids < 50 mm at PC % 15.0
Chlorine input requires dust extraction Heat to main burner % 75.0 76.2
Theoretical sulfur concentration in clinker
does not comply with practical experience Noble fuels contracted available noble fuel noble fuel
minimum (t) maximum (t) consumed (t) consumed (t)
Experiences have to be gained with drill
BZ PC
cuttings (clinker quality)
Heavy Oil 10,000 6,320
Scenario 2: Antracite
Above measures allow approx.0.5 USD/t cli Petcoke 63,583
in kiln 3. Fuel Oil 2006
Petcoke 2006 100,000
In order to keep the maximum chlorine input Mix coal coke 1:2 100,000 62,706
limit, MMOs and drill cuttings have been
reduced in kiln 3 and used in kiln 2
Alternative contracted available alternative fuel alternative fuel
National petcoke has been blended with
fuels minimum (t) maximum (t) consumed (t) consumed (t)
anthracite (2:1) in order to keep the limit of
BZ PC
sulfur input at 1.64% SO3
Tires 2,959
Further cost reduction potential by: Lodos de tanque 20,000 20,000 6,000 14,000
- In long term avoid use of lump fuel and shredded tires solidos impregna 7,550
in order to increase petcoke Waste Oil 1,000 1,000 600 1,000
- Agreement with client to produce drill cuttings with low Agua contaminad 800 800 660 800
Cl content Fine solid AF
- Increase AR Drill cuttings 27,000 40,000 10,000 27,000
MMO 9,000 3,000
Shredded Tires 10,000
Shredded Solid A 14,000 14,000 14,000
Picture 15 Gap Analysis and Fuel Mix Scenarios
6. APPLICATION
The aim of this chapter is to give a overview of different installations which are
used to burn AF. As mentioned before, there is a separate paper available dealing
only with handling and preparation of AF. [PHAF-2003]
The most often used alternative fuel tires serves here as an example that a certain
AFR can be fed to the kiln in different locations and in a variety of physical
properties. Furthermore this chapters gives reasons why it is hardly possible to
burn communal waste. To complete the chapter and to have an example of every
fuel family, the impregnated sawdust installation of Eclpens and liquid AF
installation from Ras el Ma are shown with flow charts.

6.1 Tires
Tires are one of the first alternative fuels which have been used in cement kilns.
Many installations using tires are in operation and since this AF is almost
everywhere available, they are well distributed all over the world. The
attractiveness depends on the local disposal fees. Because of the large number of
application this fuel is used in many different ways, most of them are described
further in this chapter.
Overview of methods to use tires at a kiln
Energy potential
Calorific value:18 28 to 32 MJ/kg
In a industrialized country up to ten percent of the fuel could theoretically be
substituted by tires if all tires would be available to the cement industry.
Tires are an important energy source and so far the most frequent application of
AF. The practical attractivity depends on the disposal fee that is available. Current
values (for whole tires in different areas) range from 0 60 USD/t.
18
CV depends on quantity of steel included

Typical Composition of Tires
Constituents
Rubber 36.0%
Filler (soot, SiO2) 37.0%
ZnO 1.2%
Softeners 3.0%
Sulfur 1.3%
Steel, textiles 18.0%
Rest 3.5%
Total 100%
Table 8: Tire composition
Chemical analysis
C 70% Cr 97 ppm
H 7% Ni 77 ppm
S 1 ...3% Pb 60 to 760 ppm
Cl 0.2...0.6% Cd 5 to 10 ppm
Fe 15% Tl 0.2 to 0.3 ppm
ZnO 2%
SiO2 + rest 5%
Table 9: chemical analysis of tires
6.2 Whole Tires

This is the most frequent application, based on the secondary firing principle.
Originally, this was first developed and used on dry SP-kilns but then also
extended to long dry or wet kilns.

6.2.1 Kiln inlet on suspension preheater kilns
A feed system according to the figure below involves an investment of around
USD
1- 2 Mio.USD for a fully automatic installation. Typical substitution rates are 5 - 20
% of the total heat consumption. In the case of precalciner kilns potentials are
lower.
Tire individuation : At least four 90 changes

of direction of (roller)
conveyors.
Saxlund Box feeder with push rod extraction

(cap : 1 d )
Concrete storage pit

( cap : 1 W ) Roller conveyors
Tyre size
(Metal detector
monitor
for wheel rims)
Weigh feeder
long belt Oversize tires

Other lump fuels feed point
Double flaps gate
(activation by light
barrier)
shut off gate Kiln capacity : 1522 t/h
Tire capacity : 0,8 - 1,6 t/h
PICTURE 16: Whole tire burning at kiln inlet in Koromacno (Croatia)

6.2.2 Mid Kiln on long dry or wet kilns
The principle is similar to the solution above but the introduction chute is rotating
with the kiln shell, i.e. the tire feed is coupled with the kiln revolution. Figure 9
shows the introduction chute of the "Cadence" system as used at the Joliette plant.
The Substitution rate is similar to the solution at the kiln inlet.
flap opens kiln 3.6 x 122m

tire discharge
flap closed
free opening
0.33 x 0.91m
loading position for tire
PICTURE 17: Whole tire burning at mid kiln in Joliette (Canada)

6.3 Shredded Tires
Shredded tires allow a less fluctuating fuel input and have a higher density
(advantages for transport and storage).
Shredding costs19 are in the range of 30 - 60 USD/t. Sometimes this is already
paid by disposal fees. Shredding is normally not operated by the cement producer.
Because of the better combustion properties more feed points are available.
The usage of shredded tires at the main burner is only applicable on long wet/dry
kilns, whereas the usage of tire chips in the precalciner is the way to use this fuel
on precalciner kilns.
The use of shredded tires < 300 mm on suspension preheater kilns at the kiln inlet
, similar to the installation for whole tires, is rare. Reasons: It would cost less to
use whole tires and the thermal substitution rate cannot be raised.
6.3.1 Main burner

The use of shredded tires < 50 mm has some applications on long kilns in North
America where still many long dry and wet kilns are in operation. The tire chips are
injected into the burning zone, by a pipe on top of the burner. Figure 18 shows a
typical example of a long wet kiln installation. Substitution rates of up to 15 % can
be achieved.
This application cannot be used in suspension preheater and grate preheater kilns.
1700 m3/h
Coal / oil / gas TDF injection pipe
burner 127mm 37m/s
long wet kiln

4.27 x 152m
0.34 bar
~ 20m
Fuel
TDF pipe
Walking
Floor
PICTURE 18: Shredded tire burning in a wet kiln
19
operating and owning costs

6.3.2 Precalciner
Shredded tires or TDF20 < 50 mm are successfully used on precalciner in several
plants (Midlothian, Theodore, Lgerdorf and others). Substitution rates of up to 30
% can be achieved with this type of installation.
Tire Chips
Fuel
< 50mm
TA
Theodore:
BDP: 4645 t/d
Fuel Split: 50/50
Heat cons: 3490 kJ/kg cli
Sub. Rate TDF: 17%
Push bottom
floor bunker
PICTURE 19: Shredded tire burning in the PC at Theodore (USA)
6.4 Ground Tires at the main burner

Theoretically, ground tires would be the ideal fuel for any primary firing (without
compromise). However, the costs for grinding are usually prohibitive. Ground
rubber as granulate is normally more expensive than fuel. Nevertheless, a group
plant in Germany and Holcim (Switzerland) have tried this but finally the had to
give up. Except from fines, from tires re-surfacing, there's hardly any source
available for ground tires
6.5 Pyrolisis/ Gasifier

In Japan, some gasifier for cement kilns have been reported a long time ago. They
were based on a reactor (shaft) with understochiometric air addition at 700C. The
hot gas produced was sent directly to the cement kiln.
In 1999, a new gasifier for whole tires (industrial scale) was commissioned by
Polysius at Jura Cement in Wildegg (Switzerland). The hot combustible gases are
used in the precalciner. Investment for a 3 t/h installation is about 3 Mio. USD.
The gasification can potentially help to optimize the use of tires, which still needs
to be demonstrated on a long-term basis. The costs are significantly higher than
for burning the tires directly. Direct burning - not gasification - should always be the
first option to be investigated.
20
TDF = tire derived fuel
PICTURE 20: Gasifier for tires at Wildegg (Switzerland)
6.6 Domestic Refuse (usually not recommended)
Energy Content
Domestic refuse Cement consumption

Quantity 400 kg/person year 450 kg/person year
Energy content 3.4 GJ/person year 1.7 GJ/person year
(heat energy)
Table 10: Waste figures (all numbers from Germany)
The energy content contained in the domestic refuse is twice the energy
consumption of the cement!
However, a complete use of the energy from raw domestic refuse in the cement
industry is not feasible.
Reasons:
poor homogeneity, inadequate size, difficult handling
Cl-content of 0.5...1% Cl which can cause clogging problems in the kiln
low calorific value (8 to 10 MJ/kg)
low density and high transport costs per heat unit
competition to existing incinerators
Conclusion:

Domestic refuse needs intensive processing in order to eliminate undesired
fractions and to obtain a reasonable burnable fraction. Such a fraction may
represent 30...50% of the original refuse, the rest needs further disposal. The
burnable fraction is called RDF (refuse derived fuel) and offers better properties,
e.g. a calorific value of 12...16 MJ/kg.
Experience:
The first application was in the early 80s at BCI/Westbury, was stopped.
The most important application today is the Wittekind plant in Erwitte (Germany).
According to figure 21 with 50% substitution and a chlorine bypass. Otherwise very
few plants have realized major applications.
1/3 Light fraction
Drum (ZSS)
25 t/h
RDF
(fluff)
600
2/3 Heavy fraction
kg/m3
to landfill RDF
Baling press
(former quarry) (bales)
RDF 50% fuel substitution
n = var
10 t/h Cutting M
Coal Prim.Air 2000 t/d kiln
mill 6 - st.suspension -
RDF Preheater + CI - Bypass
(fluff) Type MBM
< 50mm
13 MJ/kg
Special
130 kg/m3
Burner
Example of the 2000 t/d suspension preheater kiln of Wittekind in Erwitte (Germany)
PICTURE 21: RDF Burning in Erwitte (Germany)

6.7 Solid AF < 5 mm: Imp. Sawdust in EC
The impregnated sawdust is produced by the supplier and the properties can be
specified. Since the emission of the sawdust (containing volatile organics) is
harmful, open storage is not an adequate solution. The substitution rate depends
on the properties of the impregnated sawdust.
rotary feeder
magnetic
separator
rotary separator air
holes: 25mm blower
integrated
scale separated injector pipe
material 100mm
PICTURE 22: Impregnated sawdust burning at Eclpens (Switzerland)
6.8 Liquids (Ras el Ma)

Burning of liquids, especially waste oil, has a long tradition. The use of solvents
and waste water became more and more important since there are high disposal
fees available. To be able to use solvents, the installations have to meet the safety
regulations for handling liquids with low flashpoints.
To be able to have a better control on the product which is fed to the burner, it is
important to have two tanks instead of one. So one can be filled and the other one
stays in use for feeding the kiln with a liquid AF, where the properties have been
measured, and do not change with every new delivery.
Burner
N2
Pressurized
Air
PICTURE 23: Liquid installation in Ras el Ma (Morocco)

7. REFERENCES
[QCM -2000] AFR Quality Control Manual

Available in Environet on HolSpace
[PHAF-2003] Preparation and Handling of alternative Fuels
Alexander Flacher, 2003
On the Cement Course CD

And in the group e- Library on ICM (HolSpace)
[CoPAoK] Community of Practice AFR Application on Kiln.
All technical aspects of AFR Usage are treated there
Available on Holspace
[CoPAFR] Community of Practice AFR,
dealing with all aspects of AFR (mostly non
technical)
Available on HolSpace
[PI] Pi Pyramid
A tool for AFR business development
To be found on HolSpace
[CoPPet] Community of Practice Petcoke
Dealing with both, technical and non technical
aspects of Petcoke usage at the plants.
Available on HolSpace
[LGF-2001] Low Grade Fuel Study
Impacts of AFR on the kiln process
A. Obrist, D. Pauling, A. Flacher, TPT 00/21018/E
On the Cement Course CD
Also available on ICM (HolSpace)
[AFRCOD-2000] AFR Codification List
Current version attached in Annex 1
Actual Version to be found in Environet on Holspace

ANNEX 1
AFR Codification List:
English French Spanish
A) Fuels
Waste oil Huiles uses Aceite gastado y/o usado

Emulsions Emulsions Emulsin
Non-chlorinated solvents Solvants non chlors Solventes no clorados
Chlorinated solvents Solvants chlors Solventes clorados
Synthetic fuel Combustible synthtique Combustible sinttico
Paint sludge Boue de peinture Lodos de pinturas
Waste wood Dchets de bois Residuos de madera
Impregnated saw dust Sciures imprgnes Aserrn impregnado
Refuse derived fuel Dchets de fuel (?) Combustible derivado de residuos
Dried sewage sludge Boues d'puration sches Lodos secos de aguas residuales
Waste tires Pneus usags Llantas usadas
Shredded tires Pneus dchiquets Llantas trituradas
Waste rubber Dchets de caoutchouc Residuos de hule
Mixed plastics Plastiques mlangs Mezcla de plsticos
Polyethylene Polythylne Politileno
PET PET (polythylne trphtalate) PET (Polietileno Tereftalato)
PUR foam Mousse de polyurthane Espuma PUR ?
Polysterene Polystyrne Polistireno
Contaminated soil Terres contamines Suelo contaminado
Spent pot liners Revtements usags de chaudires Liner
Electrode dust Poussires d'lectrodes Polvo de electrodo
Distillation residues Rsidus de distillation Residuos de destilacin
Filter tar Goudron Brea de filtro
Petrochemical tars Goudrons ptrochimiques Petroqumicos de brea
Diaper residues Dchets de couches culottes Residuos de fabricacin de paales
Spent accumulators Batteries usages Acumuladores gastados
Sterilized hospital waste Dchets hospitaliers striliss Residuos hospitalarios esterilizados
Bio-medical waste Dchets bio-mdicaux
Fuller earth Terre foulon (argile) Tierro fuller
RESH RESH RESH?
High carbon fly ash Cendres volantes haute teneur en carbone Ceniza volante alta carbono
Landfill gas Gaz de dcharge Gas de confinamiento
Vegetable oil Huile vgtale Aceite vegetal
Spent lubricants Lubrifiants usags Lubricante gastado
Pressed olive cake Gteaux d'olives presses Pasta de filtro de aceite de oliva
Waste coal Dchets de charbon Residuos de carbn
Coal sludge Boue de charbon Lodo de carbn
Tar of TDI Goudrons de TDI (Tolune diisocyanate) Brea de tolven diisocinato
Acid sludge Boue acide Lodo cido
Waste charcoal Dchets de charbon de bois Residuos de carbn vegetal
Active carbon Charbon actif Carbn activado
Black carbon Noir de carbone Carbn negro
Palmnut shells Coquilles de noix de coco Cscara de nuez
Peanut shells Cosses de cacahutes Cscara de cacahuate
Oily sludges Boues huileuses Lodos aceitosos
Hydrocarbon sludges Boues d'hydrocarbures Lodos con hidrocarburos
Paper sludge Boues de papeterie Lodos de papel
Waste paper Dchets de papiers Residuos de papel
Cardboards Cartons Cartn
Textile waste Dchets de textiles Residuos textiles
Magnetic tapes Bandes magntiques Cintas magnticas
Fir cones Pommes de pin Fibras de abeto
Bark Ecorces Cortezas
Waste glycole Dchets de glycol Residuos de glicol
Chemical waste Dchets chimiques Residuos qumicos
Rice husks Coques de riz Cscara de arroz
Graphite dust Poussires de graphite Polvo de grafito
Anthracite dust Poussires d'anthracite Polvo de antracita
Animal meal Farine animale Harina de animal
Animal fat Graisse animale Grasa de animal

B) Raw Materials & Mineral Additions
Coal fly ash Cendres volantes (de centrales charbon) Carbn de ceniza volante
Foundry sand Sables de fonderie Arena de fundicin
Foundry dust Poussires de fonderie Polvo de fundicin
Blastfurnace slag Laitier de haut fourneau Escoria de horno
Metallurgical slag Laitier mtallurgique Escoria metalrgica
Steel slag Laitier d'acirie Escoria de acero
MSWI slag Laitier MSWI (?) Escoria MSWI (?)
Mill scale Dchets de meunerie (farine, bl) Escamas de molinos
FCC Catalysts Catalyseur FCC Catalizador de FCC
Absorber lime Chaux d'absorption Cal absorbente
Lime sludge Boues rsiduaires de chaux Lodo de cal
Reactor lime Chaux de racteur Cal de reactor
Water treatment sludge Boue de traitement des eaux Lodo de tratamiento de agua
Road sweepings Boues de curage Borredura de caminos
Bottom ash Scories Ceniza de fondo
Filter dust Poussires de filtre Polvo de filtro
Pyrite ash Cendres de pyrite Ceniza de pirita
Iron scale Oxyde de fer Escama de hierro
Filter press cake Gteaux de filtration Pasta de filtro prensa
Mine tailings Dchets miniers Sales minerales

Aluminium sludge Boues d'alumine Lodos de aluminio
Red mud (Alu) Boues rouges (Aluminium) Lodo rojo
Diatomaceous earth Terres de diatome Tierra diatomcea
Metal cuttings Boues de rectification Cortes de metal
Abrasive blasting sand Sables de dcapage Arena abrasiva de soplado
Paper ash Cendres de papier Ceniza de papel
Railway aggregates Ballast Agregados de ferrocarril
FGC gypsum Gypse FGC Yeso FGC?
Chemical gypsum Gypse chimique Yeso qumico
Borehole sludges Boues de trous de mines Lodos de perforacin (recortes)
Harbour dredge sludge Boues de dragage de port Lodo de dragado de puertas
AFR Codification list (V 12.00)
Actual version can be found

on the Group e- Library
in the ENVIRONet (HolSpace)

Low Grade Fuel - Study
A. Obrist, D. Pauling, A. Flacher

TPT 00/21018/E
1. Introduction.................................................................................................................... 419
2. Targets............................................................................................................................ 419
3. Maximum Substitution Rates and Feed-point Selection............................................ 420
4. Impact of LGF on clinker burning process ................................................................. 423
4.1 General ................................................................................................... 423
4.2 Basic thermal phenomena/ limitations................................................ 424
4.3 Impacts on heat consumption ............................................................. 425
4.4 Impacts on exhaust gas temperature.................................................. 428
4.5 Impacts on flame temperature ............................................................. 429
4.6 Impacts on power consumption .......................................................... 431
4.7 Impacts on production rate .................................................................. 431
4.8 Impacts on availability .......................................................................... 434
4.9 Impacts on clinker quality .................................................................... 434
5. Criteria for LGF properties (in view of cement kiln) ................................................... 436
5.1 Combustion properties (granulometry) .............................................. 436
5.2 Minimum flame temperature (water and ash) ..................................... 436
5.3 Homogeneity.......................................................................................... 437
5.4 Circulating elements: chlorine, sulfur and alkalis ............................. 438
5.5 Environmental: Heavy metals and PCB .............................................. 439
5.6 Clinker quality........................................................................................ 440
5.7 CKD (Cement Kiln Dust) quality (if CKD is discarded) ...................... 440
5.8 Health and safety................................................................................... 440
5.9 Handling properties .............................................................................. 441
6. Adaptations / Modifications to Compensate or Alleviate the Impacts of LGF......... 442
6.1 Assessment of actual situation: Process indicators ......................... 442
6.2 Enhanced process control ................................................................... 443
6.3 Enhanced control of circulating elements (SO3, K2O, Na2O and Cl)445
6.4 Enhanced preheater cleaning (cyclone preheater kilns only) .......... 447
6.5 Quality control (LGF ash compensation)............................................ 448
6.6 Improved homogenization and dosing of LGF................................... 448

6.7 Dust extraction ...................................................................................... 449
6.8 Bypass technology ............................................................................... 451
6.9 Increased gas handling capacity ......................................................... 453
6.10 Increased critical cross sections......................................................... 455
6.11 Burner adaptations ............................................................................... 455
6.12 Cooler modifications............................................................................. 456
6.13 Precalciner modifications..................................................................... 457
6.14 Mixing air on long kilns (connected to mid kiln firing) ...................... 458
6.15 O2 enrichment ....................................................................................... 458
6.16 Gasifier / Thermal Cracking.................................................................. 461
6.17 Drying of LGF ........................................................................................ 468
7. Criteria for a new cement kiln in view of future use of LGF ...................................... 469
7.1 Summary of requirements for a new kiln using AFR......................... 469
7.2 General ................................................................................................... 469
7.3 Criteria for preheater and exhaust gas system design ..................... 470
7.4 Criteria for bypass and related items .................................................. 470
7.5 Criteria for calciner ............................................................................... 470
7.6 Criteria for rotary kiln............................................................................ 470
7.7 Criteria for main burner ........................................................................ 471
7.8 Criteria for grate cooler ........................................................................ 471
8. Set of rules how to handle LGF (storage, dosing and injection into kiln)................ 472
9. Interaction LGF preparation and burning ................................................................... 474
9.1 Cost relations to prepare LGF.............................................................. 474
9.2 Substitution rate versus quality of LGF: Optimization of costs ....... 475
SUMMARY:
What are Low Grade Fuels (LGF):

Alternative fuels with particularly low CV or high ash content or high water content or high
circulating elements or coarse granulometry.
Objectives of the study:

Quantification of the impacts of LGF on the kiln process
List of criteria for LGF properties in view of the cement kiln to understand the critical items
List of the solutions to compensate or alleviate the impacts of LGF on the kiln process
Impacts of LGF:
Within the study most of the practical impacts on the cement kiln are given in quantitative
and direct form. E.g. the impact of injected water is directly expressed per ton of injected
water.
The following impacts are described and quantified:
Impact on heat consumption
Impact on kiln production rate
Impact on power consumption
Impact and conclusions on flame temperature
Impact on availability (description of practical examples)
Impact on clinker quality
Important results are:

Reduction of kiln capacity by water injection:
Cyclone preheater kiln: - 2 t clinker / t
water
Grate preheater kiln: - 1.7 t clinker / t
water
Long wet and dry kiln: - 0.8 t clinker / t
water
Increase of heat consumption by water injection:
+ 4.6 GJ / t water (directly injected
water)
+ 2.15 GJ / t water (water included in
fuel)
Increase of heat consumption by false air (or transport air):
+ 1 MJ / Nm3 FA
Increase of heat consumption with excess air:
+ 1.8% heat cons. per % additional O2
Criteria for LGF properties:

The following criteria for LGF properties in view of the cement kiln are listed and limits
are given:
Granulometry in function of kiln type and feed point
Maximum water and ash input
Homogeneity
Maximum input of chlorine, sulphur and alkali
Most important heavy metals and PCB
Limits for elements that effect the clinker quality
Health and safety aspects
Handling properties
Compensating the impacts of LGF:

The possibilities and solutions how to compensate or alleviate the impacts of LGF are
explained:
Assessment of actual situation: Process indicators
Enhanced process control
Enhanced control of circulating elements
Enhanced preheater cleaning (cyclone preheater kilns only)
Quality control (LGF ash compensation)

Improved homogenisation and dosing of LGF
Dust extraction
Bypass technology
Increased gas handling capacity
Increased critical cross sections
Burner adaptations
Cooler modifications
Precalciner modifications
Mixing air on long kilns (connected to mid kiln firing)
O2 enrichment
Gasifier
Drying of LGF
New kilns:
Criteria for new kilns in view of future use of LGF are listed.
Even if the future LGFs are not clearly known in the project phase, sufficient flexibility for
LGFs must be provided. In particular the design of rotary kiln, precalciner and exhaust gas
system must be reviewed as they hardly can be corrected later on.
LGF equipment:
A set of rules is given how the equipment at the plant for storage, dosing and injection into
the kiln should be designed to ensure the maximum possible flexibility with regard to
changing LGF properties.
Preparation costs:
Finally the costs for preparation (shredding) of different types of solid LGF are summarised
and an example of substitution rate versus particle size is outlined.

1. INTRODUCTION
What is a low grade fuel (LGF):
Alternative fuel where at least one of the following factors is critical:
low calorific value (CV)
high ash content
high water content
high circulating elements content
coarse granulometry
From the view of the clinker kiln LGFs are far away from the requirements and rules
that were traditionally applied for normal fuels. LGFs are thus a challenge as they
may disturb the process. The incentive is their low price. The question is how far we
can challenge the process until negative impacts offset the low price. And how can we
avoid or alleviate these negative impacts ?
This study offers the tools to quantify impacts. To keep it simple and systematic the
various impacts (e.g. water, ash, ..) are presented as single items at once (assuming
there would be no other simultaneous impacts). In reality various impacts may occur
simultaneously. Logically, at combined impacts, maximum loads can not be cumulated
(water plus ash plus CKD plus poor combustion plus ). The single items have to be
finally put together by the user himself. A combined numerical example is listed in
annex 1.
2. TARGETS
The study has the following targets :
Be a technical guideline
Define the problem, explain limits of LGF and the solutions
Usable by LGF providers so that they understand what the critical AFR items in
view of the cement kiln are
Usable by plants burning LGF, as a set of rules
Tool of the CIE-Pyramid
Element of future courses / seminar (AFR part)

3. MAXIMUM SUBSTITUTION RATES AND FEED-POINT SELECTION
Main firing versus secondary firing:

For LGF with stable toxic organic components only the main firing is used because of
the higher temperatures and the longer retention time which allow to completely burn
these substances (example: waste oil with traces of PCB).
If the main firing can not be used and the LGF contains toxic organic components,
tests are needed with emission monitoring of possible harmful decomposition
products.
Kiln Feed:
The feed point via kiln feed is forbidden because of the emission problems generated
during preheating (VOC, CO). This feed point is reserved for alternative materials with
no organics. The only exception would be kiln systems where the kiln feed enters the
combustion zone without preheating (e.g. one stage precalciner kiln at Fort Collins) or
kilns with VOC removal systems (activated carbonfilter at Siggenthal, thermal oxidiser
at Dundee).
Gasifier:
In exceptional cases solid fuels can be transferred into a combustible gas by means of
a process integrated gasifier (chapter 6.16). This allows an application at any feed
point accessing the high temperature zone of the kiln (not displayed in the figure
below).
Possible feed points :

Possible feed point for LGF
Exhaust gas
Raw meal
PRECALCINER KILN
Precalciner firing Kiln
PC inlet firing
Main firing
Clinker
Exhaust gas
Raw meal
PREHEATER KILN
Secondary firing
Main firing
Exhaust gas
Raw meal
Water
Clinker
Secondary firing
Nodules
GRATE PREHEATER KILN
Granulating
table
Main firing
Clinker Grate preheater
Mid kiln firing

LONG WET AND DRY KILN Exhaust gas
Chain zone
Main firing
Raw meal
Clinker

Maximum Substitution Rates
All substitution rates refer always to total substitution of the fuel and can not be
accumulated (e.g. preheater kiln at main firing: either 30% substitution with solids <5mm or
max. 50% substitution with solids <1.5mm).
Kiln Type Feed-point Max. Substitution Rate Solids Max. Substitution
Rate Liquids
Precalciner Kiln 1) Main Firing < 5 mm or foils < 50 mm: max. Up to complete heat
(with Tertiary 10-15% substitution input at main firing
Air) < 1.5 mm: max. 25% (= 40-60% substitution)
substitution
2) Precalciner < 50 mm: max. 30% Up to complete heat
Firing substitution input of calciner
(half of calciner heat input; (= 40-60% substitution)
example LD)
3) Kiln Inlet Lumps: max. 10% substitution Not reasonably
(example: whole tires at GS) applicable
Preheater Kiln 1) Main Firing < 5 mm or foils < 50 mm: Up to 100% substitution
(no Tertiary Air) max. 30% substitution (e.g. waste oil,
(example tire granulate at HD) solvents)
< 1.5 mm: max. 50%
substitution
(example: saw dust at RK)
2) Secondary Lumps: max. 10% substitution Max. 20-25%
Firing (example whole tires) substitution
(kiln inlet or < 50 mm: max. 20%
riser duct) substitution
(example: plastic chips)
Grate Preheater 1) Main Firing < 5 mm or foils < 50 mm: Up to 100% substitution
Kiln max. 30% substitution (e.g. waste oil,
(Lepol) < 1.5 mm: max. 50% solvents)
substitution
2) Secondary Lumps: max. 10% substitution Up to 20% substitution
Firing (e.g. whole tires):
(kiln inlet or < 50 mm: max. 20%
hot chamber) substitution
Long Wet or 1) Main Firing < 50 mm: max. 15% Up to 100% substitution
Dry Kiln substitution (e.g. waste oil,
(example tire chips at Holnam) solvents)
< 5 mm or foils < 50 mm:
max. 50% substitution
2) Mid Kiln Lumps (whole tires or bales): Not applicable
Firing max. 20% substitution
The maximum substitution rates are empirical and based on experience.

4. IMPACT OF LGF ON CLINKER BURNING PROCESS
4.1 General
The use of LGF always has certain impacts on the process, whether it is immediately
visible or not.
If LGF is used at very low quantities it may happen that the effects are not measurable
within the given accuracy or the effect disappears within the normal fluctuations. It
would be incorrect to extrapolate such observations from small to large quantities or to
large substitution rates.
Within this study most of the practical effects can be given in quantitative form and the
preferred reference to express this impact is always the direct factor causing this
impacts. E.g. the impact of injected water is expressed per ton of injected water. This
gives a precise and fair consideration of the impacts, even if only small relative
quantities of water are injected that disappear in the noise level of a large kiln.
Why can AFs increase the thermal losses on a cement kiln system and thus create
induced losses? There is a defined number of reasons that contribute to such effects
as follows:
6 factors for increased heat consumption due to AF
Fluct.
1 Water 4 AF feed
Cold air
2 Ash 5 introduct.
Poor
3 comb. 6
Circul.
Phenom.
Consequence : More exhaust gas, less kiln capacity

Impact chains:
The various possible impacts and their logical impact chains can be displayed as
follows:
LGF Properties
Ballast Inhom Coarse Circ.El.

(water, (Cl,S)
ash)
Combus
tion
Plugging
Heat
Cons.
Availa
bility
T/d Prod
Prod costs ( ) OEE Clinker

Qual
The figure illustrates that the various impacts are inter-related. The impact chains start
from the LGF properties and finally all end-up at production costs, OEE and clinker
quality.
4.2 Basic thermal phenomena/ limitations

Most of the impacts can be quantified already on paper. E.g. for the impact of water
tests are not really necessary as this can be predicted. There are minor exceptions: If
combustion problems due to poor granulometry occur only a semi-theoretical
prediction is possible by using the empirical increase in oxygen content that is
necessary to maintain proper combustion.
Heat consumption
Behind the following results there is a model which simulates the particular
behavior of a cement kiln system (preheating, calcining, sintering and cooling).
The logical sequence of calculating impacts always starts with the heat
consumption. Other impacts can then be deducted logically. The basic heat
consumption impact formula is :
q= + w*4.6 +a*1.1 +pa*1.0 +q* O2*.018 + v*1.2
MJ kg kg Nm3 MJ, %O2 Nm3
q= add. w=water a=ash pa=Pair O2 =incr. v= increase

Heat (inject) Oxygen in % of Vmin
This formula refers to the logical specific parameters causing the impacts (as e.g.
1 kg of water) and if useful - can also be used as per kg of fuel or kg of clinker.
From this formula most of the impacts (gas quantity, power consumption, capacity
reduction) can be deducted logically.

Flame temperature
The flame temperature is critical in the main firing. Below a certain minimum flame
temperature it is not possible to burn clinker (at acceptable freelime). However it is
not possible to express a flame temperature in form of an absolutely real and
measurable flame temperature. Flame temperature considerations make only
sense as theoretical flame temperatures calculated within the same model. The
model used here considers the following main factors :
Adiabatic flame temperature
Preheated secondary air coming from a standard grate cooler
Secondary air heat depending on heat consumption
Oxygen at kiln inlet always 2%
Standard fuels = fuel oil or coal (natural gas not standard)
Injection of solids into flame (inert or non inert) can be considered
The results are relative figures for the flame temperature (depending on the model).
To find the minimum possible flame temperature the model has to be calibrated with
practical experience. The practical reference is injection of inert material of max. 40%
into a wet kiln which produces a theoretical flame temperature of approx. 2100C. The
same temperature results if injection of 22% wet kiln dust (containing uncalcined
CaCO3) is assumed, which also corresponds to reality.
To get the maximum possible quantities of water and ash injection for any kiln type,
the same model is used to verify at what relative quantity the minimum flame
temperature is reduced down to 2100C.
Finally a general hint regarding flame temperature: The net CV of any fuel or LGF
cannot be used to characterize its impact on flame temperature. The generally correct
consideration is to start from flame impact factors (injection of water, ash, CKD)
expressed per kg clinker.
4.3 Impacts on heat consumption

By definition the heat consumption of a kiln is based on net calorific value (net CV)
of the fuels in use, including LGFs, of course. LGFs usually increase the heat
consumption compared to conventional fuels.
The effects of water, ash, false air introduction and increased O2 are displayed in
this chapter.
These effects are valid for impacts that actually occur in the high temperature range
(>800C) of the kiln.

Influence of water injection on heat
consumption
5 4.6 4.6 4.6
4.5
4
GJ/t water
3.5
3
2.5 2.15 2.15 2.15
2
1.5
Water injection
1
0.5 If water incl. in fuel
0
Dry SP Grate PH Wet
If wat er in f uel -> 2.45 M J/ t deduct ion
already included in NET CV Kiln Type
Water injection can produce strong impacts. A distinction is necessary on whether

water is part of the fuel (LGF) or injected separately.
If the water is included in the fuel and a correct deduction for this water
(deduction = 2.45 GJ/t water) has been made from the net calorific value, the
increase in heat consumption is 2.15 GJ/t water.
If the pure water is injected directly into the kiln, the increase in heat
consumption is 4.6 GJ/t water.
(Note : An alternative and physically logical approach would be to allocate a
negative net CV of -2.45 GJ/t to injected water and consider this for the kiln
heat consumption. By doing so there would be only one common factor of 2.15
GJ/t for any water/fuel mixture. However we recommend to treat water injection
separately because nobody in the cement world includes pure water injection
in the calculated kiln heat consumption).
Important: For the impact in terms of gas volume increase, the effect is the same for
both cases.

Influence of ash on heat consumption
( inert cold ash into high temp. zone )
1.2 1.1 1.1 1.1
0.8
GJ/ t ash
0.6
0.4
0.2
0
Dry SP Grate PH Wet
Kiln Type
Influence of false air on heat consumption
1.6
1.4 1.4
1.4
GJ/1000 Nm3 FA
1.2
1 1 1
1
0.8 0.7
0.6
Standard case (e.g. PA)
0.4
0.2 If FA cannot be used for
combustion
0
Dry SP Grate PH Wet
Kiln Type
False air (FA) due to AFR occurs mainly in form of additional transport air (TA) input
and leakage air at feed chutes for lump fuels.

Influence of O2 increase on heat consumption
% increase heat cons. per % O2 2.0% 1.8% 1.8% 1.8%
1.5%
1.0%
0.5%
0.0%
Dry SP Grate PH Wet
Kiln Type
Other than in case of water, ash and false air impact, the results are given here as %
increase per 1 % O2 increase. In this form the results do not depend on the kiln type.
The above rule is applicable for O2 increase in most of the practical cases such as:
O2 after precalciner or kiln inlet O2 of any kilns with 100% of the fuel at the main firing
(sintering zone).
Exception to this rule: If the kiln inlet gas is followed by a precalciner or secondary
firing (so that a normal O2 level of say 2% can be maintained after such a
subsequent firing) the impact of the kiln inlet O2 on heat consumption is 0% increase.
However this does not mean that such an O2 increase has no impacts, but they do
not fall into the category heat consumption. These other categories are: Loss of
flame temperature
(-80/-90C per 1% O2), higher temperatures and higher gas velocities and dust
generation at kiln inlet.
Increase of O2 after firing is e.g. required in case of combustion problems or if the

LGF energy input (feed or CV) is fluctuating (short term fluctuations). In case of
fluctuations an additional rule of thumb applies :
A fluctuation (short term [minutes]) of +/- 6% in total heat input (all fuels)
requires + 1 % O2 after combustion (which causes 1.8% more heat
consumption)
Consequence: Limit fuel fluctuations to the minimum.
4.4 Impacts on exhaust gas temperature

Only the water effect is displayed here (often the strongest effect):

Influence of water injection on exhaust gas temp. on
SP kiln
80
70
60
Temp. increase
50
4 stages
(C)
40
5 stages
30
20
10
0
0 0.05 0.1 0.15
Water injection (t w /t cli)
4.5 Impacts on flame temperature

The following diagram shows the impacts on the theoretical flame temperature
when introducing either water or ash or cement kiln dust (CKD) into the main
flame. For CKD the effect is stronger compared to ash because calcination
produces additional cooling.
Theoretical Flame Temp in Function of liquid water, ash, CKD
injected into flame, at 2%O2
Dry Kiln 3.3 MJ/kg 100% MF
2500 PC Kiln 3.3 MJ/kg 40%MF
2450
Wet kiln 6 MJ/kg
2400
Minimum flame temp 2100 C
2350
2300
2250
2200
2150
2100
2050
2000
Within the model used for the calculation the practical limit means a minimum flame
temperature of 2100C. So when one of the above bars comes down to 2100C this
means that the particular impact has reached the possible maximum.
If the O2 deviates from the standard 2% O2, each % more O2 reduces the flame
temperature by -80C (-90C on PC kiln).
The following diagrams (for water & ash influence only) are based on the same
figures as above.

Water influence on th. flame temp
2 % O2, oil/coal flame, 15% PA, standard grate cooler
2500
if gas firing : - 130 C +1% O2 : - 80 C
2450
2400 Dry SP 100% MF
Dry PC 40% MF
F.Temp. [C]
2350
Wet
2300
2250
2200
2150
2100 Min = 2100 C
0 0.02 0.04 0.06 0.08 0.1 0.12
Water into flame [kg/kg cli]
Ash influence on th. flame temp.

2 % O2, oil/coal flame, 15% PA, standard grate cooler
2500
if gas firing : - 130 C +1% O2 : - 80 C
2450
Dry SP 100% MF
2400
Dry PC 40% MF
F.Temp. [C]
2350 Wet
2300
2250
2200
2150
2100 Min = 2100 C
0 0.1 0.2 0.3 0.4
Inert ash into flame [kg/kg cli]
Results : practical impact limits according to the diagrams are :

For a suspension preheater kiln (with 100% main firing) :
Max 0.11 kg H2O/kg cli or max 0.36 kg ash/kg cli
For a precalciner kiln :
Max 0.042 kg H2O/kg cli or max 0.14 kg ash/kg cli
For a wet kiln :
Max 0.12 kg H2O/kg cli or max 0.40 kg ash/kg cli or max 0.22 kg CKD/kg cli.
Above limits consider only the flame temperature and in some cases they can be
overruled by stronger criteria. E.g. it has to be verified whether the ash to be

considered is also compatible to the clinker chemistry and can actually be absorbed in
the sintering zone. So the effective limit may be lower than the flame temperature
limit.
The basic impact limits apply for coal and oil firing as main fuel. In case of a gas
flame, flame temperatures are 130C lower and thus all above limits become
considerably lower.
4.6 Impacts on power consumption
Influence of water injection on electrical power

consumption
40.0 32.8
kWh/t water
30.0
20.0
10.0 5.0 5.0
0.0
Dry Grate Wet
Kiln type
The effect as indicated is at the ID fan for constant clinker production, so that the
preheater pressure drop increases above normal. Not applicable if the pressure drop
is kept constant.
Cost-wise the KWh effect gives only a minor contribution.
4.7 Impacts on production rate

The following graphs show the impacts of water, ash, false air and increased O2
on production capacity. The calculations are based on constant exhaust gas
volume and the effects are deducted from the previously shown impacts on heat
consumption (chapter 4.3).
The % increase in heat consumption and the additional exhaust gas produced are
often roughly proportional. However such simple rules of thumb are not accurate
and can fail completely as in case of the water impact. The basis for the following
results are complete calculations, of course.

Influence of water injection on kiln capacity
2.50
Prod. loss in t cli/t water

1.98
2.00
1.65
1.50
1.00 0.83
0.50
0.00
Dry SP Grate PH Wet
Kiln type
This applies for directly injected water and water included in the fuel. Incidentally in
case of water impacts the increase in heat consumption (chap. 4.3) alone would
be misleading for estimating the production loss.
Influence of ash on kiln capacity

due to thermal effect of cold ash
0.30
0.26
Prod. loss in t cli/t ash
0.25 0.22
0.20
0.15
0.11
0.10
0.05
0.00
Dry SP Grate PH Wet
Kiln type
The above ash effect considers thermal effects only (cold and inert ash into high
temperature zone).

Influence of false air on kiln capacity
Standard case
0.25 0.24
Prod. loss in t cli /1000 Nm3 FA
0.20
0.20
0.15
0.10
0.07
0.05
0.00
Dry SP Grate PH Wet
Kiln type
The above influence refers to additional false air (or primary air) into the high
temperature zone that occurs because of LGF introduction. It is therefore obvious that
above figures are not applicable for other false air occurring after the high
temperature zone (e.g. in the upper part of a cyclone-preheater or in the exhaust gas
system).
Influence of O2 increase on kiln capacity
6.0% 5.7%
5.0%
% prod. loss per % O2
5.0% 4.4%
4.0%
3.0%
2.0%
1.0%
0.0%
Dry SP Grate PH Wet
Kiln type
The same comments and restrictions as already mentioned under chapter 4.3 (O2
increase) apply. That means the above rule is applicable for O2 increase at kiln inlet
of any kilns with 100% fuel to the main firing or for O2 increase after a precalciner.
Exception: If the kiln inlet gas is followed by a precalciner or secondary firing and if no
increase of the final O2 level after such a subsequent firing spot is required the impact
of the kiln inlet O2 on heat consumption is 0% increase. In such a case any O2
increase at the kiln inlet has no impact on kiln capacity.
O2 increases are usually necessary in case of coarse solid fuels that create
combustion problems. A theoretical prediction is not possible. For a given type of LGF
an empirical correlation of O2 versus substitution rate can be evaluated. But even
then it still depends on the practical operating priorities (e.g. O2 can be kept constant

at the disadvantage of more combustion problems or more emissions). Production
capacity can be very sensitive on O2 increases.
4.8 Impacts on availability

Practical experience shows that the use of LGF can have a major impact on kiln
availability. These impacts can not be quantified precisely because they vary a lot
from case to case. The most common reasons for these kiln stops are blockages
and refractory damages.
The burning of lump LGF always creates local reducing conditions in the kiln
(sintering zone as well as kiln inlet). This leads to an increased volatilization of
sulfur which increases build-up formation, blockages and thus can reduce
availability.
On many kilns which already burn LGF, these problems are the major
limiting factor for the use of LGF.
These impacts can largely be compensated by process optimisation (see chapter
6.1-6.4).
Examples:
In Lgerdorf the availability was reduced to 75-85%. One of the reasons was
the use of LGF. By an optimisation of the kiln process the negative impacts
were compensated and the availability went up to >95% [14].
At HCB in Switzerland an increased market demand forced the plants to
maximise their production rate and availability. The goal was not to reduce the
use of LGF. Through an intensive process optimisation program called HCB
Working Group PRO the OEE was increased by 10% and at the same time
the use of LGF was increased [15].
4.9 Impacts on clinker quality

Ash effects
If ash is introduced into the kiln by LGF in significant quantities relative to
clinker, an adaptation of the raw mix must be made to meet the LSF (also SR
and AR) target in the clinker. Usually the LSF in the raw meal must be
increased to compensate the low CaO content of the LGF ash. If the main
component of the raw mix has a very high CaCO3 content (as the chalk in
Obourg or Lgerdorf) the kiln has a bigger potential to use LGF with high ash
content. If the main component of the raw mix is low in CaCO3, the use of LGF
with high ash is limited (example Untervaz: High grade lime stone has to be
purchased to compensate the ash from dried sewage sludge).
See also chapter 6.5 for LGF ash compensation in case of sudden LGF feed
failures.
Effects of locally reducing conditions
If lump fuels (e.g. tire chips) are injected into the burning zone, local reducing
conditions can not be avoided. This leads to an increased volatilization of
sulfur and can cause brown clinker cores (brown color inside the clinker
granules).
In Lgerdorf, during periods with high amounts of brown clinker core, a
reduction of the early strength has been reported. Additionally the brown color
can have an undesirable influence on the cement color. The brown clinker
cores were eliminated through the process optimization campaign (see
chapter 4.8).

In Obourg the early strength has a strong correlation with the SO3 content of
the clinker. An increase of the LGF substitution rate results in a reduced
incorporation of sulfur into the clinker. This effect still has to be solved in
Obourg, possibly by a burner optimization to reduce local reducing conditions
caused by coarse LGF.

5. CRITERIA FOR LGF PROPERTIES (IN VIEW OF CEMENT KILN)
5.1 Combustion properties (granulometry)

The required granulometry of solids depends on:
Kiln type
Feed point
Substitution rate
See table Maximum substitution rates in chapter 3.
5.2 Minimum flame temperature (water and ash)

The following values are maximum possible inputs of water and ash which still allow to
have a hot enough flame to produce clinker.
The maximum inputs of water and ash cannot be cumulated. If the maximum input of
water is introduced into the sintering zone, no ash can be introduced anymore
otherwise the flame is not hot enough anymore to produce clinker.
For high ash inputs a correction of the raw mix is necessary. Depending on the ash
chemistry it may happen that the chemical limit is reached before the flame
temperature limit is reached.
The limits are independent from the type of LGF.
Maximum possible input of water and ash:
Kiln Type: Maximum input:

Suspension preheater kiln Max. input H2O = 0.11 kg H2O/ kg cli or
(no tertiary air)
Max. input ash = 0.36 kg ash/ kg cli (if no other limit is
reached first)
Precalciner kiln Max. input H2O = 0.04 kg H2O/ kg cli or
(with tertiary air)
Max. input ash = 0.14 kg ash/ kg cli
Grate preheater kiln Max. input H2O = 0.11 kg H2O/ kg cli or
Max. input ash = approx. 0.05 kg ash/ kg cli (due to
poor ash incorporation)
Wet kiln Max. input H2O = 0.12 kg H2O/ kg cli or
Max. input ash = 0.40 kg ash/ kg cli
If these maximum inputs are used on a kiln, significant impacts on the kiln
performance (heat consumption and production rate) are the consequence (see
chapter 4). Also note that the values do not take into account the limitations of the kiln
system (e.g. limited ID fan capacity).

5.3 Homogeneity
The required homogeneity of the LGF depends on the following factors:
1. Fluctuation of the heat value in the LGF (short term [minutes])
2. Fluctuation of the mass flow of the LGF into the kiln (given by the quality of the
LGF dosing system; short term [minutes])
3. Substitution rate
The combined influence of the first two factors produces the fluctuation of the heat
input into the kiln by the LGF. The fluctuation of the total heat input into the kiln needs
to be below a certain level.
Rule of thumb for the required homogeneity of a LGF:
% fluctuation (heat input) * % substitution <

100%
% fluctuation (heat input) = fluctuation heat value + fluctuation mass-flow
If the fluctuations are higher, impacts caused by the inhomogeneity have to be

expected (e.g. CO formation, increased heat consumption, reduced production
capacity, higher sulfur cycles ...). See also chapter 4 (impacts).
In chapter 6.6 measures are described how to improve the homogenization of LGF
(e.g. by mixing of liquids, separate dosing for different solid LGF types or pre-
treatment centers).
Example:
100% Substitution: < 1% fluctuation of heat input to avoid negative impacts
Practical example:
Impregnated saw dust with a heat value of 11-13 MJ/kg = +/- 7% fluctuation of HV.
The dosing system delivers a mass flow between 2.0 and 2.1 t/h with a set-point of 2
t/h
= +/- 3% fluctuation of mass-flow.
Total fluctuation of heat input by the impregnated saw dust = 10%
Maximum 10% substitution without impacts on kiln operation caused by the
inhomogeneity

5.4 Circulating elements: chlorine, sulfur and alkalis
Chlorine Limit (for SP)

< 200 - 300 g Cl per t cli
200 Fuels Materials
Kiln system
Chlorine Clinker
Chlorine:
Chlorine is by far the most important circulating element with regard to LGF utilization.
The maximum possible input of chlorine from LGF depends on the following factors:
1. Kiln type
2. Input of chlorine by the LGF (concentration and mass-flow)
3. Input of chlorine by the raw material and other fuels
To define the maximum acceptable chlorine concentration for a certain type of LGF,
the actual situation with regard to total Cl-input into the kiln must be assessed first
(balance of circulating elements).
The following limits are expressed as total input of chlorine based on clinker [mg
Cl/kg cli]:
Kiln type: Maximum chlorine input (all sources):

Cyclone preheater kiln Total Cl input < 200-300 mg Cl/kg cli
without bypass
(with or without precalciner)
Cyclone preheater kiln +1% bypass rate +100 mg Cl/kg cli *
with bypass
If the bypass dust is used in the cement, the
maximum Cl limit of the cement
(usually Cl < 0.1%) has to be respected.
Grate preheater kilns (Lepol) < 200-300 mg Cl/kg cli
< 600 mg Cl/kg cli if preheater chamber dust is being
extracted
Long wet and dry kilns With 100% reintroduction of CKD:
Total Cl input < 200-300 mg Cl/kg cli
With extraction of CKD more input of Cl is
possible. If the CKD is used in the cement, the
maximum Cl limit of the cement
(usually Cl < 0.1%) has to be respected.
If all the CKD is discarded, a maximum Cl input of

up to 5000 mg Cl/kg cli has been reported
* On a suspension preheater kiln, a bypass at kiln inlet allows an additional Cl input of
100 mg Cl/kg cli for each percent of bypass rate (example: with 8% bypass, 800 mg
Cl/kg cli can be withdrawn from the kiln).
In chapter 6.3, 6.4 and 6.7 measures are described for enhanced control of circulating
elements. In chapter 6.8 different bypass systems are described and the costs of a
conventional bypass are given.
If bypass dust extraction or CKD extraction is used to handle excessive inputs of
circulating elements by LGF, the use of the dust must be clarified (the best solution is
to mix the dust into the cement). Never should this dust be landfilled.
Sulfur and Alkali:

The sulfur and alkali limits are usually not restricting the use of LGF. These limits are
more important for fuels such as petcoke.
As a rule of thumb the following 2 criterias can be used to calculate the maximum
possible sulfur input (both criteria have to be fulfilled):
1. Alk / SO3 ratio (corrected with chloride)
> 0.8
2. Total SO3 input < 1.5%SO3 in clinker
See also reference [9] Circulating element for the calculation of the Alk/SO3 ratio.
5.5 Environmental: Heavy metals and PCB

The following values give a short overview of practical and typical limits that are
actually used.
Only the most important and critical elements are listed below:
Heavy metals:
To define local heavy metal limits, kiln specific heavy metal balances can be made to
establish more fact based heavy metal limits.
Mercury (Hg): Almost all the mercury input is emitted. Therefore the Hg
concentrations in LGF have to be strictly limited.
Typical limit for AFR: 2-20 ppm *
Thallium (Tl): It is enriched in the outer dust cycle. The concentration in the
direct operation dust (dry SP kiln) can finally become so high (if all the dust is
returned to the kiln) that EP trips are a environmental and health hazard.
Therefore the Tl concentrations in LGF have to be strictly limited.
Chromium (Cr): The Cr concentration in the cement should be as low as
possible to prevent health problems by hexavalent chromium when handling
wet / fresh concrete. Therefore the Cr concentrations in LGF have to be strictly
limited.
PCBs:
Especially in liquids, the PCB content must be limited to prevent the abuse of the
cement kiln as a hidden, cheap disposal possibility for this problematic substance.
Typical limit for AFR (without permit to burn PCB): 50 ppm *
In special cases AFR with higher PCB content are burned (special permit based on
test campaign needed; safety measures for handling needed).
* Manual: AFR Quality Control Schema [13]

5.6 Clinker quality
The following elements, which can be introduced into the kiln by LGFs in significant
concentrations, have an effect on the clinker quality:
Phosphorus (P2O5): Maximum total input < 1%P2O5 in clinker
Higher concentrations can cause a delayed setting time. Phosphorus sources
can be sewage sludge or animal meal (P2O5 can also combine with CaO
which reduces the C3S).
Fluorine (F): Maximum total input < 0.2%F in clinker
Higher inputs need an adaptation of the raw mix (Fluorine is a mineralizer).
Zinc (Zn): Maximum total input < 1000 mgZn/kgCli
Higher inputs can cause a shorter setting time and a higher water demand
(even if the Zinc is added to the cement via kiln dust).
Lead (Pb): Maximum total input < approx. 100-200 mgPb/kgCli
Higher inputs can cause a shorter setting time and a higher water demand.
Sulfur (S, SO3):
High sulfur inputs as well as high sulfur cycles lead to a finer (dusty) clinker.
This has consequences on kiln operation and grindability. A reduced SO3
content in the clinker can lead to a lower earlier strength. This reduction can
be a consequence of combustion problems related to the use of LGF (long
flame, local reducing conditions, ..).
5.7 CKD (Cement Kiln Dust) quality (if CKD is discarded)

If the CKD is discarded, the enrichment of certain elements (e.g. chlorine and heavy
metals) creates a potential long term liability problem of the disposed material
(leaching).
Therefore the input of critical elements has to be limited.
5.8 Health and safety

Combustible dust:
Explosion and fire hazards with regard to combustible dust have to be respected
when storage and handling systems for (dusty) LGF are designed.
In Untervaz for example, a dedusting system for the installation of plastics in a hall
had to be added on later.
Fine dust:
Fine dust is a health hazard if the fine particles are inhaled. Additionally, harmful
substances in the dust can pose a health hazard.
For example the vanadium content in Flexicoke (not a LGF but a possible fuel for the
cement kiln) can be harmful and therefore the inhalation of that dust has to be
prevented.
Flash-point of liquids:
If the flash-point of liquid fuels is below 55C, the storage and handling equipment
must be designed according to special safety standards to prevent explosions and
fires (self-ignition).
Flammable and poisonous vapors (of solids):

If a LGF emits flammable or poison vapors, the installations for handling and storage
have to be designed in a way that no vapors can accumulate in dangerous (fire,
explosions and health hazards) concentrations.
For example in Obourg, the impregnated saw dust is stored outside and the transport
belts are ventilated into the kiln. This example can still be improved.
Odor:
If a LGF has a strong odor, the installations for storage and handling have to be
ventilated.
5.9 Handling properties

Stickiness:
Solid LGF have to be extractable from flat bottom bunkers with activated floor (best
solution for difficult solids). If the LGF is so sticky that a continuous extraction is
impossible, the fuel can be mixed for example with saw dust, kiln dust, raw meal or
other materials to improve handling properties (example: Preparation of impregnated
saw dust in pretreatment platforms).
Tire chip Spiders:
If tire chips are not properly cut (more ripped apart than cut), pieces of wires stick out
of each chip, which makes a continuous extraction from a feed bin (and handling in
general) almost impossible. The only solution is that the chips are cut properly, thus
the shedder knifes have to be sharpened or replaced.
Transport volume:
LGF with a density of approx. less than 50 kg/m3 need to be compacted for transport
(e.g. bales of plastics), because of the too high transport volume (high transportation
costs per ton). Palletizing is expensive (approx. 25 USD/t) and therefore usually not
feasible.
For tires, shredding reduces the transport volume significantly.
Health and safety as well as environmental aspects (see also chapters 5.8 and
5.5):
In the cement plant skin contact of workers with the LGF can never be ruled out. The
LGF properties must allow that without health hazards.

6. ADAPTATIONS / MODIFICATIONS TO COMPENSATE OR ALLEVIATE THE
IMPACTS OF LGF
6.1 Assessment of actual situation: Process indicators

The list of process indicators serves as a check-list to assess how well the burning
process has been optimized. The first step of any modification or adaptation should be
the assessment of the actual situation of the process.
One of the main reasons which limits the use of LGF in practical applications is an
insufficient or not optimized process control.
In the following list of process indicators, the indicators are listed and described
(Subject), the method of evaluation (Procedure) and the recommended limits
(Target Range) are given:
Subject Indicator Procedure Target Range

1) Kiln Feed LSF Sampling campaign: Fluctuation of
Homogeneity Hourly spot samples LSF < +/- 2
during 3 day-shifts of 8
hours within one week
2) Coal and Petcoke Residue 90m sieve Sampling campaign: Fluctuation:
Fineness and Hourly spot samples, R90m < +/- 1%
residue 200m sieve during 3 day shifts of 8 Set-point coal:
hours within one week. R90m < of %
volatiles
R200m < 2%
Set-point petcoke:
R90m < 5%
R200m < 1%
3) Degree of CaOf and (liter weight) Sampling campaign: CaOf = 0.8 1.5%
Burning of Hourly spot samples (upper limit according
Clinker simultaneous to the to plant specific
kiln feed samples. requirements)
4) Enrichment of Cyclone preheater Sampling campaign: Cyclone preheater

Circulating kiln: During one week once kiln:
Elements / %SO3 and %Cl in hot per shift one spot SO3 < 2.5% and
Tendency of meal sample of the hot meal Cl < 1.0% in hot meal
Build-up or CKD
Formation
Long wet and dry kiln: Long wet and dry kiln:
%SO3 and %Cl in CKD The limits are roughly
5 times higher.
Interpretation of the
results together with
Alk/SO3-balance.

Subject Indicator Procedure Target Range
5) Sulfur / Alkali / Total inputs and Calculation of balance see chapter 5.4
Chloride Balance outputs of circulating (inputs and outputs) of
elements circulating elements
Molar alkali-sulfur ratio (SO3, K2O, Na2O, Cl)
from typical values of
concentrations [%] and
feed rates [t/h] *
6) Combustion: %O2, %CO at kiln inlet Measuring of gas CO < 0.1%
Composition of composition at least O2: sufficient (e.g. 3-
Exhaust Gas at every half hour during 4% for calciner kiln,
Kiln Inlet 3 day-shifts of 8 hours approx. 2% for long
within one week. wet and dry kilns as
well as preheater kilns
without calciner)
7) For Calciner Set-point and Print out of Temperature
Kilns: fluctuation of temperature during 3 fluctuations < +/- 10C
Control of calcination day shifts of 8 hours Set point = 90% - max.
Calcination temperature (bottom within one week 95% calcination
Degree cyclone, or hot meal degree
temperature)
Correlation of Sampling campaign: Temperature and
calcination Hourly spot samples of calcination degree
temperature and hot meal during 3 day must show clear
calcination degree shifts of 8 hours within correlation (higher
(Calculated from LOI one week. Recording temperature higher
of hot meal) of calcination calcination)
temperature at the
time of sample taking.
* see reference [9]
6.2 Enhanced process control

The use of LGF often increases instabilities during kiln operation. These fluctuations
can be compensated or at least alleviated by enhanced process control.
Procedure:
1. Definition of key parameters for kiln operation with clear set-points, tolerable
range and countermeasures in case of deviation.
These definitions are specific to each kiln and have to be done in close
cooperation with the kiln operators.
Set-point, range and countermeasures have to be fixed in a written form and
be displayed in the control room. Operators have to be trained to use them.
2. Deviations of the set-points and tolerable ranges have to be analyzed
(PARETO) and the causes have to be eliminated. Typical reasons of such
deviations are bad measurements (e.g. temperatures and gas concentrations),
defective equipment (e.g. tertiary air damper) and insufficient dosing of fuel
(noble fuel and LGF).
3. As a very advanced and effective measure, the installation of a high level
control system such as LINKman can be done if the process is stable

(LINKman is well suited to handle kilns burning LGF; example Untervaz and
Altkirch).

Example of enhanced process control / process optimization at Lgerdorf:
Essential Set-point and Control / Improvements /

parameter range countermeasures measures realized
O2 kiln inlet > 3% Tertiary air damper Replacement of
and ID-fan defective tertiary air
damper;
relocation and new
operation strategy
of kiln inlet probe
Temperature after 850 890C Calciner fuel rate Improved dosing,
lowest cyclone handling and
stage homogeneity of
LGF
Free lime 0.8 1.2% Mandatory reaction
with fuel rate of
main burner if 2
values are below
0.8 (no over-
burning)
6.3 Enhanced control of circulating elements (SO3, K2O, Na2O and Cl)
The use of LGF often increases the internal circulation of volatile elements and thus
creates or worsens coating problems. The burning of lump LGF always creates local
reducing conditions in the kiln. This leads to a increased volatilization of sulfur which
increases build-up formation and blockages. This can have a major impact on kiln
performance (availability and production rate). On many kilns which already burn LGF,
these problems are the major limiting factor for the LGF substitution rate.
One of the most important points is the presence of sufficient oxygen (excess air) in
the kiln. The volatility of sulfur depends strongly on the oxygen content in the kiln
gases. Therefore the O2 content at kiln inlet must be kept in an optimum range (see
chapter 6.1). The locally created reducing conditions by burning lump LGF can thus
not be avoided. Their impacts can only be alleviated by other measures.
The first step to solve coating problems is an analysis of the actual situation with
regard to circulating elements.
Assessment of the coating problems:
1. Excess air in the kiln:
Measurement of O2 (and CO) at kiln inlet
Is enough O2 available in the kiln for complete combustion ?
2. Systematic analysis of the indicators:
Suspension preheater kiln: Indicator = SO3 and Cl in hot meal
Grate preheater (Lepol) kiln: Indicator = SO3 and Cl in chamber dust
Long wet and dry kilns: Indicator = SO3 and Cl in CKD
Comparison with the limits for circulating elements *
Is it a chloride problem, sulfur problem, alkali problem or a combination ?
3. Performance of a balance of circulating elements:
Where do the inputs and outputs come from ?

Comparison with typical, tolerable inputs of circulating elements in raw
materials and fuels *

4. Calculation of the molar alkali-sulfur-chloride ratio:
Comparison with typical values *
Are enough alkali available to bring out the sulfur with the clinker ?
5. Calculation of the sulfur volatility:
Comparison with typical values *
Why do we have a sulfur problem (too high inputs, unfavorable alkali-
sulfur ratio or combustion problem) ?
From the results of the analysis, the actions to solve the problems can be derivated. *
Counteractions have to the defined and taken if the indicators (see above) exceed the
limits. For chlorine this can be the reduction of fuel quantities with high Cl
concentrations or the increase of dust extraction (see chapter 6.7). For sulfur this can
be an increase of O2 at kiln inlet or the reduction of SO3-input by the fuel (e.g.
percentage of petcoke).
* Circulating Elements [9]
6.4 Enhanced preheater cleaning (cyclone preheater kilns only)

One of the most important measures to control the circulating elements (besides
respecting the input limits) is an optimized preheater cleaning.
The most important points are the following:
The preheater must be checked for coatings and cleaned at least once per
shift.
Manual cleaning must be intensive, short and periodical.
Clear measurable limits must be defined and when these limits are exceeded,
mandatory cleaning must be done.
Example Ternate: If the pressure difference between kiln inlet and calciner
exceeds the limit cleaning is done.
Example Lgerdorf: If the pressure at kiln inlet exceeds the limit cleaning must
be done.
Kiln inlet and riser duct: On every platform and in every corner openings have
to be available. In critical sections additional access openings have to be
provided.
In critical points where coating is always forming, air blasters should be
installed.
Riser ducts (from cyclones and at kiln inlet) should be protected against build-
ups by lowering the meal entry point to condense the volatile elements on the
cold meal. Correct splash box design is important to distribute the meal
evenly across the duct.
Endangered cyclones should have enlarged cyclone outlets (double cone)
and meal ducts with increased diameter to enable falling coating to pass
through.
The preheater should be absolutely tight. False air leaks increase the
formation of build-ups.
When coating formation is very strong, a high pressure water pump should be
used to help remove coatings.

Application of air blasters:
6.5 Quality control (LGF ash compensation)

If ash is introduced into the kiln by LGF in significant quantities relative to clinker, an
adaptation of the raw mix must be made to meet the LSF (also SR and AR) target in
the clinker. Usually the LSF in the raw meal must be increased to compensate the low
CaO content of the LGF ash.
High inputs of ash by LGF need important corrections of the raw mix. In this case the
input of LGF can not be interrupted on a short term basis because clinker quality will
otherwise be out of specification. Raw mix adaptations take several days to arrive in
the kiln because of the homogenizing silo capacity. Therefore the supply of the LGF
and the handling system (careful equipment design important!) must guarantee an
uninterrupted ash input. If this can not be guaranteed, a back-up system to
compensate the ash input must be installed. For example in Obourg such a back-up
system for fly ash dosing is installed.
6.6 Improved homogenization and dosing of LGF

For small substitution rates, the negative impacts of poor homogenization and dosing
may be not measurable within the given accuracy and normal fluctuations. But for
higher substitution rates (e.g. >10%) the dosing and homogenization of LGF is key!
The goal is to have a constant heat input into the kiln by the LGF. The heat input
depends on the fluctuation of the heat value (homogeneity) and on the quality of the
dosing system.
See chapter 5.3 for the tolerable limits of these fluctuations.
Homogeneity of liquids:

Mixing in the storage tank is mandatory (either mechanical or by circulation flow).
The tank must be big enough, or even better 2 medium size tanks should be installed
(example Obourg or Lgerdorf), to avoid unacceptable fluctuations during discharge
of new deliveries.
Homogeneity of solids:
The homogenization of solid LGF needs certainly more effort than for liquids, but has
the same or even more importance because of the more difficult dosing properties.
Mixing of different types of deliveries of LGF with a bridge crane does not work. With a
front loader in a big hall or open area a homogenization is possible.
A proper solution to burn 2 different LGF at the same time is to install 2 separate
dosing systems (example Lgerdorf).
An other proper approach is a pretreatment platform or company which can produce
from many different sources of LGF one new and homogeneous LGF (e.g.
impregnated saw dust).
Dosing of LGF (in particularly solids):
1. Suitable equipment:
The equipment for LGF dosing must be designed with the same importance as
the equipment for the noble fuels (e.g. coal dust).
For the important design criteria of the equipment see chapter 8.
(Example: Long pneumatic transport lines with bends and corners are just not
suitable for most LGF. In this case a complete redesign (installation of belt
conveyors) is the only solution.)
2. Continuous improvement and adaptation to changing LGF properties of these
handling systems is needed. Small modifications can make a big difference.
Remark: The intermittent feed of whole tires at kiln inlet is feasible only because of the
special combustion properties of tires (delayed combustion). It would be problematic
to intermittently feed LGF which burns faster (consequence: incomplete combustion -
high CO peaks).
6.7 Dust extraction

General:
Dust extraction is used to take out a certain fraction of the circulating elements, mainly
chloride (on long wet and long dry kilns also sulfur and alkali) as well as to prevent
high heavy metal accumulations (in particular Thallium). The effectiveness depends
on the kiln type (see below).
If the extraction is used to handle excessive inputs of circulating elements by LGF, the
use of the dust must be clarified. The best solution is to add the dust into the cement
mill. The tolerable amount of chlorine in the cement is 0.1% Cl. Normally this limit can
by far not be exploited due to fluctuating concentrations of Cl in the dust and due to
the fact that not all types of cement can be blended with dust. The plant in Hver for
example has a practical average potential of 0.05% Cl to be mixed into the cement.
If LGF causes inputs of circulating elements (usually chlorine) this creates impacts
on the process, depending on the relative quantity (expressed in g Cl per t of
clinker). To compensate for this input an output by using partial dust extraction
can be realized. The highest efficiency is obtained if the extraction takes place at a
point in the process where the highest enrichment of the volatile element occurs.
Extraction [g Cl / kg cli] = relative dust quantity [kg dust / kg cli] * concentration in
the dust
[g Cl / kg dust]

Cyclone preheater kiln:
The easy (low cost) extraction of normal filter dust as in case of long kilns does hardly
produce a substantial effect on a cyclone preheater kiln because the enrichments of
Cl in the outer dust cycle are very low. Moreover the gases are usually utilized in the
raw mill where kiln dust is heavily diluted with fresh raw meal. But still there is a
marginal possibility to improve the chlorine situation without major investment which is
worthwhile to be considered in almost any case.
Lets assume a real bypass installation cannot be justified but we want to slightly
increase the chlorine load. If total chlorine inputs exceed only slightly the tolerable
limits (reaching some 300- 350 g Cl / t cli) it is recommended to discard the filter dust
during direct operation and send these small dust quantities to cement grinding.
Incidentally this has also other benefits, it will also break the outer metal cycles
(whereof the Thallium is the most critical).
Experience from Lorca with dust extraction (2 stage preheater which is favorable for
taking out chlorine by dust extraction): [17].
Grate preheater kiln (LEPOL):
Due to the comparatively low dust quantities generated by the nodule feed system
high enrichments in the intermediate dedusting and the main dedusting equipment are
achieved. Significantly higher chlorine inputs (e.g. by LGF) can be tolerated if these
dusts are taken out. Normally they are being extracted and create an outlet for Cl.
Long dry and long wet kilns:
Circulating elements (sulfur, alkali and chlorine) can be removed comparatively easily
on a long wet or on a long dry kiln by discarding the fine filter dust. In particular the
dedusting with electrostatic precipitators offers the most favorable possibility to extract
a minimum quantity at highest enrichment. From experience we know that up to 0.5%
chlorine input on clinker (0.5% on clinker = 5000 gCl/tcli) basis can be handled and
extracted by simply discarding enriched filter dust.
6.8 Bypass technology
Conventional bypass:

A small bypass of e.g. 10% can much increase the flexibility to use chlorine containing
LGF and to alleviate problems with sulfur cycles. If a by-pass is installed, the use of
the bypass dust has to be clarified (see below and also chapter 5.4).
Chlorine extraction efficiency :
With a by-pass a higher chlorine input can be tolerated. The amount of
chlorine depends on the by-pass rate. Rule of thumb:
For each percent of by-pass rate, 100 mg Cl/kg cli more chlorine input can
be tolerated (without bypass the maximum chlorine input must be less than
200-300 mg Cl/kg cli). This rule of thumb is applicable up to a maximum of
15-20% by-pass rate.
Example:
10% BP : chlorine extraction approx. 1000 g Cl/t cli
100% BP : chlorine extraction approx. 5000 g Cl/t cli
Impacts due to bypass (on PC kiln):
Impact on heat consumption = 8 kJ/kg cli per % BP
Bypass dust = approx. 0.001 kg dust/kg cli per %BP
Environmentally sound utilization/disposal of bypass dust must be found
Impact on emissions (because of SO2 and HCl that bypass the efficient
absorption in the suspension preheater)
Investment
To add a 10% bypass means a considerable investment, typically in the
order of 2 3 Mio USD. So far the installation of a by-pass has not been
realized in the Holderbank Group only to increase the use of LGF with the
exception of Holly Hill.
Special consideration referred to chlorine
The bypass transforms chlorine contained in LGF into inorganic chlorine in bypass
dust. Chlorine promotes leachability of metals, which may be a concern when
landfilling this dust. Using again the above figures (to be consistent) each ton of
chlorine that must be extracted via bypass dust creates approx. 10 t of enriched
bypass dust and an additional heat consumption of 80 GJ/t of chlorine which allows
to allocate the costs to one ton of removed chlorine. Depending on the local cost
factors (for landfill and energy) this may result in costs in the order of 200 1000 USD
per ton of chlorine removed.
Therefore the pros and cons of a chlorine bypass installation due to LGF must be
evaluated. All these bypass impacts do of course not occur if the suspension
preheater kiln is only loaded below its natural chlorine bearing capacity (200 300 g
Cl/t cli).
Maury- Bypass:
This bypass was realized in Erwitte (Germany) [6]. After extraction and quenching, the
bypass gases are dedusted in a cyclone. The gas including the fine dust is
reintroduced into the preheater. A bypass filter is thus saved. Only the coarse bypass
dust can be extracted (from cyclone). The chlorine extraction was sufficient to use
50% RDF on a suspension preheater kiln [7].
Conclusion: Not recommended for standard cases (not as effective as normal
bypass).
Mid kiln bypass for long kilns (Cadence system):

The chlorine extraction efficiency compared to the normal dust discarding (via the
main EP fine dust) is questionable. The side effects are a creation of additional kiln
draft and lower velocities in chain zone which increases kiln capacity.
Conclusion: Not recommended for standard cases.
6.9 Increased gas handling capacity
Suspension preheater / pre-calciner kiln:

The first step before any modification is done is always the elimination of all false air
leaks.
On SP kilns the increase of gas handling capacity normally brings the kiln capacity up.
Areas to be considered :
ID fan:
Enlarge ID fan and/or motor
Effect :
Power consumption increase +0.15 kWh/ t cli per % volume increase (at
a SP kiln with 5 kWh/t cli consumed at the ID fan)
Absolute amount of false air increases but percentage of false air remains
constant (at constant leaking gap size)
If increased pressure reaches almost -100 mbar --> do not push further,
reduce pressure drop by preheater modification
Water injection into exhaust gas:
Water injection into the top cyclone stage (either before or after the cyclone)
is a relatively easy fix to boost the capacity of an existing system with
minimum investment. The injection into the downcomer needs very fine
droplets, if this is not achievable injection into top stage is not ideal but safer.
Cooling tower:
The volume of the cooling tower which can usually not be upgraded is critical
(retention time of the gas for evaporation of the water). In some cases it is
enough to upgrade the number and performance of the water injection
nozzles and the pumping capacity.
Filters:
With an electrostatic participator the filter efficiency is decreased with
increased gas volume. If the dust emissions are already close to the limit, the
filter has to be upgraded (e.g. installation of additional chambers).
With a bag filter a higher gas volume will lead to an increase of the cleaning
cycle frequency (reduced bag life and higher power consumption) and / or a
higher pressure loss across the filter (higher fan capacity needed and higher
power consumption).
Long wet or long dry kiln:

Increasing the exhaust gas capacity of a long wet kiln fan does not automatically
cause a proportional increase of the kiln capacity, which is an essential difference to a
SP kiln. Such proportional capacity increase would only work if the dust generation
effect (dust leaving the kiln) would be negligible. But actually there is an over-
proportional increase of dust generation (approx. plus 3 % dust per every % more gas
velocity). In many cases the dust must almost fully be returned to the kiln. So the kiln
capacity is limited rather by the dust generation effect than by the waste gas handling
capacity.

Depending on the dust generation and dust return possibility it is not always sufficient
to only increase the gas handling capacity. Also the dust generation must be reduced
simultaneously by appropriate measures (reduce velocity of kiln inlet zone, adapt
chain design for low dust generation).

6.10 Increased critical cross sections
Cyclone preheater and precalciner kilns:
Measures to draw more gas quantity at less pressure drop (consider shape of
ID fan curve):
Cyclone dimensions (increase thimbles when Di/Do <= 0.6)
Cyclone design with low pressure drop
(Riser ducts design)
Calciner orifice / TA damper design
Measures related to internal dust generation/cycles:
Kiln inlet chamber (modify narrowest section for < 21 m/s, for dust
reasons)
Long wet and long dry kiln:
The critical area of the kiln is the chain section. Higher gas velocities lead to more
dust generation. An optimized chain design can alleviate this increase.
At the kiln inlet section a shell diameter extension can be made to reduce dust
generation and thus bring the capacity up. This is a major modification, the feasibility
has to be evaluated from case to case. In particular kilns that have a constant
diameter over the whole kiln length are well suited to such a modification. Example:
Extension of inlet section at TR from 12 to 14 with total cost of 2.75 Mio USD (incl.
new kiln drive, chains and lining).
The achievable gain is strongly connected to the previously mentioned relations
between inlet zone velocity - dust generation - dust return - kiln capacity. On a wet kiln
with already high dust generation, the increase of gas handling capacity alone would
not help.
6.11 Burner adaptations

To compensate or alleviate the impacts of LGF on the burning process, burner
adaptations are a preferred measure to improve the combustion of LGF.
Even when the LGF is not burned in the main firing (but e.g. at kiln inlet, mid kiln or
precalciner) an optimized combustion in the sintering zone has a strong impact on the
kiln performance and thus on the substitution rate of the LGF.
The goal of the adaptations is to create a strong, hot and stable flame.
The optimized burner should operate with the following key figures:
Key Figures for Main Burner:
Primary air ratio [%PA]: 10 12%
based on stoichiometric combustion air (Amin);
without transport air
Specific axial momentum 7 10 N/MW (incl. fuel and transport air)
[Gax]: therefore primary air fan pressure must be at
least 200 mbar (better 250 mbar)
Injection velocity coal dust: 25-30 m/s
Additional important criteria:
Coal and petcoke fineness Residue on 90m and 200m sieve: See
[R90m] and [R200m]: chapter 6.1
Oil temperature and Oil temperature sufficient to reach a viscosity of
pressure (for heavy oil): 15 cSt; pressure according to atomizer supplier
specification

Correct atomization of liquid Atomization with pressurized air (or steam);
LGF: pressure of liquid and air (steam) as well as
amount of air (steam) according to atomizer
supplier specification
Theses rules apply also to kilns with no LGF, but in case of LGF an optimum burner
set-up is even more important to compensate the impacts of LGF.
To fulfill these figures, a new burner might be necessary or at least a new primary air
fan is needed in many cases.
Injection point of LGF through the main burner:
The injection of LGF can either be done through the center of the burner or through a
separate pipe in the burner refractory on top of the burner.
Experience from Origny show that the combustion of light weight material such as
plastic foils is better if injected through the center of the burner, whereas heavier
material such as impregnated saw dust or dried sewage sludge is better to inject
though a separate pipe on top of the burner (more details [10]).
For practical applications the injection through the center of the burner is in most
cases only possible if a new burner is purchased, whereas a separate pipe on top of
the burner can be added to any existing burner. Therefore, for the first tests the
injection on top is usually applied. When purchasing a new burner, in any case one or
more additional channels in the center should be foreseen for test or later application
of AFR.
Alignment of the burner inside the kiln:
The burner should be positioned parallel to the kiln axis and in the kiln center. To
shorten the flame the burner can be retracted towards the kiln outlet.
6.12 Cooler modifications

The goal of the cooler modifications is to create a high and stable secondary air
temperature which is favorable in view of LGF combustion.
The main requirements for a cooler are the following:
1. High heat recuperation efficiency:
Which means a high secondary and tertiary air temperature to improve the
combustion of the fuel, especially difficult to burn fuels like LGF.
2. Stable operation:
Kiln operation and cooler operation are closely linked together. An unstable
cooler operation will destabilize the whole kiln system.
An optimized cooler performance does help to alleviate the impacts of LGF.
Grate Cooler:
To optimize an existing cooler to improve the combustion of LGF the following
parameters should be assessed:
The total cooling air should be in the range of 1.8-2.3 Nm3/kg cli (low value
only for modern coolers).
The specific cooling air load per m2 of aerated cooler surface [Nm3/s,m2],
calculated for each chamber should show a descending pattern. The inlet load
should be 1-1.5 Nm3/s,m2 (max. 2 Nm3/s,m2 with fixed inlet).
The chamber pressure distribution should show a descending pattern (low
pressure in the back of the cooler, higher pressure towards the front).
Operation with as high bed height as possible, but still below the height where
instabilities occur (air breaking through). Therefore the installation of a view

port is needed to visually assess the clinker bed. Chamber pressure at the
start of the first moving grate should be in the range of 30-50 mbar.
The grate speed control loop should have a control range (e.g. 38-42 mbar
chamber pressure at beginning of first moving grate) in which it does not vary
the grate speed to avoid too much fluctuations of the cooler operation.
The kiln hood pressure must be kept continuously on set-point by the waste air
fan control loop.
Modifications:
Installation of a fixed inlet to improve clinker distribution and thus cooler and
kiln operation.
Installation of modern plates (and cooler fans with enough pressure) to
improve heat recuperation.
Planetary coolers:
The performance and thus optimization potential of a planetary cooler depends
strongly on the efficiency of the internal heat transfer equipment (lifters).
The second important point is an optimum elbow design (connection between kiln and
planetary tube) to create an even distribution of the clinker in all planetary tubes and
to avoid excessive dust back-spillage.
Travelling grate cooler (Recupol):
This cooler type has a limited potential for optimization.
The installation of a modern fixed inlet (possibly together with a short oscillating grate
section) and improved sealing of the grate sides as well as the discharge hoppers is
possible to improve recuperation.
6.13 Precalciner modifications

For an existing precalciner the following modifications can be made to improve the
combustion of LGF (listed according to feasibility in terms of costs):
Hot spot
Turbulence generators
Pre-combustion chamber
Retention time
Hot spot:
The idea of the hot spot is to create a (hot) zone in the calciner where the ignition and
combustion is facilitated by not yet introducing cold, not calcined meal. A hot spot
design is the repositioning (lifting up) of the meal entry point (meal from the second
lowest cyclone into the calciner) above the fuel entry point. Even more effective is the
splitting of the meal entry to the calciner, into an upper and lower entry point
(=controlled hot spot). The hot spot is then between the fuel entry point and the upper
meal entry point. A meal split is only possible if the second lowest cyclone is installed
high enough. For existing precalciners such a hot spot can be installed to improve the
combustion and thus substitution rate of LGF.
Turbulence generators:
All modern precalciners are designed with turbulence generators such as counter flow
introduction of tertiary air, KHD-Pyrotop, Polysius Deflection Chambers and others.
The turbulence does improve the burn out of the calciner fuel which is specially
important for difficult to burn fuels such as LGF. For existing, old precalciners such
modifications can be made to improve the combustion and thus substitution rate of
LGF.

Pre-combustion chamber:
In the pre-combustion chamber the fuel is ignited and partly burned in pure air (tertiary
air). This type of precalciner is designed for low reactive fuels such as LGF and
petcoke. A modification of an existing calciner (addition of pre-combustion chamber) is
costly but can be economically feasible especially in combination with a capacity
increase or a conversion to petcoke firing. Several projects have already been
realized.
Retention time:
The retention time of the gas in the calciner and thus the time for the fuel to burn out
depends on the calciner volume. For difficult to burn fuels such as coarse LGF, higher
retention times are favorable (4.5 seconds recommended). The modification of the
calciner volume is usually very costly.
General aspects on burning LGF with coarse granulometry in the precalciner:

Not in all calciner types, LGF with oversize pieces or coarse granulometry can be
burned. For example in a separate-line calciner these types of LGF can not be used.
In-line calciners and pre-combustion chambers of the downdraft type are suitable. See
also chapter 7.5.
6.14 Mixing air on long kilns (connected to mid kiln firing)

This technology is only applicable to long wet and long dry kilns and is connected with
the use of mid kiln firing.
Mixing air is a fan which is installed on the rotating kiln between the mid kiln firing
entry point and the beginning of the chains. This fan blows air with high pressure into
the kiln. Thus turbulence (mixing) is created inside the kiln to improve the burn out of
the mid kiln LGF.
Mixing air is installed on only a few kilns word wide. In Obourg it is installed since
1999 (2x10000 m3/h 12% primary air; 170 mbar). According to the plant the mixing
air improves the burn-out of the mid kiln fuel (compacted lump fuel bales). This leads
to a reduced chain entry temperature, which allows to increase the substitution rate at
mid kiln.
6.15 O2 enrichment
Conclusion on O2 enrichment:
The use of O2 enrichment is a high cost item. It can be justified if the costs can be
paid by the AFR business. The availability of cheap O2 sources (say < 50 USD/tO2)
would also favor such applications but so far this is not yet proven in our industry (The
Air Liquide company claims that the cost for O2 could be reduced up to some 50%
by applying non-cryogenic on-site oxygen supply units but we are not yet there).
Even without considering any LGF the use of O2 on a kiln boosts the capacity.
However every additional ton of clinker produced (1.1 t cli / t O2 for wet kiln and 2.54 t
cli / t O2 dry kiln) would cost so much that O2 enrichment can usually not be justified.
It is still better to design the system properly (e.g. adequate sizing of fans and cross
sections).
The use of O2 to merely boost kiln capacity is not recommendable regardless whether
LGF is being burnt or not.
If the main problem when using LGF is actually insufficient flame temperature it makes
sense to get the temperature back by using O2 enrichment. In particular when the
theoretical flame temperature would already drop to the minimum of 2100C virtually

most of the common methods (e.g. increase of cross sections and flame shape
adjustments) cant help any more.
General aspects of O2 enrichment:

The concept to enrich the combustion air with O2 is very old. The standard technical
realization is the injection of pure O2 via burner channel into the sintering zone. The
O2 source is usually liquid oxygen, however also alternative sources as O2 producing
units at the plant have been proposed (not realized) to decrease the O2 costs.
The use of O2 decreases the nitrogen load in the kiln gas that occurs when using
normal ambient air with 21% O2 content. The logical consequences are as follows :
1) Less combustion gases (Vmin)
2) Reduced exhaust gas losses (because of 1)
3) Less air from the cooler (lower recuperation)
4) Overall reduction of heat consumption (result from 2 and 3)
5) Increased flame temperature (because of 1)
6) Increased NOx (as a consequence of 5)
7) Decreased specific exhaust gas quantity
8) Increased potential production rate
Most of the effects of O2 enrichment can be predicted by model calculations. Results
are shown in the following graphs:
Influence of O2 enrichment on kiln capacity
3.00
2.54
2.50
2.11
2.00
t cli/ t O2
1.50
1.07
1.00
0.50
0.00
Dry Grate PH Wet

Influence of O2 enrichment on heat
consumption
0.00
-0.20 Dry SP Grate PH Wet
-0.40
GJ / t O2
-0.60
-0.80
-1.00
-1.20
-1.40
-1.38 -1.38 -1.38
-1.60
Theoretical Flame Temp, incl also O2 injection
0.02 kg O2
Dry Kiln 3.3 MJ/kg 100% MF
0.05 kg O2
PC Kiln 3.3 MJ/kg 40%MF
2800
Wet kiln 6 MJ/kg
2700
0.05 kg O2
Minimum flame temp 2100 C
No impact
No impact
No impact
No impact
2600
Theoret. Flame temp (C)
No impact
No impact
2500
0.05kg H2O
0.05kg H2O
2400
0.10kg H2O
0.10kg H2O
0.042kg H2O
2300
0.12kg H2O
0.11kg H2O
0.40kg ash
0.14kg ash
0.36kg ash
2200
2100
2000
The flame temperature model used is consistent to the previous chapter 4.5. The
theoretical flame temperature is only a relative indication that must be compared to
temperatures calculated within the same model.
The following points can be seen (or deducted) from the graphs :
1 ton of injected O2 boosts the kiln capacity by a gain of 1.1 2.5 additional
tons of clinker. The lowest gain would be applicable for the wet kiln.
For a constant O2 addition (kg O2/kg cli) the strongest flame temperature
effect occurs at the main firing of precalciner kilns (due to the 40% firing rate).
An addition of only 0.02 kg O2/kg cli produces 280C more flame
temperature.
Assuming 100% firing rate to the main firing, the O2 effect on a suspension
preheater kiln (without precalciner) is stronger than on a wet kiln. E.g. at 0.05
kg O2/kg cli the SP kiln flame temperature increases by 250C (wet kiln =
140C).

The O2 can be used to compensate LGF effects. Assuming a SP-kiln with
100% main firing the impact of 1 ton injected water can be compensated with
approx. 1 ton of O2 in order to get the same flame temperature.
The Holly Hill case:

The so far longest and most extensive experience is available from Holly Hill. The
basic idea behind the O2 enrichment was to compensate the negative impacts of
the liquid AF (Synfuel or HWDF = hazardous waste derived fuel) used in their wet
kilns.
Main data kiln 2 (long wet kiln).
Normal capacity (coal) 2180 mt/d
HWDF CV 23 MJ/kg and less
HWDF substitution rate 45 43%
Production loss by HWDF 320 mt/d (- 14.5%)
O2 injection 60 93 mt/d
O2 injection volume rate 1750 2710 Nm3/h
O2 enrichment from 20.9 % to 23.5 Vol %
Production gain by O2 180 - 230 mt/d
Other effects :
Holly Hill is in a special situation regarding emissions compliance (THC). In order
to keep THC < 20 ppm they have to run the kiln at an abnormally high O2 of 4
5.5% which reduces the flame temperature almost to the lowest limit. So the O2
brings the flame temperature back, keeps gas velocities low and reduces the dust
generation.
In addition the O2 enrichment is also used to compensate the effect of CKD
injection into the flame.
Cost factors :
CAPEX approx. 220000 USD
Monthly fixed costs (leases) approx. 7000 USD
Variable cost 65 USD per ton of liquid O2
Cost of purchased clinker 52 USD/t (locally used for benefit calculation)
According to the general rules the gain on a wet kiln should be approx. 1.1 t clinker
per ton of O2. However Holly Hill reported a rate of approx. 2.5 t cli/t O2 which is
surprisingly high and must be related also to other reasons, such as the unusual
O2/THC situation. The plant is convinced that the use of O2 pays out, the strongest
impact being the production gain that compensates the production loss by HWDF.
The costs for O2 can be shared with the supplier of the HWDF (Safety Kleen).
6.16 Gasifier / Thermal Cracking

Instead of feeding solid LGF directly to the kiln, they can be fed indirectly by
integrating a gasifier into the kiln process. Within the considerations made here we will
concentrate on kiln process integrated gasification / thermolysis. The hot combustible
gases are directly sent to the kiln. In such cases the gasifier can also be considered
as a pre-combustion chamber working at under-stoichiometric combustion.
The general advantages are
Lump fuels do not need to be ground to small particle size
Better control of fuel input because the gasification rate can be controlled
better (at least better than feeding lump fuel directly to the kiln).

Better stability of burning process
No reduction of theoretical flame temperature and no increase of heat
consumption in case the pre-combustion involves no cold air but hot air from
the cooler.
To some extent ash or solid residues can be kept away from the burning
process if necessary (except from fine dust).
Exceptionally undesirable chlorine can be kept away from the burning process
General disadvantages are :
High costs (compared to direct feeding to the kiln)
Safety issue (hot explosive gases that may ignite when adding air)
On unexpected stops of the kiln firing the gasification process cannot be
stopped immediately and hot combustible materials remain in the gasifier.
Therefore a flare or an extinguishing system is needed.
External gasifiers (not process integrated):
External gasifiers produce from waste different fractions: Combustible gas, liquid
fuel (oil) and solid fuel (coal). They have significant disadvantages compared to
process integrated gasifiers when they are combined with a cement kiln. The
occurrence of three products (gas, liquid and solid residues) is just one of the
numerous drawbacks. External gasifiers are not recommended and not
considered further in this study. [18]
Conclusion on Gasifiers:
The gasification / thermal cracking process has a few applications on cement kilns
that show that it is technically feasible. The thermal efficiency of the gasification
process can potentially get close to 100% if it is integrated in the kiln system as a
pre-combustion chamber using tertiary air from the cooler.
Long term and large scale operating experience is available from the CFB gasifier.
This system works but has not met all expectations because it cannot dampen the
fluctuations of the AFR feed and the originally expected use of tire chips was not
feasible. Other systems seem to be also promising but are less known from
practical experience.
Gasifiers do not dispense from installing advanced LGF handling and dosing
equipment.
The LGF applications involving gasification are comparatively expensive and can
only be justified if disposal fees are available. To some extent impacts of LGF on
the kiln can be alleviated but not always be avoided. Therefore we would not
recommend the gasification as a standard solution to use lump LGF. Only if all
other possibilities to use LFG by direct combustion are unsuccessful gasifiers can
be used in a high cost environment.
Available solutions for cement kilns:
1) LURGI circulating fluidized bed (CFB) gasifier

Holderbank
Within Holderbank the use of a CFB was studied extensively for gasifying
waste wood at RK [4] and alternative fuels at UV [5], but it was not realized.
Main reasons for not using it :
costs more than direct feeding to kiln
technical risks

Rdersdorf
A CFB was realized in Rdersdorf/ Germany [2]:
The Rdersdorf kiln is designed for 6000 t/d. The circulating fluidized bed
gasifier is connected to the precalciner and is designed for 100 MW. It can
supply max. 2/3 of the calciner fuel.
According to the latest info the CFB is operated with 100% pre-processed
municipal waste and fluff (< 50 mm). In contrast to original expectations the
use of tire chips was not feasible in practice. Investment costs for the gasifier
amount to 35 Mio DEM in 1995 (approx. 18 Mio USD) and might possibly be
lower for the next installation.
The experience shows that the fluctuations (of mass-flow and CV) of AFR
input cannot be dampened by the CFB. Fluctuating fuel input directly
influences the precalciner.
Because cold air is introduced at the aerated gasifier bottom the same effects
must occur as for cold air introduction due to AFR. The evidence is that
Rdersdorf experienced a 40C higher exhaust gas temperature.
2) NESA at Rochefort
Multiple hearth (NESA):

Bridge crane
Product 1 Conveyor
Rotary kiln
Product 2
Screen
Pyrolysis furnace
Gas
Mixer burners
Screw conveyor
Storage silo
Combustion fan Lift

Cooling screw elevator
Cooler To raw meal mill
Solid waste with organic content (up to 25%) that can be used as alternative
raw materials (AR) after thermal treatment is gasified with this installation. It is
designed for 20000- 30000 t/a of solid waste. The gases from the pyrolisis
(net CV of 0 6.5 MJ/Nm3) are sent to the main firing of the Lepol kiln. The
system is designed for 6000 Nm3/h combustible gases at 350-450C which
are introduced into the main firing with a separate burner tube parallel to the
main burner.
Investment is in the order of 7.5 Mio USD.
At the moment the installation is still in the start-up phase. 1 t/h feed rate has
been achieved so far.

3) Polysius gasifier
Installed at Jura Cement, Wildegg, Switzerland:
The installation is designed for 3 t/h whole tires. Wood is also possible with
some modifications. The pre-combustion chamber is fed with tertiary air. The
extraction of steel wires at the bottom is done towards the kiln inlet.
Approximately 1/3 of the combustible matter is extracted together with the
steel wires in the form of soot and reaches the kiln inlet. About 2% water
referred to clinker is injected into the gasifier for temperature control. Total
investment is approx. 5 Mio DEM (some 2.5 Mio USD) (excluding tire
feeding). Report on the installation: [8].

4) FLS HOTDISC
The HOTDISC developed by FLS was tested in pilot scale and the first
industrial installation is expected to come in operation in 2000 in northern
Norway.
The system can be considered as a pre-combustion chamber for lump fuels
which is closely linked to the precalciner. The lumps are retained on a slowly
rotating disc that is exposed to the tertiary air. Temperature control is by
diverting meal on the disc.

5) Onoda REG gasifier
The gasifier has been realized in Japan long ago and published 1987 [11]. It
was studied 1992 for application in the Olten plant [12] (with contacts to
Onoda) and was recommended as favored solution. It has not been realized.
In contradiction to early publications, only tire chips can be used (no whole
tires). No recent publicity or activities are known.

6.17 Drying of LGF
To eliminate the impact of the water introduced into the kiln by LGF, the drying of LGF
before burning could theoretically be a possibility.
One example is sewage sludge. After mechanical de-watering normal sludges are
solid but contain still as much as > 60% water. Drying of such materials is
problematic, mainly because of strong odors. Technically it cannot be realized like
normal fuel drying (e.g. using hot waste gases that get in contact with the sludge
would create a big emission problem).
In Switzerland an extensive study has been done how to use sewage sludge in the
cement kiln. Conclusion: Drying of sewage sludge is not feasible in the cement plant.
Dying is done today by the producer of the sludge and the dried sewage sludge is
burned in the cement kiln. [16]
In Obourg an installation (BEMTI) has been realized to eliminate odor (thermally)
from drying and grinding waste coal sludge mixed with sewage sludge. Investment:
Total 570 Mio BEF (some 13 Mio USD); thermal gas treatment alone 300 Mio BEF
[15].
Another example is impregnated saw dust having 30% water. For similar reasons as
mentioned above it is not feasible to dry impregnated saw dust.

7. CRITERIA FOR A NEW CEMENT KILN IN VIEW OF FUTURE USE OF LGF
7.1 Summary of requirements for a new kiln using AFR
Requirements for New Kiln using AFR

Kiln system Main requirements for AFR Remarks
part
Preheater 10 % gas volume-flow reserve compared to
conventional fuel
Exhaust gas Same as above (larger ID fans and filters)
system
Calciner Gas retention time min 4 - 4.5 sec, in-line type, Downdraft type pre-
controlled hot spot must be possible combustion chamber preferred
for low reactive fuels
Bypass Decision and bypass size depending on chlorine Make sure bypass dust does
input by AFR (and general process factors) not need to be landfilled
(especially at high bypass rate)
Rotary kiln L/D min 15, +10% cross section compared to
standard design for conventional fuels
Kiln burner High momentum, min. one spare channel
(inside)
Cooler Modern grate cooler High secondary air
temperature favorable
7.2 General
Within the Holderbank Group basically all new kilns must be designed for future use of
LGF and petcoke i.e. for high fuel flexibility.
In cases where the AFR use is already clearly defined in the project phase it is
obvious that the kiln system design must be chosen accordingly. However in certain
projects for new kilns it was decided that at the project stage only the presently
available AFR (or almost no AFR) were taken into consideration without looking into
the future flexibility for other AFR.
Under the influence of a tight budget and the stripped down concept there was a
tendency to sacrifice the future fuel flexibility for cost reasons. Unfortunately a later
change for AFR is not always possible by simple modifications or add-ons. Certain
items are just not correctable or would involve high costs and downtime.
The following criteria refer to a standard kiln system with precalciner and grate cooler.
Capacity of kilns is usually 3000 t/d 6000 t/d (majority of new kilns). The criteria are
generic and can also become overruled by additional criteria (if clearly defined
requirements for LGF usage or other specific process requirements exist).

7.3 Criteria for preheater and exhaust gas system design
The preheater design shall take into account +10% reserve volume-flow compared to
standard fuel design (pure oil or coal firing). The recommended pressure drop over
the preheater (including +10% more volume) shall be 50 mbar.
Kiln ID fan, conditioning tower, raw mill gas system (as far as applicable), filter and
fans must be designed accordingly.
7.4 Criteria for bypass and related items

A small bypass of 10% can much increase the flexibility to use chlorine containing
LGF and alleviate problems with sulfur cycles, but a bypass installation must be
decided upon from case to case.
If a bypass is needed just to handle the possibly excessive chlorine input caused by
LGF the use of a bypass dust must be clarified. There are different solutions, the
easiest being the addition to the cement. At moderate chlorine inputs to the kiln (<
0.1% Cl referred to clinker that would require a 10% bypass) the dust admixture to the
cement can be feasible, depending also on the local standards. If landfilling of bypass
dust cannot be avoided the use of a bypass due to LGF creates a negative
environmental impact which is not acceptable.
If no bypass is foreseen the kiln system shall be designed to extract the filter dust
during direct operation and send it to the cement grinding section. This will give a
slight extraction for chlorine and break high metal accumulations (in particular
Thallium) in the outer dust cycle.
7.5 Criteria for calciner

The calciner shall be designed for a gas retention time of 4 - 4.5 sec to offer better
burnout for LFG.
Calciner types where large LGF particles cannot be accepted should be excluded.
The problem is that certain calciners cannot handle lumps or heavy particles because
they fall down into the tertiary air duct and create blockages. These are the separate
line calciners and the calciners with tertiary air leg where solids can accumulate at
the lowest point. Such calciner designs are not acceptable any more in view of solid
AFR. Acceptable are in-line calciners where oversize large particles can fall down to
the kiln inlet. In addition the generation of a hot spot by an adjustable meal split (high
and low position) shall be possible. Very favourable are also calciners having
precombustion chambers of the downdraft type (RSP, Polysius new type CC, Fuller
downdraft) which are especially suited to ignite and burn 100% low volatile fuels as
petcoke.
7.6 Criteria for rotary kiln

In view of solid AFR with coarse granulometry fed to the main firing or lump fuel fed to
the kiln inlet the rotary kiln must be designed to serve also as a combustion chamber
for solid AFR. The aim is to get a high gas retention time to enhance combustion and
to achieve comparatively low temperatures at the kiln inlet to keep sulfur cycles (will
be enhanced by poor combustion) as much as possible in the rotary part rather than
in the preheater. A kiln with a L/D ratio of min 15 (on 3 supports) is recommended. For
the kiln diameter a 10% addition in cross section (compared to the kiln size required
for conventional fuels) is recommended.

The following diagram shows the necessary kiln diameter increase (compared to
standard diameter) needed due to AFR.
Kiln diameter: L/D min. 15

6.00
5.80
5.60
Kiln diameter IS
5.40
5.20
(m)
5.00
4.80
AFR kiln
4.60
4.40 Standard kiln (no
4.20 AFR)
4.00
2000 3000 4000 5000 6000 7000 8000
Kiln capacity t/d
7.7 Criteria for main burner

Today a burner with high primary air ratio (min. 12% based on stoichiometric
combustion air, excluding transport air) and high specific momentum (min. 7 N/MW) is
standard anyway for process reasons. In addition, provision of at least one additional
jacket/channel for solid AFR in the burner center (needs a larger central section than
the standard design) shall be provided.
In contrast to other key equipment parts (as rotary kiln or calciner) the later
replacement of a kiln burner pipe because of particular AFR requirements is usually
still affordable.
7.8 Criteria for grate cooler

A grate cooler of modern standard design will usually be sufficient and no extra
requirements are needed. Such a cooler will usually offer a high and stable air
temperature (about 1000C) which is favorable in view of flame cooling effects due to
AFR. The standard arrangement is a tertiary air take off from the kiln hood. Although
tertiary air extraction from the cooler roof would produce even higher secondary air
temperatures this is not recommendable because of protection of the nose ring
section.

8. SET OF RULES HOW TO HANDLE LGF
(STORAGE, DOSING AND INJECTION INTO KILN)
Set of Rules / Important Points:

SOLIDS
Storage in the plant The storage should be covered. Dust and harmful gases have
to be considered (dedusting or ventilation needed).
Extraction from the storage hall and filling of the dosing hopper
with bridge crane or front loader.
Dosing system Feed hopper with activated flat bottom and vertical walls (e.g.
Saxlund moving floor) in connection with belt scale. *
Transport to the kiln Only regular belt conveyors should be used (no pneumatic
transport except on burner platform because of inevitable
blockages with most LGF).
Secondary firing and Triple flap gate
precalciner:
Feeding into the
precalciner or
preheater riser duct
Main Firing: For the main firing a separation of oversize pieces is
Separation of mandatory. Recommended system: Rotating screen.
oversize pieces Magnetic separator (cross belt) is recommended to protect the
rotary valve before the pneumatic transport.
Main Firing: Short and straight pneumatic line starting at the burner
Injection at the main platform.
burner Injection through burner center or on top of the burner.
LIQUIDS
Storage tank: Continuous mixing is mandatory (mechanical agitator or
circulating flow).
Pump: The screw pole type pump is preferred (not sensible to
particles, continuous flow and high pressure).
Special solutions are needed for solvents (e.g. membrane
pump) and sludges (e.g. piston pump).
Two pumps should be installed in parallel (one stand-by).
Atomizer Atomization must be done with pressurized air (or steam).
* Examples for flexible dosing systems of solids:
Recommended system for the main firing: Dosing hopper from Obourg and
Eclpens (Saxlund moving floor with screw extraction, belt scale).
Recommended system for secondary firing or precalciner: Dosing hopper from
Lgerdorf (Saxlund moving floor, belt conveyor with stripper drum, belt scale)
For detailed recommendations on AFR equipment: see reference [10]
Examples of handling systems for solids and liquids:

Precalciner or Secondary Firing
Belt Conveyors
Storage Hall
Double or
Triple Flap
Bridge Crane Gate
Belt Scale
Feed Hopper
with Activated
Flat Bottom Main Firing
Rotary Separator
Rotary Feeder
Belt Conveyors
Burner
Mechanical Mixing
Flow and Density

Measurement
Discharge
Filter
Burner
Pressurized Air

9. INTERACTION LGF PREPARATION AND BURNING
9.1 Cost relations to prepare LGF
Shredding Cost vs. AFR Granulometry

180
Tyres, Coarse Shredding
160
Tyres RUMAL
Owning & Operating Cost [USD/t]

PNEUREC 140
HEMING 120
PLASTREC
Waste BIOREC
100
Wood REKINGEN 80
60
Industrial
Plastic 40
Waste 20
0
110 100 90 80 70 60 50 40 30 20 10 0
Max. Particle Size [mm]
Estimation of Shredding Cost for Tires, Wood and Plastic

Although only rough estimates of shredding costs are possible, the following
conclusions can be made:
The cost for shredding increases exponentially with the fineness of the AFR
product.
Shredding of tires is more expensive than for industrial plastics and waste
wood. Although not shown, plastics with mineral or metal contamination are
shredded at higher costs than clean industrial plastic waste.
Coarse tire, plastic or wood chips for SP/PC firing (0 50 mm) can be
produced at overall costs of less than about USD 65.-/t. Fine granulation for
main burner firing (< 10 mm) costs more than USD 80.-/t.
Plastic shredding: see also [20]

9.2 Substitution rate versus quality of LGF: Optimization of costs
Economics of using tires as a function of size reduction
Basic data/assumptions
Precalciner kiln 3000 t/d Operating hou 6000 h/a
Heat consumption (convent. fuel) 3.3 MJ/kg cli Amortisation 7 years
Specific fuel cost (conventional) 3 USD/GJ Interest rate 5%
Inline calciner with 60% fuel
CV of whole tire 28 MJ/kg
Shredded Shredded Granulate

Whole tires tires to kiln tires to to PC and
Item at kiln inlet inlet precalciner kiln
1 Size specification (max particle size) mm max 1200 200 50 8
2 Density (at specified size) kg/m3 200 500 500 700
3 Possible fuel substitution % heat 10% 12% 20% 40%
4 Efficiency of substitution % 80% 85% 90% 95%
5 Tire consumption (capacity) t/d 35 42 71 141
6 Tire consumption (eff) t/a 8 839 10 607 17 679 35 357 40
7 Equipment for handling and feeding Mio USD 1.5 1.5 1.5 1.5 28
22
8 Specific investment cost USD/tpy 170 141 85 42 20
9 Cost for size reduction USD/t tire 0 20 50 160 6
0
10 Financing costs USD/t tire 29 24 15 7
10000 1000 100 10 1
11 Operating cost USD/t tire 10 5 5 5 -20
12 Total costs (items 9,10,11) USD/t tire 39 49 70 172
max particle size (mm)
13 Fuel saving effect referred to 1 t of tire* USD/t tire 67 71 76 80 -40
Gain (item 13 minus 12), excluding
14 USD/t tire 28 22 6 -93 -60
any disposal fees etc
15 Increase in heat consumption * % 2.0% 1.8% 2.0% 2.0% -80
16 Loss of kiln capacity % 2.0% 1.8% 2.0% 2.0% -93
17 Loss of kiln capacity t/d 60 54 60 60 -100
18 Loss of kiln capacity t cli/ t tire 1.7 1.3 0.8 0.4
19 Relative transport costs (at spec size) C/Co 2.0 1.2 1.2 1.0
* considering item 4 (efficiency of substitution)
Note : Additions or deductions to the gain (item 14) depend much on the local project conditions and are left open here. Adding a disposal fee
obviously has a high impact on the feasibility. A possible penalty for lost clinker depends on whether maximum production is actually needed
or not. Transport cost depend on the local logistics and the product size/density.

Annexes:
1) Example of impact calculation for solid LGF
2) Summary of LGF Impacts
3) References
CTS-Process Technology
A. Obrist D. Pauling A. Flacher

Example of impact calculation for a solid LGF
(showing impacts of water, ash, transport air, increased O2 level)
Data Given:
Basic LGF data Basic kiln data
Water content % 30% Kiln type SP kiln (100 % main firing)
Ash content % 15% Heat consumption GJ/ t cli 3.3
CV net GJ/ t LGF 15.0 LGF feedpoint Main Firing
Transport air Nm3/kg LGF 0.7 for pneumatic injection Substitution rate % 10%
O2 increase % O2 0.5 to burn LGF in kiln
Results:
Ratio LGF/clinker t LGF/ t cli 0.0220
Impact on heat consumption Impact on flame temp. Impact on prod.
(for this type of kiln)
Red.
Actual Critical Flame
referred to LGF referred to clinker quantity quantity temp. Loss of production
% of kiln % of kiln
Influence factor unit GJ/t LGF % of CV MJ/ t cli heat cons kg/kg cli kg/kg cli C t cli/ t LGF capacity
Water % 30% 0.65 4.3% 14.2 0.43% 0.007 0.110 -24 -0.59 -1.31%
Ash % 15% 0.17 1.1% 3.6 0.11% 0.003 0.360 -4 -0.04 -0.09%
Transport air Nm3/kg LGF 0.7 0.70 4.7% 15.4 0.47% - - - -0.17 -0.37%
O2 increase % O2 0.5 1.35 9.0% 29.7 0.90% - - -40 -1.29 -2.85%
Total 2.86 19.1% 62.9 1.91% -68 -2.09 -4.62%

=primary results based directly on factors from LGF study chapter 4 Critical reduction = -400C Fl. Temp
Conclusions:
- 19.1% of the CV of this LGF is lost due to increased kiln heat consumption
- Kiln heat consumption is increased by 1.91%
- Kiln production is reduced by 2.09 ton clinker per each ton of this LGF (or by 4.62% for 10% substitution)
- Reduction of flame temperature is not critical

Annex 1b
Detailed calculation of results from Annex 1 example:
Ratio LGF/clinker: 3.3 GJ/t cli / 15 GJ/t LGF * 10% = 0.022 t LGF/t cli
Water
Impact on heat consumption:

GJ increase/t LGF: Factor from chapter 4.3: +2.15 GJ/t H2O
2.15 GJ/t H2O * 0.3 t H2O/t LGF = 0.65 GJ increase/t LGF
% increase referred to CV of LGF: 0.65 GJ/t LGF / 15 GJ/t LGF = 4.3%
MJ increase/t clinker: 0.022 t LGF/t cli * 650 MJ/t LGF = 14.2 MJ increase/t cli
% increase of kiln heat consumption: 14.2 MJ increase/t cli / 3300 MJ/t cli = 0.43%
increase
Impact on flame temperature:
kg H20/kg clinker: 0.022 kg LGF/kg clinker * 0.3 kg H20/kg LGF = 0.007 kg H2O/kg
clinker
Critical quantity from chapter 4.3: 0.11 kg H2O/kg clinker resp. 400C
Reduction flame temperature: 0.007 kg H2O/kg cli / 0.11 kg H2O/kg cli * 400C = -24C
Impact on production:
t cli lost/t LGF: Factor from chapter 4.3: -1.98 t cli lost/t H2O
0.3 t H2O/t LGF * -1.98 t cli lost/t H2O = -0.59 t cli lost/t LGF
% lost of kiln capacity: 0.022 t LGF/t cli * -0.59 t cli/t LGF = -1.31% lost
Ash and transport air

calculation same as for water
O2 increase (excess air)
Impact on heat consumption:

% increase of kiln heat consumption: Factor from chapter 4.3: +1.8% increase per 1%
O2
0.5 * 1.8% = 0.9% increase of heat consumption for 0.5% O2 increase
MJ increase/t clinker: 3300 MJ increase/t cli * 0.9% = 29.7 MJ increase/t cli
GJ increase/t LGF: 0.0297 GJ increase/t cli / 0.022 t LGF/t cli = 1.35 GJ increase/t LGF
% increase referred to CV of LGF: 1.35 GJ/t LGF / 15 GJ/t LGF = 9%
Impact on flame temperature:
Factor from chapter 4.3: -80C decrease per 1% O2 increase
Reduction flame temperature: 0.5 * -80C = -40C decrease by 0.5% increase of O2
Impact on production:
% lost of kiln capacity: Factor from chapter 4.3: -5.7% production lost/ 1% O2 increase
0.5 * -5.7% production lost = -2.85% production lost

t cli lost/t LGF: 0.0285 t cli lost/t cli produced / 0.022 t LGF/t cli prod = -1.29 t cli lost/t
LGF

Annex 2
Summary of LGF impacts
Maximum possible factor set

impact on heat by flame cooling in main
Influence factor consumption impact on production rate firing *
Kiln type all kilns Kiln type Dry SP Grate Wet Kiln type SP PC Wet
Unit Unit Unit
H20 GJ/t H20 2.15 t cli/t H2O -1.98 -1.65 -0.83 t H2O/t cli 0.11 0.042 0.12
Ash (inert) GJ/t ash 1.1 t cli/t ash -0.26 -0.22 -0.11 t ash/ t cli 0.36 0.14 0.4
False Air GJ/1000 NM3 1 (0.7) t cli/1000 Nm3 FA -0.24 -0.2 -0.07 N/A N/A N/A
O2 conc. % heat per 1 %O2 1.8 % cap. per % O2 -5.7 -5.0 -4.4 % O2 increase 5 5 3
*applicable for coal/oil as main fuel

Annex 3: References
Ref Author Title Source Remark

1 Obrist A. Cement Course, Use of PT 00/14674/E
Alternative Fuels
2 Flacher A., Bauer Rdersdorfer PT 00/30/E
C., Pauling D. Zementwerke. Visit to
Rdersdorf Plant,
Fluidised Bed Gasifier
4 Obrist A. Vergasung von Holz VA 89/5565/D Waste wood

zwecks Befeuerung eines Rekingen,
Zementofens survey
gasification
5 Vorprojekt PA 92/10847/D Alternative
Sekundrbrennstoffe im Fuels via CFB
Zementofen at UV
6 Maury H.D.(MBM) Une installation pour Ciments Bton Use of RDF at
l'utilisation direct des Pltres Chaux Erwitte, using
dchets mnagers en fuor Jan/Feb 2000, p. Maury bypass
de cimenterie / A plant for 433-444
direct use of domestic
waste in a cement kiln
7 Maury H.D.(MBM) Chlor-Bypasss zur Zement-Kalk-Gips Chlorine
Erhhung des Nr 11/1988 bypass at
Brennstoffeinsatzes aus Erwitte to burn
Mll beim Klinkerbrennen more RDF
8 M. Suter Polysius tire gasifier at TPT 00/21025/E
Wildegg
9 D. Brassel, U. Circulation Phenomena PT 99/14503/E
Gasser
10 D. Pauling Recommended AFR TPT 01/21054/E
Equipment
11 I. Saito, K. Sakae, Effective use of waste World Cement Onoda gasifier
T. Ogiri tires by gasification in Sept. 1987
cement plant
12 A. Obrist Verbrennen von Altreifen VA 90/5758/D Proposal
hauptbrennerseitig im Onoda gasifier
Zementofen, Studie fr for tires chips
den Fall at Olten
Portlandcementwerk
Olten
13 R. Stenger Manual: AFR Quality HMC / CIE
Control Scheme
14 Dr. H. Rosemann, Process Optimization and Presentation Group
R. Hasler AFR at Lgerdorf Meeting 1999
15 B. de Quervain Kiln Optimization in View Presentation Vesta
of AFR Forum IV
16 A. Obrist, Th. Lang Nicht-landwirtschaftliche VA 88/5503/D
Verwertungsmglichkeiten

Ref Author Title Source Remark
von Klrschlamm unter
besonderer
Bercksichtigung der
Verbrennung im
Zementofen; Teilstudie 1
Schlammbehandlungsver-
fahren
17 C. Daidone Chlorine Balance and Vesta Forum IV,
Optimization at Lorca April 2000, Mexico
Plant
18 L.P. Seirlehner Opportunity Study HMC / CIE (to be finished
(SCS-Technology) Thermal Cracking approx.
10/2000)
19 R. Spilliaert Presentation BEMTI Presentation for BEMTI in
C.E.T.I.C. Obourg
Technique Plenire
5.6.1998
20 A. Flacher Technical file: Plastic HMC / MPT / AFL
Shredding
21 CIE/Euremi Technical files on AFR AFR Pyramid
preparation and handling (Holderspace)

PREPARATION AND HANDLING OF ALTERNATIVE FUELS
A. Flacher
MPT 03/14882/E
Table of CONTENT
1. From Waste to Alternative fuel..................................................................................... 485

1.1 Waste Properties............................................................................................... 485
1.2 Processing of Waste ......................................................................................... 486
2. Preparation and Handling of Liquid Alternative Fuel................................................. 487
2.1 Filter .................................................................................................................. 488
2.2 Tank .................................................................................................................. 489
2.3 Pump................................................................................................................. 489
2.3.1 Eccentric Screw Pole Pump .......................................................................... 489
2.3.2 Membrane Pump ........................................................................................... 490
2.3.3 Centrifugal Pump........................................................................................... 490
2.4 Flow Meter ........................................................................................................ 491
2.5 Injection Nozzle................................................................................................. 491
2.6 Safety Equipment.............................................................................................. 491
3. Preparation and handling of pasty alternative fuel .................................................... 492
3.1 Direct Sludge Pumping to Kiln .......................................................................... 492
3.2 High Viscosity Liquid Facility............................................................................. 493
3.3 Impregnation Facility ......................................................................................... 493
3.3.1 Reception and Preparation of Primary Materials........................................... 493
3.3.2 Production of Impregnated Alternative Fuel and Dispatch ............................ 494
4. Preparaton and Handling of Solid alternative Fuel .................................................... 495
4.1 Shredding and Granulation of Solid Waste ....................................................... 495
4.1.1 Shredders ...................................................................................................... 495
4.1.2 Granulators.................................................................................................... 496
4.1.3 Multi-Stage Size Reduction ........................................................................... 496
4.2 Handling of Solid Alternative Fuel at the Cement Plant .................................... 497
4.2.1 Storage .......................................................................................................... 498
4.2.2 Extraction & Dosing System .......................................................................... 498
4.2.3 Kiln Feeding................................................................................................... 499

5. Examples........................................................................................................................ 501
5.1 Typical Liquid Waste Handling.......................................................................... 501
5.2 SCORIBEL Impregnated Sawdust Production.................................................. 502
5.3 Industrial Plastic Waste Shredding at PLASTREC AG ..................................... 504
5.4 The whole Tyre Feeding System at Siggenthal Plant ....................................... 505
5.5 Plastic Chip Feeding System at Untervaz Plant................................................ 505
5.6 Feeding System for Coarse Shredded Waste (Tyre & Plastic Chips) at Altkirch506
5.7 Impregnated Sawdust Feeding System at Eclpens Plant ............................... 506
5.8 Used Oil Facility at Lgerdorf Plant................................................................... 507

SUMMARY
Unlike most standard bulk materials, alternative fuel is a waste product. It often doesn't have
suitable physical properties for the kiln process. This is where alternative fuel preparation is
needed.
Waste markets and alternative fuel properties may change quickly. A corresponding kiln
feeding system must therefore be of a flexible design to adequately handle a wide range of
products.
This paper focuses on common waste treatment techniques for alternative fuel production
and handling. For a complete understanding of the subject, the lecture of the Cement
Course paper Use of Alternative Fuels is recommended.
1. FROM WASTE TO ALTERNATIVE FUEL
1.1 Waste Properties

When waste material is considered for alternative fuel use, first of all, its chemical
composition (circulating elements, ash composition) must be suited for the kiln process.
Then there are other properties and aspects, which need to be looked at:
? The viscosity of liquids determines storage and handling techniques as well as
possible kiln feeding options.
? According to the granulometry of solids, a size reduction may have to be considered
to fit the fuel to the desired kiln feeding point.
? Water contained in liquids requires mixing for better homogeneity of the alternative
fuel. Water contained in solid waste has an impact on the handling properties.
? Foreign bodies and impurities need to be separated in order to reduce wear on
installations and risk of blockages.
? Safety is very important when dealing with waste, e.g. toxicity and flashpoint need
to be considered.
? Additional aspects refer to the waste supply, such as annual tonnage as well as
form and schedule of delivery.

1.2 Processing of Waste
Same as for any other process, for the preparation of alternative fuel the most simple and
inexpensive but also safe process suits best. A mechanical waste treatment is therefore
chosen in almost any case.
The preparation of alternative fuel comprises the following main processes:
? Screening to separate foreign and oversized material, which might cause handling or
wear problems in fuel preparation and firing installations.
? Size reduction to fit the fuel to the foreseen firing point (e.g. preheater, main burner),
as well as to allow a proper dosing and to ensure a good combustion.
? Homogenizing and mixing to feed to the kiln process a fuel of constant quality.
Picture 1 gives an overview of the different processes from waste to alternative fuel.
Explanation is given in the following chapters.
From Waste to Alternative Fuel

WASTE Alt. Fuel
Liquid Decanting, Liquid
Filtration, Mixing
Screening, Mixing
Sludge Sludge
with Liquid
Screening
Solid Shredding, Impregnation,

Screening Solid
Shredding Gas
Gasification
Picture 1: Processes from waste to alternative fuel
Thermal waste processing (gasification/pyrolysis) is applied in a few cement plants. It is

operated in conjunction with the kiln. Gasification allows separating the combustible fraction
of a waste in a gaseous form from the non-combustible part. The investment cost for such
installation is comparably high, which makes this process worthwhile only when
considerable disposal fees are received.

2. PREPARATION AND HANDLING OF LIQUID ALTERNATIVE FUEL
In many cement plants, liquid waste has been the first waste material to be accepted and
used as alternative fuel because little preparation is necessary and, in certain cases, existing
installations may be used. However, there are three main aspects that need to be
considered when handling liquid waste, such as waste oil and solvents.
? Pollution with Solid Material
Liquid alternative fuel is likely to be polluted with foreign bodies (metal pieces, sand,
plastics, etc.). A filtration at reception is necessary. However, for an efficient
operation not all particles are screened. Small ones remain in the liquid. Therefore,
pumps and kiln injection nozzles must be designed accordingly.
? Inhomogeneity
Liquid alternative fuel often consists of mixtures with different densities.
Homogenization is required to feed a product of constant quality to the kiln. This is
achieved by mechanical mixers.
? Safety
Solvents or liquids containing solvents represent a safety risk since there is a high
potential for explosion. The decisive parameter is the liquid's flashpoint (the
flashpoint is the temperature at which the evaporations of a combustible liquid form
an inflammable gas). Pure motor oil has a flashpoint of up to 200C, whereas
solvents can have one below 0C.
Another safety risk to workers and the environment represent toxic substances.
Special protection and safety procedures are required - according to the standards of
the chemical industry.
Picture 2: Protected worker sampling liquid waste

The flowsheet below shows a typical facility for the handling of liquid alternative fuel. The
main equipment is described in what follows.
Vertical Tanks
with Mechanical Mixing
Feed Pumps
Flow Meter
Discharge Discharge
Filters Pump
Burner
Pressurized Air
for Atomization
Picture 3: Flowsheet of a typical handling installation for liquid alternative fuel
2.1 Filter
At the truck or rail unloading station two filters are installed in parallel to allow the cleaning of
one filter without interrupting the discharge. Typically, basket type filters are used which
allow manual cleaning without evacuation of the system. A second filtration is installed
before the kiln feeding pumps for their protection.
The filter mesh size is determined by the atomizer type used. The finest filter mesh size is
usually chosen between 1.5 and 3 mm. A finer screening is not required as long as the
installation, in particular the pumps, is adequately designed.
Picture 4: Double basket filter

2.2 Tank
The occurrence of mixtures of liquids of different densities requires a continuous
homogenization. A mechanical mixing is always the preferred way to achieve this. Tank
recirculation flow is by far not as efficient, and operating costs are higher (power
consumption of the recirculation pump).
For an efficient mechanical mixing storage tanks are vertical and have a volume of mostly
less than 200 cubic meters. A conical bottom allows for a better sedimentation of solid
particles when the discharge pipe is installed at an elevation. In the above picture discharge
is at the bottom so that no sediments accumulate in the tanks.
Homogenization is done batch wise. Therefore, at least two storage tanks are needed - one
for discharge of new deliveries, the other for kiln feeding.
2.3 Pump
The selection criteria for a kiln injection pump are twofold:
? Dosing Accuracy
For accurate dosing of liquids with varying density and viscosity positive
displacement pumps are used. These are pumps where the volume flow is linear to
the pump speed. The positive displacement pump can deliver a constant pressure
over the full flow rate range. This is a prerequisite for a good atomization at the
injection nozzle.
? Wear Cost
Only certain designs of positive displacement pumps can pump liquids containing
solid material. Gear pumps, for example, show excessive wear rates with polluted
liquids. The internal parts and seals of some pump designs are made of polymer
material. Only some of them can withstand the chemical attack of solvent-type
alternative fuel.
For redundancy, often two pumps in parallel are installed for kiln injection.
2.3.1 Eccentric Screw Pole Pump

This pump is the first choice for the dosing of liquid alternative fuels, such as waste oil. Its
advantage is the small flow rate pulsations (< 1%). The stator inside the pump is made of a
plastic material, which wears out quickly with solvents or liquids containing abrasive solid
material. Standard pumps can deliver a pressure of up to 24 bar.
Picture 5: Cross sectional view of an eccentric screw pole pump

2.3.2 Membrane Pump
For solvent applications this type of pump is widely used. The membrane (usually made of
Teflon) shows adequate lifetimes. The main disadvantage of this type of pump is the high
flow rate fluctuations of up to 5%. Often dampers are installed which allow for a reduction to
less than 2%. There are various designs of membrane pumps with achievable pressures up
to 100 bar.
Picture 6: Electromechanical double diaphragm pump
2.3.3 Centrifugal Pump

For liquid transport, e.g. between truck discharge and storage tank, a constant flow
rate at a certain pressure is not required. This allows using a centrifugal pump, which
is able to handle polluted liquids.
Picture 7: Centrifugal pump of open-wheel design

2.4 Flow Meter
The flow meter controls the pump speed to achieve the flow rate set by the kiln operator.
The commonly used flow meters measure with the coriolis principle. With this principle there
are no moving parts or flow restrictions.
Besides the flow rate, this instrument can also measure the liquid density. This allows for a
correlation with the liquid's water content and can be used for process control as well as to
assess the homogeneity of the liquid.
2.5 Injection Nozzle

Traditional liquid fuel like heavy fuel oil is heated to reduce viscosity so that a proper
atomization is possible. Liquid alternative fuel mostly has a much lower viscosity at ambient
temperature. Heating is not required. For a good atomization, however, it should be
dispersed in a twin-fluid nozzle with the aid of compressed air or steam.
Picture 8: Cross section of a twin-fluid nozzle
Twin fluid nozzles of various suppliers are available on the market. Depending on the design
the required compressed air (steam) flow is about 10 - 12% by weight of the liquid flow. The
required liquid pressure at the nozzle is typically between 3 and about 10 bar. It is important
to select the nozzle in view of the maximum particle size in the liquid. The max. particle size
should not be more than one third to one half of the smallest nozzle cross-section.
2.6 Safety Equipment

Low flashpoint liquids, like solvents, require special attention regarding the design of the
storage and handling facility. At ambient temperature, evaporations - together with the
oxygen of the ambient air - can form explosive gas mixtures.
There are two steps of safety measures applied to minimize the risk for explosion. First of
all, all electric equipment installed at or near the facility has to be explosion proof. This
includes motors, switches, cabling, etc.. The closer the equipment is operated to source of
existing or potential evaporations the better the explosion protection has to be.
Secondly, all areas where there are open liquid surfaces (e.g. inside tanks) nitrogen is
introduced to lower the oxygen concentration below the limit where an explosion can occur.
For the case of fire, fire-fighting equipment (water gun, sprinklers) is installed.

3. PREPARATION AND HANDLING OF PASTY ALTERNATIVE FUEL
In industrial areas there are often considerable amounts of pasty or sludgy waste
materials available, such as
? Resin, paint, varnish
? Oil sludges from tank cleaning
? Distillation residues
? Grease, soap.
Many of these materials are hazardous. With regard to the variable clinker production cost,
they can therefore be of particular interest as considerable disposal fees may be received.
It is important that hazardous waste is burnt at high temperatures with sufficient residence
time (The combustion criteria for destruction of halogenated waste for example is 1200 C
with 2 seconds residence time). This is only ensured at the main burner of a cement kiln.
Hence, a pre-treatment is required to feed such waste at the main burner.
For common types of sludges a pre-treatment can be avoided when pumping it directly to
the kiln.
3.1 Direct Sludge Pumping to Kiln

Waste materials like filter cakes, tank bottoms, sewage sludge, etc. may have to be
screened for removal of coarse foreign bodies before pumping. The pumps used are either
concrete pumps or specially designed double-piston sludge pumps. Sludge feeding is at the
kiln inlet, riser duct or meal ducts. Feed rates are generally limited due to the high water
input or the poor combustion because of the lack of a dispersion of the sludge inside the
preheater.
Picture 9: Sludge feeding installation to kiln inlet (Apaxco plant)

3.2 High Viscosity Liquid Facility
A so-called high viscosity liquid facility aims to liquefy the sludgy waste. This is achieved by
mixing the sludge with liquid fuel. Usually, the solids content of the mix is about 30%. The
important aspect of this process is the compatibility of the substances involved. To avoid
chemical reaction, sludge and liquid need to fulfill certain requirements. This however, limits
the use of such pre-treatment system to certain substances, which is a drawback with regard
to flexibility.
3.3 Impregnation Facility

The other common sludge pre-treatment process aims to prepare a pourable, fine, solid
alternative fuel. This is achieved by mixing the sludges with an impregnation support. Saw
dust is commonly used for this purpose. It has a good absorbency and the impregnated
product shows little stickiness. Alternatively, filter cakes and also animal meal has been
used as impregnation aid.
In what follows the different steps of an impregnation facility are explained.
3.3.1 Reception and Preparation of Primary Materials

The sludgy waste is delivered either in bulk or small containers (typically 200 l steel drums)
and emptied into pits. Drums that cannot be emptied by gravity are either cleaned out with
special equipment or shredded as a whole
Picture 10: Drum emptying by gravity

According to chemical analysis of the incoming waste and the product requirements, the
unloaded materials are mixed to a batch with the desired chemical composition.
3.3.2 Production of Impregnated Alternative Fuel and Dispatch

To this batch the impregnation aid is added. In simple facilities the mixing of the two
components is made by means of a wheel loader. Depending on the quality of the sludge
and sawdust the mixing ratio of the two components is about 1:1. The more sophisticated
process is to use a continuously operated intensive mixer. The better mixing of sludge and
sawdust allows a mixing ratio of about 2:1 (slude:sawdust).
Legend:
1: mixing container
2: mixing tool
3: material guiding arm
4: discharge opening
5: housing
Picture 11: Intensive mixer (type Eirich)
Before storage, metals are separated and the product is screened in a drum screen. The
product is fine (mostly 10 mm) and well suited for incineration at the main firing.

4. PREPARATON AND HANDLING OF SOLID ALTERNATIVE FUEL
Solid waste can be fed in a coarse form to the kiln (e.g. whole tyres) but often needs to have
a finer granulometry. A size reduction is necessary and determined mainly by the following
process factors:
? Transportation cost to the cement plant
? Handling properties
? Kiln feeding point
? Thermal substitution rate
? Kiln behavior.
The mechanical processes used for the size reduction of solid waste are shredding and
granulation. They are typically applied for materials such as wood, rubber, plastics or waste
tyres.
4.1 Shredding and Granulation of Solid Waste
4.1.1 Shredders
A shredder is a machine for primary size reduction of solid materials. Its operating principle
is based on a tearing and shearing action of slowly rotating tools (10 40 rpm) agitated by a
high torque drive.
Picture 12: Typical shredder design with two counter-rotating shafts
Typically, a shredder consists of two counter-rotating shafts. On these shafts engaging disks
with hook type knives are mounted to grab and tear the material fed. Such shredder is
preferably used for the size reduction of high-density waste (e.g. wood and tyres). Waste
materials with a low bulk density such as plastic films require a shredder with large feed
opening and a mechanism to push the material towards the cutting rotor. For such
application single rotor shredders are used.

Depending on the knife size and operation (multiple stage, closed circuit) a minimum chip
size of 20 50 mm can be achieved.
4.1.2 Granulators
Granulators (also called grinders) are used for secondary size reduction. The typical design
consists of a single rotor with a fixed counter knife and an outlet sieve. There is a
mechanism to push waste material towards the rotor where knives grab it. After being cut at
the fixed counter knife the waste chips are further milled before they pass the outlet sieve. A
granulator operates at higher rpm than a shredder, above about 100 rpm.
Pusher
Sieve
Rotor
Screw
Picture 13: Typical granulator design with one shaft and outlet sieve
Due to their high rotor speed, granulators are used for low abrasive material. Waste
containing considerable amounts of metals and/or mineral material should not be granulated
because of excessive wear. The spacing between the rotor knife and the fixed counter knife
needs to be small (about 0.5 mm) to efficiently cut fine material like plastic films.
4.1.3 Multi-Stage Size Reduction

For the firing of waste materials at the preheater/precalciner, a single stage size reduction is
sufficient in most cases. If bulky waste material is to be fired at the main burner a multiple
stage shredding and granulation is required for the efficient production of an alternative fuel
of a few millimeter size. This process is however very costly (electrical energy consumption,
wear parts) and only of interest where high disposal fees are received (see Low Grade Fuel
Study for cost of shredding of solid waste).

4.2 Handling of Solid Alternative Fuel at the Cement Plant
The handling at the cement plant comprises storage, extraction, transport, dosing and
feeding. In view of the various possible feeding points of solid alternative fuel (see LGF-
Study) a classification in three different size classes can be made, see below.
Size Approx. Typical Kiln
Denomination Example Reduction Granulometry Feeding Points
Lump Fuel
Whole Kiln Inlet
None 0.5 1.2 m
Tyres Mid-Kiln
Coarse Solid Single

AFs Stage 100 250 mm Riser Duct
Tyre Shredding
Chips
Multiple
Stage +/- 50 mm Precalciner
Shredding
Fine Solid
AFs Impreg.
Granulation < 50 mm* Main Burner
Sawdust
* Max. Size determined by type of waste and substitution rateTable 1: Classification of

solid alternative fuels
The flowsheet below shows a typical facility for the handling of solid alternative fuel. The
main equipment is described in what follows.
Precalciner or Secondary Firing
Belt Conveyors
Storage Hall
Double or
Triple Flap
Bridge Crane Gate
Belt Scale
Feed Hopper
with Activated
Flat Bottom Main Firing
Vertical walls
Drum Screen
Rotary Valve
Belt Conveyors Burner
Picture 14: Typical flowsheet of a kiln feeding installation for solid alternative fuel

4.2.1 Storage
Alternative solid fuel is mostly stored on a stockpile inside the plant. Storage should be
made on a concrete floor to avoid any ground contamination. For protection against rain and
wind the storage yard is covered.
Movement of the often difficult to handle alternative fuel is successfully made with a front
loader or crane. Because of the difficult handling properties the storage volume of feeding
hoppers is kept small to reduce compacting of the material inside the hopper.
For environmental, health or safety reason some solid alternative fuels (example animal
meal) are handled and stored in confined systems (pneumatic transport, storage silo).
4.2.2 Extraction & Dosing System

Because of its coarse granulometry, high moisture content or stickiness, solid alternative fuel
often shows difficult handling properties. Specially designed extraction and feeding systems
are required in such case. There are some general design guidelines:
? To avoid material build-ups on walls and bridge building a feeding hopper is best
equipped with vertical walls.
? Feed hopper extraction is best made with an activated floor extraction. Key is
to have a system, which allows for form closure with the material. This can be
achieved with screw or chain conveyors, extraction arms or so-called moving
floors, see picture below. An extraction by means of friction (e.g. belt conveyor)
is often not suited.
Picture 15: Moving floor hopper

? For volumetric flow rate control, activated extraction aids are giving better
results than simple limiting gates, see below.
Picture 16: Extraction problem with limiting gate Picture 17: Extraction with counter-
rotating drum
For the dosing of solid alternative fuel belt weigh feeders are commonly used. They show
good flexibility and can serve for a wide range of solid fuels. Other dosing systems like rotary
or loss-in-weight scales are applied in special applications.
4.2.3 Kiln Feeding
4.2.3.1 Gravimetric Kiln Feeding

At precalciner, kiln inlet or riser duct solid alternative fuel is best fed by a mechanical
transport system followed by a gravimetric feeding. The lock between feeding system and
kiln is a double or triple pendulum flap. The flaps are pneumatically driven and alternately
open or close to minimize false air intake.
During up-set kiln operation with positive pressure at the feeding point the flap gates do not
sufficiently close the duct and hot gases may escape. Therefore, shut-off gates above or
below the flap gates are installed.
4.2.3.2 Pneumatic Kiln Feeding

At the main burner fine solid alternative fuel has to be fed pneumatically. A corresponding
system includes first of all a screening of foreign material that could lead to blockages in
feeding pipes. Drum screens are commonly used for this purpose.
Solid alternative fuel for main burner firing (e.g. animal meal, impregnated sawdust, plastic
chips) often causes clogging in pneumatic transports due to its chip size or stickiness. To
minimize handling problems the pneumatic transport should be straight, as short as possible
and have a constant cross section. The feeding device to the pneumatic transport is a rotary
valve.

When solid alternative fuel is introduced with an existing burner, the injection pipe is placed
parallel to and on top of the burner because there is usually no wide enough pipe in the
center. In case a new burner is purchased it should have a central feeding option.
Depending on the type of alternative fuel either one of the injection points works better.
Mostly, it is found after a series of trials.
Injection Pipe
Picture 18: Injection pipes for solid alternative fuel on top and in the center of the
main burner
The injection velocity is lower than for traditional fuels, often in a range of 20-30 m/s. Many
plants have conveying air blowers with VFD (variable frequency drive) drives for better fine-
tuning of the injection velocity.

5. EXAMPLES
5.1 Typical Liquid Waste Handling

A typical handling facility for the reception of liquid waste and the production of liquid
alternative fuel is shown in picture 9.
The installation consists of the following main elements:
? Liquid waste decanting tank with primary filtration
? Shredder to crush oversized particles for protection of the pump and to avoid
blockages.
? Discharge pump, typically of centrifugal type
? Liquid storage tank with mechanical mixer
? Loading pump, typically of centrifugal type
? In case of low flashpoint liquid, a nitrogen inertisation system
? Sprinkling system for fire suppression
Typical Liquid Waste Handling
Fire Protection
N2
M
Liquid Waste Liquid A.F.
CC00-032.dsf / Kma 7.3.2000

5.2 SCORIBEL Impregnated Sawdust Production
SCORIBEL is a Belgium based subsidiary of Ciments dObourg. At the impregnation
workshop of its Seneffe plant solid, pasty and liquid waste is mixed with an impregnation aid
to form a fine, solid alternative fuel for incineration at the main firing of cement kilns.
The installation consists of three main parts:
? A reception section for industrial waste in bulk or containers as well as the
impregnation aids sawdust and filter cake. In a closed building the liquid, pasty and
solid waste is premixed and shredded.
? By means of a continuously operated intensive mixer the waste sludge is mixed with
the impregnation aid before being screened.
? In two moving floor type silos the product is stored and homogenized before being
loaded to trucks.
For reduction of organic emission air is aspirated at the main sources of emission and sent
to a thermal treatment unit.
The plants design production capacity is 15 t/h, presently it is operated at more than 20 t/h.
Investment cost for the impregnation facility (excl. air treatment): BEF 250 Mio.

SCORIBEL - Impregnated Saw Dust Production
Saw Dust Solid Waste Waste in Drums Bulky Waste Liquid Waste
Mineral/Organic
Bulk Storage
Building
Pit Building
Drum Storage
Hangar
Liquid Waste
Drum Recycling Mixing Pit
Unprocessible Drums
>250
Mixing Tower
Fe
50/250 12/50 0/12
Product Storage
CC00-031.dsf / Kma 7.3.2000

5.3 Industrial Plastic Waste Shredding at PLASTREC AG
PLASTREC AG, a subsidiary of Holcim (Schweiz), operates a facility for the shredding of
industrial plastic and rubber waste to a product for use at the primary firing of cement kilns.
The waste materials received are cables, textile reinforced plastic, residues from
manufacture of hard plastic items, plastic films and rubber.
The facility is designed to mechanically process the incoming waste to a product of 10 mm
size. This is achieved by pre-shredding and subsequent grinding of the material. Two
shredders and two granulators, each operated in parallel, are used for the two-stage
process. Material transport is done by belt conveyors for raw and pre-shredded material, and
pneumatically for the finished product.
The plant has an hourly production rate of 1.5 2.0 t. The annual production is about
5000 t. The consumption of electrical energy is 100 kWh/t for the whole facility.
Investment cost for the whole installation: CHF 1.2 Mio.
FLOWSHEET PLASTREC AG
Vibration Feeder
55kW Shredder
30kW Shredder
Ni, Cr Al, Cu
30kW 45kW
Granulator Granulator
Feed Hopper for

Granulated Plastic
Big Bag
Afl-012.dsf / Kma 3.11.98

5.4 The whole Tyre Feeding System at Siggenthal Plant
AFR : Tires at Siggenthal
Tire individualization:
Four 90direction change
with roller conveyors
photoelectric
barriers
max 300t
hopper with moving
floor feeder (Saxlund)
manual feed
point
oversize
tires
double flap gate
Kiln capacity : 2000 t/d safety flap gate
Tires capacity : 0-2,5 t/h

In operation since : 1992
Investment cost : 2.3 Mio CHF
WVO-004.dsf / Kma 2.10.01
5.5 Plastic Chip Feeding System at Untervaz Plant

Handling of Plastics and other solid Alternative Fuels at BCU Untervaz
Kiln Big-Bag Discharge

Gas Grannulated Material
Hot Weight
Meal Feeder
WEIMA-Shredder UNTHA-Shredder
45 m 3
Capacity : 1.2 t/h

In operation since : 1992 Weight
Feeder
Investment Cost : 1,4 Mio sFr
Dp-061.dsf / Kma 25.2.03

5.6 Feeding System for Coarse Shredded Waste (Tyre & Plastic Chips) at Altkirch
Handling of Tyre Chips and Lump Alternative Fuel at ALTKIRCH
Automatic
Feed Control
Shut off
Gates
5m3
Extraction
with variable
Speed Stock Bunker
300m3
Double Flap
Gates
Opening 0,8x0,85m
Capacity : 2.0 t/h
Belt Scale
In operation since : 1994
Investment cost : 6.2 Mio FF
Dp-030e.dsf Kma 8.6.98
5.7 Impregnated Sawdust Feeding System at Eclpens Plant

Handling of Impregnated Saw Dust and Plastic Chips (CSS) at Eclpens
level
sensor
135m3 rotary valve
magnetic
SAXLUND separator
feeder with to main
movable floor injector burner
rotary separator
holes: 25mm
belt
variable scale oversize
speed screw material
Kiln capacity : 1700 t/d

CSS capacity : 2 -6 t/h
Investment cost : 1.5 Mio CHF
Dp-059.dsf / Kma 22.8.01

5.8 Used Oil Facility at Lgerdorf Plant
Burning of Waste Oil at LGERDORF Plant
Burner
Storage Area Platform Rotary Kiln
High Pressure Pumps (2) Flowmeter
4 t/h
Central Channel of
Pillard Burner
Pressurized
Air
Storage Tank Storage Tank Purge Tank Main firing of 3'100 t/d
80m3 80m3 Lepol Kiln (until 1996)
max. 30% Waste Oil
Filling Pumps (2)
Investment for Equipment 1.7 Mio DM

Costs for Test (expected) 1.0 Mio DM
Design for class A1
(Flashpoint < 21C)
CC00-10.dsf Kma 28.1.00

Chapter 8
Circulation Phenomena

Circulation Phenomena in the Clinkerization Process
Ren Hasler, Daniel Brassel

PT 99/14503/E
1. INTRODUCTION ............................................................................................................. 510

2. MECHANISM OF THE CIRCULATION PHENOMENA .................................................. 513
3. CIRCULATING ELEMENTS IN THE KILN SYSTEM ..................................................... 520
3.1 Input of Circulating Elements ............................................................................ 520
3.2 Enrichment of Circulating Elements / Endangered Zones for Encrustation
Formation ........................................................................................................................ 520
3.3 Output of Circulating Elements ......................................................................... 521
3.4 Volatility of Circulating Elements....................................................................... 525
3.5 Condensation of Circulating Elements .............................................................. 536
4. TYPICAL APPEARANCE OF BUILD-UPS..................................................................... 538
5. KILN OPERATION PROBLEMS DUE TO CIRCULATING ELEMENTS........................ 547
6. IDENTIFICATION OF PROBLEMS WITH ENCRUSTATIONS AND BUILD-UPS ......... 547
6.1 Material Balance ............................................................................................... 548
6.2 Criteria and Indicators to Assess the Build-up Problem.................................... 549
6.3 Example of a Circulation Phenomena Problem ................................................ 553
7. MEASURES AGAINST BUILD-UP FORMATION .......................................................... 556
7.1 General Measures............................................................................................. 556
7.2 Intelligent Cleaning ......................................................................................... 556
7.3 Measures against Chloride Problems ............................................................... 561
7.4 Measures against Sulfur Problems ................................................................... 561
7.5 Measures against Alkali Problems .................................................................... 564
8. MATHEMATICAL MODEL TO SIMULATE THE CYCLES OF THE CIRCULATING
ELEMENTS ..................................................................................................................... 565

SUMMARY
This chapter describes the circulation of volatile elements in the kiln system. It indicates the
tolerable inputs of circulating elements so that no excessive build-up and clogging problems
arise.
In particular it shall serve as guideline how an encrustation problem, caused by volatile
elements, is systematically solved.
1. INTRODUCTION
Alkali, sulfur and chlorine compounds (hereafter called circulating or volatile elements, see
below) in raw materials and fuels utilized for the cement manufacture, when present in high
concentrations often given rise to difficulties in kiln operation with build-up formation, mainly
in the preheater and the kiln inlet section.
Volatile Elements (VE):

Sulfur SO3
Potassium K2O
Sodium Na2O
Chlorine Cl
The build-up formations cause cyclone blockages or restrictions of the kiln inlet area so that
the kiln has to be stopped for cleaning. In extreme cases more than 200 annual kiln stops
due to blockages may occur, so that the impact on kiln availability and productivity can
become a severe issue. Suspension preheaters armed with dozens of shock blowers (Fig.
1a, 1b) to prevent build-up formation illustrate the importance of this problem.

Fig. 1a

Fig. 1b

2. MECHANISM OF THE CIRCULATION PHENOMENA
Depending on the degree of volatility, the circulating elements evaporate in the sintering
zone of the cement kiln and are carried by the gases to colder zones, where they mainly
condense on the raw meal and partly also on the surrounding walls.
Afterwards they return with the raw meal into the sintering zone where they partly
reevaporate depending on the degree of volatility. This repeated process through the kiln
leads to the establishment of internal cycles (see Fig. 5). Finally the cycles reach equilibrium
so that the output of circulating elements is equal to their input by the raw materials and the
fuels (cp. Fig. 2).
Fig. 2 Circulation of Volatile Elements within the Kiln System

Almost all the circulating elements finally leave the system with the clinker. However, this is
only the case when beforehand sufficiently high internal cycles of the volatile elements have
been formed. The extents of these cycles depend on the degree of volatility of the circulating
elements. As the latter recondense on the colder raw meal and the surrounding walls, the
formed sticky molten salts are able to reduce the fluidability of the raw meal and, if present in
sufficient quantities, to glue it finally on the walls. From time to time, especially during a
change of the temperature profile, pieces of build-ups fall down and mainly block the cyclone
outlets (Fig. 3).

Fig. 3 Build-ups due to Circulating Elements in the Various Kiln Systems

If the amount of molten salts becomes too high, either because of an excessive input of
volatile elements or due to a high degree of volatility, the installation of a kiln gas bypass
becomes necessary in order to extract part of the circulating elements from the kiln system
(Fig. 4a, 4b, 4c).
Fig. 4a Conventional Bypass with Water Injection

Fig. 4b Hot Gas Bypass without Water Injection
Fig. 4c Bypass with Gas Feedback

A small part of the circulating elements leave the kiln system with the main exhaust gas dust.
The latter is efficiently precipitated in the dedusting unit and is normally reintroduced into the
kiln system. This is called the external cycle of the circulating elements (Fig. 5).
Fig. 5 Circulation Phenomena: Internal and External Cycle
In normal cases a negligible amount of circulating elements is emitted by the stack into the
atmosphere. However, the emission of SO2 is not negligible anymore, when the sulfur in the
raw material is present in form of sulfide (FeS2, PbS, ZnS) or organic compounds. In this
latter case the sulfides are volatilized in the temperature range of 400 to 600C and leave
the kiln system partly as gaseous SO2 emission (Fig. 6 and section 3.3.4).

Fig. 6 SO2 Emission in case of Sulfides

3. CIRCULATING ELEMENTS IN THE KILN SYSTEM
3.1 Input of Circulating Elements

In the following the typical inputs of circulating elements by the raw materials and the fuel(s)
are indicated. It is differentiated between input ranges where usually no encrustation and
build-up problems arise and input ranges that usually lead to severe clogging problems.
Important: All figures may serve as rough guidelines only.
The real limits for the build-up and clogging formation depend on a lot of
individual parameters like the degree of volatilization, the temperature
profile, the completeness of combustion and the excess air factor as well
as the kiln system itself. The indicated values are guidelines for suspension
preheater kiln only.
3.1.1 Input by the Raw Materials (loss free basis)

Alkalis (K, Na):
Generally appear as interlayer cations in the clay minerals and the feldspars.
Sulfur:
The sulfur is introduced in several mineralogical forms:
as sulfate: gypsum CaSO4 2H2O
anhydrate CaSO4
as sulfide: pyrite FeS2, organic compounds
In the following only the sulfates are considered.
Chlorine:
The chlorides are mainly introduced as NaCl (from seawater) or KCl.
Note: Thekiln feed normally includes already external cycle (see Fig. 2) so that its
concentration on circulating elements is higher than the one of the pure raw mix.
Typical limits for 4 to 5 stage SP kilns are listed in section 6.2.2.
3.1.2 Input by the Fuel(s)

Mainly sulfur is introduced by the fuels such as Coal, Coke, fuel oil and so on. In Addition
alternative fuels often contain a reasonable amount of sulfur.
Typical limits for 4 to 5 stage SP kilns are shown in section 6.2.2.
3.2 Enrichment of Circulating Elements / Endangered Zones for Encrustation

Formation
Depending on the individual circulating element and its compounds, the condensing point
lies in the temperature range of 650 to 1000C.
Condensation: 650 800C : Chlorides and its compounds
800 1000C : Sulfates
Therefore the endangered zones for the formation of build-ups by the condensed circulating
elements depend on one hand on the circulating element and its compounds and on the
other hand on the kiln system itself (see also Fig. 3).
In the following the build-up zones for the various kiln systems are indicated. At the colder
end usually the encrustations by the chlorides are found whereas at the hotter end the ones
of the sulfates are met.

Kiln System Zones of Build-ups
Precalciner kiln / 4 The two lowest cyclones stage and their riser
or 5 stage SP kiln ducts, kiln inlet area (first section of rotary part)
Lepol kiln Second preheating chamber, kiln inlet (first
section of rotary part)
3 stage SP kiln Lowest cyclone and its riser duct, kiln inlet, first
section of rotary part
2 stage SP kiln Lowest riser duct, kiln inlet, first section of rotary
part
Long dry kiln / wet Calcining zone of rotary part (transition zone)
3.3 Output of Circulating Elements
3.3.1 Clinker
The chlorides are so volatile that they hardly leave the kiln via clinker. In exceptional
cases, when the sintering zone has largely cooled down or when embedded in big
material lumps, the chlorides may leave the kiln in major quantities. However, normally
the chlorides form a large cycle within the kiln system and they need to be extracted by a
kiln gas bypass.
The sulfur and the alkalis leave the kiln system normally via clinker either as definite
compounds:
K2SO4, K3Na(SO4), Na2SO4
Ca2K2(SO4)3
CaSO4 (rare!)
or in solid solution in clinker minerals:
K with the Belite, Aluminate
Na with the Aluminate
SO3 with the Belite
The calciumanhydrate CaSO4 is rather volatile (decomposes at temperatures > 1000C)

and forms therefore a large sulfur cycle in the kiln system. Therefore it is very important
that there is sufficient alkalis to combine with the sulfur and to leave the kiln system as
alkali/sulfur compound.
3.3.2 Kiln Gas Bypass

Kiln gas bypasses (see Fig. 4) are mainly used to extract the very volatile chlorides from SP
preheater kilns. Chloride bypasses withdraw typically 5 to 15% of the kiln gases. The hot
gases from the kiln inlet are quenched down by fresh air, sometimes supported by injection
of water into the quench chamber, to a temperature below 600C. The gaseous chlorides
condense onto the withdrawn dust particles and are separated finally in an electrostatic
precipitator or a bagfilter (see Fig. 5a, 5b, 5c).
In rare cases kiln gas bypasses are also applied in case of too much CaSO4 and not
sufficient alkalis.

Also for the production of low alkali clinker large kiln gas bypasses are used (20 to 50% of
the kiln gases) in order to withdraw the alkalis from the kiln charge and to produce a clinker
with an alkali content of < 0.6% (i.e. Na2Oeq < 0.6%). Thereby the alkalis are volatilized by
hard burning and by injection of chlorides into the kiln (see also section 0).
The bypass dust must be discarded e.g. into the quarry or is partly reutilized as additive to
the cement, or in rare cases may be sold as filler material.
A ballpark figure for the amount of bypass dust: 15 g/kg cli per 10% bypass
Due to the extraction of the bypass gases additional heat loss arise. The specific value per
% bypass depends on the kiln system:
Precalciner kilns: 8 11 kJ/kg cli
Preheater kilns: 20 kJ/kg cli
Typical chemical concentrations of bypass dusts are found in Fig. 7 and 8.
3.3.3 Kiln Dust

Another possibility to withdraw circulating elements from the kiln system is via external cycle
(see Fig. 5) which is interrupted and partly discarded. The enrichment of this kiln dust by
circulating elements depends on the kiln system. Whereas kiln dusts from long wet kilns are
highly enriched, the ones from SP kilns show nearly the same chemical composition than the
raw meal. Typical chemical concentrations of dusts from the different kiln systems are shown
on Fig. 7 and 8. Thereby the following legend applies:
WL = long wet kiln
DL = long dry kiln
DG = dry kiln with grate preheater (LEPOL)
DS = SP kilns: F Filter dust from kiln exhaust gas
BP Bypass dust

Fig. 7 Kiln Dust Analyses

Fig. 8 Kiln Dust Analyses
3.3.4 Emission by Exhaust Gas

Normally there is no emission of circulating elements by the exhaust gas. The only gaseous
component, the SO2 which does not condense at low temperatures is effectively absorbed
by the free lime at the lowest cyclone stage.
However, a source of SO2 emission is the raw material sulfur in form of sulfide (e.g. pyrite).
The sulfides decompose at temperatures between 400 to 600C, forming SO2.
Approximately 70% of the SO2 immediately reacts with the CaCO3 of the kiln feed to CaSO3
and finally at higher temperatures to CaSO4. The residual 30% of the SO2 are partly
absorbed on their way with the flue gases to the stack, mainly in the raw mill and in the
conditioning tower. However, a certain portion of the SO2 is finally emitted by the stack,
depending on the kiln and raw mill system (see Fig. 6).

3.4 Volatility of Circulating Elements
3.4.1 Definition of Total Volatility

The total volatility of a circulating element or a compound of it is defined as shown in Fig. 9.
Fig. 9 The Volatility of a Circulating Element
The total volatility of a circulating element or of a compound of it indicates which portion of it

is volatilized in the kiln and does not leave the kiln straight with the clinker.
The total volatility is:
ccli
= 1
cHM
whereas = factor for total volatility
cHM = concentration of the circulating element at the kiln inlet
ccli = concentration of the circulating element in the clinker

3.4.2 Affinity of the Circulating Elements
Circulating elements have a strong affinity to other elements and form chemical compounds.
The following affinity order has been observed (see also Fig. 10):
1) The chlorine combines first with the alkalis, forming KCI, NaCl.
The residual chlorine, if any, combines with the calcium, forming CaCl2.
2) The residual alkalis combine with the sulfur, forming K2SO4, Na2SO4. They form also
double salts Ca2K2(SO4) 2, K3Na(SO4) 2.
The residual alkalis, if any, combine with CO, forming K2CO3, Na2CO3.
3) The residual sulfur (SO3 or SO2) combines with CaO, forming CaSO4.
Fig. 10 Affinity

3.4.3 Volatility of the Circulating Element Compounds
3.4.3.1 General
In the kiln the alkalis are liberated from the clay mineral lattice. They partly dissociate
into the gas phase and recombine with other elements according to the affinity order as
described in paragraph 3.4.2. The rest recombines directly in the charge material with
chlorine or sulfur or is integrated into the lattice of the clinker minerals (belite and
aluminate, see paragraph 3.3.1). Especially the sodium is little volatile and goes
preferably direct into the clinker (Fig. 11).
Fig. 11 Circulation of Alkalis

The chlorides liberated during heating of the material and combustion of the fuel react with
the alkalis to form alkali chloride. This reaction takes place either in the charge material or,
after vaporization, in the kiln gas (Fig. 12).
Fig. 12 Circulation of Chlorine
Upon volatilization of sulfur at burning zone temperatures in the cement kiln, sulfur
dioxide is the predominant component. The gaseous SO2 derives either from the
dissociated sulfur of the combustibles or the decomposed CaSO4 and the partly
volatilized Alk2SO4 from the charge material.
Subsequently if there is an excess of alkalis, the internal sulfur cycle is caused primarily by
the reaction with these, which takes place almost entirely in the rotary kiln. The alkali
sulfates formed in this reaction are, in so far as they are present in vapor form, precipitated
on the material. This occurs mainly in the rotary kiln itself, but partly also in the preheater.

The precipitated alkali sulfates thus travel through the kiln again, pass through the sintering
zone, and are partly discharged from the kiln with the clinker, while some of the alkali
sulfates remain in the cycle.
The calcium sulfate (CaSO4) is formed as a result of the reaction between SO2 and CaO. It
is partly present already in the raw materials or is formed preferably in the temperature
range of 800 to 900C as illustrated in Fig. 13. Above a temperature level of 1000C the
calcium sulfate starts decomposing, first rather slowly and beyond a temperature of 1300C
rapidly. Sulfur dioxide again is formed. If the alkali input is not high enough to combine the
whole incoming sulfur as alkali sulfate, high sulfur dioxide concentrations in the kiln gas
arise. In such circumstances calcium sulfate may pass un-decomposed through the sintering
zone leaving the kiln embedded in the belite mineral or even as CaSO4. The sulfur cycle is
illustrated in Fig. 14.
Fig. 13 SO2 Absorption by CaO resp. CaCO3

Fig. 14 Circulation of Sulfur

3.4.3.2 Volatility of the Compounds of Circulating Elements
Fig. 15 indicates the relation between vapor pressure of various alkali compounds at
different temperatures. This demonstrates that the volatilization mechanism is dependent
upon the form of combination of the particular volatile element.
Fig. 15 Vapor Pressure
Chlorides: KCl, NaCl, CaCl2

At a temperature level of 1200 to 1300C the chlorides are already volatilized to a great
extent. At sintering zone temperatures they are almost entirely volatilized so that the total
volatility factor is approx. 0.97 to 1.

Sulfates: Alk2SO4, CaSO4
Basically, the Alk2SO4 are little volatile, whereas the CaSO4 is highly volatile.
Therefore the most important criteria for the total sulfur volatility is the molar ratio between
the alkalis and the sulfur itself, corrected by the chlorine:
Molar alkali / sulfur ratio:
K 2O Na2O Cl
+
Alk
= 94 62 71
SO3 SO3
80
Alk
Desirable value: 1 .2
SO3
If there are sufficient alkalis available to combine with the total sulfur input (from raw
materials as well as from the fuel), the total volatility for the sulfur is = 0.3 to 0.5.
However, if there is a surplus of sulfur over the alkalis, the highly volatile CaSO4 is formed.
Its volatility is approximately = 0.9, but can increase up to = 1, depending on the
operating conditions of the burning process.
In general the sulfur volatility is very much depending on the operating conditions of the
burning process, such as
the maximum temperature in the sintering zone
the retention time of the kiln charge material at high sinter zone temperatures
the granulometry of the kiln charge material (diffusion to the surface of the granules)
the partial pressure of O2 in the kiln atmosphere
the partial pressure of SO2 in the kiln atmosphere
For more details see next paragraph 3.4.4.

3.4.4 Parameters Influencing the Volatility of the Sulfur Compounds
3.4.4.1 Maximum Temperature in the Sintering Zone

The essential feature of the volatility - temperature curve is that above a certain critical
temperature the volatility increases first exponentially and then slowly reaches a maximum
(Fig. 16).
Fig. 16 Volatility-Temperature Curve
For sulfur this critical temperature is in the order of the burning temperature. Consequently
hard burnable raw mixes or overheating of the kiln charge material (low free limes in the
clinker) have a very important bearing on sulfur volatility.
3.4.4.2 Retention Time of the kiln Charge Material at High Burning Temperatures
Volatilization of the sulfur compounds is a dynamic process. The longer the kiln charge
material is exposed to the high burning temperatures, the higher the total amount of
volatilized sulfur.
3.4.4.3 Granulometry of the Kiln Charge Material

If the kiln charge material is well granulated, the sulfur takes more time to diffuse to the
surface of the granule. Consequently less sulfur volatilizes.
On the other hand poor granulometry conditions result in higher dust formation which
negatively influences the temperature profile in the kiln and shifts it towards the kiln inlet.

3.4.4.4 Kiln Atmosphere
The composition of the kiln atmosphere is an essential factor for the sulfur volatility:
m
Men (SO4 )m n MeO + m SO2 + O2
2
The equilibrium of the dissociation is shifted with increasing O2 partial pressure and SO2
partial pressure in favor of the sulfates.
Too little excess air or also local reducing kiln atmosphere increases the volatility of the
sulfur.
This behavior is illustrated for the K2SO4:
K 2 SO 4 K 2 O + SO 2 +O 2 O2 excess in burning zone:
K 2 SO 4

K 2 O + SO 2 + O 2
O2 deficiency in burning zone:

K 2 SO 4 K 2 O + SO 2 + O 2
Laboratory tests have been executed for the primary volatility 1, i.e. the volatility of the
sulfur when exposed the first time to high temperatures. It is remarkable that already at a
temperature level of 1000C, the sulfur volatility goes up to 1 (100% volatility) if there is no
oxygen in the kiln atmosphere. This is the case if the combustible produces a local reducing
atmosphere in the kiln charge material e.g. when burning whole tires at the kiln inlet. From
these tests it can be derived that the main influence of the oxygen content on an increase of
the sulfur volatility is produced between 0 to 2% O2. Beyond 2% O2 the influence is
negligible (see Fig 17).

Fig. 17 Sulfur Volatility depending on Oxygen content
The partial pressure of the sulfur dioxide (SO2) also decreases the sulfur volatility. That
means that if the sulfur cycle in the kiln is large enough, the dissociation of the sulfates
decreases and the sulfates can gradually leave the kiln with the clinker. However, this is not
much of a help, as the required high sulfur cycles already cause encrustation problems!

3.5 Condensation of Circulating Elements
As previously stated, the circulating elements volatilize in the hot zones of the kiln and
condense at the colder areas of the kiln system (internal cycle).
Melting point and boiling point (at 1 bar) of some chlorides, sulfates and carbonates are
listed below. Generally speaking it can be stated that the chlorides condense at lower
temperatures than the sulfates.
Melting Point [C] Boiling Point [C] at 1 bar

Chlorides NaCl 801 1413
KCl 776 1500 (sublim.)
CaCl2 772 > 1600
Sulfates Na2SO4 884 ?

K2SO4 1069 1689
CaSO4 1280 (d) -
Carbonates Na2CO3 851 (decomp.)

K2CO3 891 (decomp.)
In the melting phase secondary compounds are formed which are more complex than the
original volatile elements:
Secondary Compounds:
Chlorides Sulfates Carbonates Spurrites

KCl K2SO4 K2CO3, 2 C2S CaCO3
(NaCl) K3Na (SO4)2 (Na2CO3) 2 C2S CaSO4
Na2SO4
Ca2K2 (SO4)3
CaSO4
Furthermore, investigations have revealed that calcium sulfate, calcium oxide and alkali
sulfate form eutectic melts which can have much lower melting temperatures than the single
compounds (Fig. 18). Under presence of chlorides these melting temperatures are even
more lowered to temperatures below 700C. Very low melting temperatures show also the
alkali carbonates. However, in such a case the alkalis can be sulfatizated by the addition of
gypsum so that it presents normally not a major problem.

Fig. 18 Melt Intervals in the System
CaSO4 - K2SO4 - Na2SO4
All these melts, when present in sufficient quantities, lead to severe encrustations and build-
up formation in the preheater and kiln inlet area.
A further problem is given by the fact that the volatilization process goes along with an
endothermic reaction whereas the condensation is exothermic. By this way an important
amount of heat is drawn from the sintering zone to the kiln inlet so that the area of
condensation is shifted even more towards the cold end of the kiln system.

4. TYPICAL APPEARANCE OF BUILD-UPS
Typical appearance of build-ups and deposits are shown in Fig. 19, 20 and 21.
Fig. 19 Macrographs of Deposit Types

Fig. 20 a) .. d) S.E.M. Micrographs of Deposits
Fig. 20 a) REM 84/506

HD - 1st cyclone: KCl blocks embedded in fine matrix

Fig. 20 b) REM 84/525
HD - 2nd cyclone: CaSO4 and C2S

Fig. 20 c) REM 84/502
HV - 4th cyclone: Binding nature of glassy KCl matrix

Fig. 20 d) REM 84/550
GM - 4th cyclone: KCl crystal structure filling pore

Fig. 21 a) .. d) S.E.M. Micrographs of Deposits
Fig. 21 a) REM 84/113

TU - Riser pipe: CaSO4 and C2S

Fig. 21 b) REM 84/117
TU - Riser pipe: K2SO4 and Ca SO4 border

Fig. 21 c) REM 84/72
WU - Kiln inlet: Fly ash balls

Fig. 21 d) REM 84/69
WU - Kiln inlet: Primary spurrite

5. KILN OPERATION PROBLEMS DUE TO CIRCULATING ELEMENTS
The consequences due to high internal cycles of the circulating elements are rather severe.
First it starts by an increased encrustation and build-up formation at the kiln inlet area and
the lower part of the preheater. As a consequence the pressure loss across the system
increases and at the same time also the inbleeding false air quantity increases. This reduces
the maximum kiln draft and thus the maximum clinker production. When the kiln operator
tries to compensate the lower available kiln draft by a lower excess air rate, the situation
becomes even worse as the sulfur cycle further increases.
Studies have revealed that a high chloride cycle impedes also complete combustion. This in
turn further increases the sulfur cycle.
High sulfur cycles lead to a poorly granulated clinker and therefore to dust formation. The
dust entrains the heat from the burning zone to the kiln inlet, so that the cycles of the volatile
elements further increase due to a longer residence time at high temperatures.
Furthermore, the high cycles of volatile elements transport the heat of the sintering zone to
the area of condensation by the endothermic - exothermic reactions, causing the same
effect as the dust cycles described above.
The result of these mechanisms is always the same:
Frequent kiln stops due to encrustation and clogging problems, i.e. reduced kiln
utilization factor ( reduction of OEE)
Reduction of the maximum kiln production
Higher heat consumption
Formation of unstable coating at the transition zone and thus high refractory
consumption
6. IDENTIFICATION OF PROBLEMS WITH ENCRUSTATIONS AND BUILD-UPS

Chlorides, sulfates, alkalis or any combination of them can cause encrustation and build-up
problems. Therefore a detailed analysis is a prerequisite to find an appropriate solution.
The analysis can consist of eight steps as follows:
Systematic analysis of the hot meal (cp section 6.1)
(every shift: LOI, SO3, Cl, K2O, Na2O)
Comparison with the defined limits of circulating elements in the hot meal!
Chlorine problem, sulfur problem, alkali problem of combination?
Performing of a material balance with the inputs and outputs of the circulating elements
(cp section 6.1).
Where do the circulating elements come from?
Comparison with typical tolerable inputs of circulating elements by raw materials and
fuels.
Which amount is critical?
Calculation of the molar alkali / sulfur ratio and comparison with standard.
Are there enough alkali present to withdraw the sulfur within the clinker?

Calculation of the sulfur volatility and comparison with the standard.
Why sulfur problem?: too high inputs?
unfavorable alkali / sulfur ratio?
because of kiln operation?
Detailed recording where the build-ups occur and possible a chemical analysis of a
typical build-up peace.
Specific measures against build-up formation (see section 7).
6.1 Material Balance

Definition of the balance boundary (Fig. 22):
Shall the external cycle be included or excluded?
Please note that the kiln feed includes the external cycle so that the concentration on
volatile elements is higher than the real inputs by the raw material. In this case the
kiln dust must be taken into account as further output!
Taking material samples over a sufficiently large period
The sampling period should be at least 8 hours of representative kiln operation.
From every input and output an hourly spot sample shall be taken. The spot samples
of the individual inputs and outputs are finally combined to one integrated sample.
The following quantities for the individual spot samples shall be taken:
raw mix (before grinding): 20.0 kg
meal or dust: 0.5 kg
clinker: 5.0 kg
fuel: 0.5 kg
Besides the samples for the material balance also samples of the hot meal at kiln inlet
(outlet of the lower most cyclone stage) shall be taken, observing the same frequency
and sample quantity as stated above.

Fig. 22 Material Balance Boundaries
6.2 Criteria and Indicators to Assess the Build-up Problem
6.2.1 Process Variables

Position of the build-ups
If the build-ups are found up to the second lowest cyclone stage or even to the third
lowest cyclone stage the problems could origin from excessive chloride cycles or in
rare cases from too high alkali cycles.
If the build-ups are found mainly at the lowest cyclone stage, the lowest riser dust
and the kiln inlet, the problems normally are caused by excessive sulfur cycles.
Completeness of combustion
CO at kiln inlet > 0.1% and/or O2 < 1.5 to 2% cause high sulfur cycles.
If no secondary firing or precalcination is applied also the CO and O2 concentrations
after preheater can be taken to judge the completeness of combustion (CO must be
< 0.1% and O2 usually 3 to 4%, depending on the inbleeding amount of false air).
Temperature profile
Excessive temperatures within and after the preheater indicate a disturbed
temperature profile in the rotary kiln and thus a too high sulfur volatilization.
Pressure profile
The pressure profile over the preheater indicates where the build-ups are located.
6.2.2 Material Balance

First a material balance must be performed as described in paragraph 6.1. Afterwards, the
following evaluation shall be done:
Input of circulating elements

It has to be determined in which form the sulfur is present in the raw materials, as
sulfate or sulfide. When present as sulfide, part of it will leave the kiln system as
emission via exhaust gas stack! Therefore the sulfur emission by the stack has to be
deducted from the overall sulfur when judging a potential sulfur problem in the kiln.
The inputs of circulating elements (without emitted sulfur through the stack!) should
be compared to the ones given below in order to check whether they are in a normal
range where usually no build-ups are formed or whether they are too high. This, of
course, is only a rough guideline.

Typical Input Limits for Circulating Elements (for 4 to 5 stage SP kiln only)
Raw Material
(loss free basis)
Cl < 0.02% Normal case, no problems
> 0.05% Heavy clogging problems,
depending on the sulfur cycle
SO3 1) < 0.5% Normal case, no problems

> 1.25% Heavy clogging problems
K2O < 1.0% Normal case, no problems

> 1.5% Problems with encrustations,
depending on degree of sulfatization
(molar alkali/sulfur ratio)
Na2O Little volatile, thus no problems due to recirculation of Na2O

1)
Sulfates: Sulfur in form of sulfides behaves in a different way,
see section 3.3.4 Emission by Exhaust Gas
Fuel(s)
Coal: S < 1.5%

No clogging problems,
(corresponding to
depending on sulfur and alkalis input by
Coke: S < 2% approx.
the raw materials and the completeness
< 5 g SO3/kg cli)
of combustion
Fuel oil: S < 2.5%
Coal: S > 3%
Severe clogging problems,
(corresponding to
depending on the alkalis and sulfur
Coke: S > 4% approx.
input by the raw materials and the
> 10 g SO3/kg cli)
completeness of combustion
Fuel oil: S > 5%
Gas: Sulfur content is normally zero!
Alkali / sulfur ratio (A/S)

The A/S ratio should be preferably at 1.2 or in a range between 0.8 to 1.5. Compare
it to the A/S ratio of the investigated case, deducting first the emitted sulfur from the
total sulfur input by the raw mix and the fuel.

6.2.3 Enrichment of the Circulating Elements in the Hot Meal
Fig. 23 illustrates the tolerable concentrations of circulating elements in the hot meal,
i.e. in the meal ex lowest cyclone stage.
Fig. 23 Concentrations in the Hot Meal
Normally no Increasing Severe build-up formation,

Sulfur encrustation build-up impeding of
problems formation continuous operation
2.5% 5% (SO3-%)
Normally no Increasing Severe build-up formation,

Alkalis encrustation build-up impeding of
problems formation continuous operation
2.5% 4% (Na2Oeq-%)
Increasing Severe build-up formation,

Chlorine build-up impeding of
formation continuous operation
0.8% 2% (Cl-%)
Fig. 24 considers the fact that normally both chlorides and sulfates are present in the hot
meal and that both together determine the tolerable limits of concentrations. The actual
case shall also be compared with this diagram.
Fig. 24 Maximum Tolerable Concentrations in the Hot Meal
% Cl
2.0 Fre
que
nt b
l ock
Zon age
e of i s
1.4 ncr
eas
ing
enc
1.0 rus
tati
on
0.8 pro
ble
No e ms
ncru
stati
ons
% SO3
0 1.0 2.0 3.0 4.0 5.0
2.5 3.75

6.2.4 Total Sulfur Volatility
With the formula for the total volatility (see section 3.4.1) the individual total volatilities of
the sulfates shall be calculated.
If of the sulfur is > 0.7 and if at the same time the A/S ratio is within the range of
0.8 to 1.2, the sulfur cycle is definitely too high and needs improvements by
measures as stated in section 7.4.
6.3 Example of a Circulation Phenomena Problem
6.3.1 Material Balance and Hot Meal Analysis
Material Balance
Inputs
Raw mix: SO3 = 0.5% cli
(loss free basis) K2O = 0.5% cli
Na2O = 0.2% cli
Cl = 0.05% cli
Coal: S = 2%, SO3 = 5%
Hu = 26800 kJ/kg, 6400 kcal/kg
Specific heat q = 3350 kJ/kg, 800 kcal/kg
consumption:
Outputs
Clinker: SO3 = 0.85%
K2O = 0.5%
Na2O = 0.2%
Cl = 0.03%
SO2 emission: SO2 = 400 mg/Nm3 (= 0.1% SO3)
Hot Meal Analysis
(loss free basis) SO3 = 4.5% cli
K2O = 4% cli
Na2O = 0.3% cli
Cl = 2% cli

6.3.2 Analysis of the Example
6.3.2.1 Material Balance
SO3 K2O Na2O Cl

Material Balance
[% cli] [% cli] [% cli] [% cli]
Input Raw Mix 0.5 0.5 0.2 0.05
Coal 0.6
Total 1.1 0.5 0.2 0.05
Output Clinker 0.85 0.5 0.2 0.03
SO2 emission 0.1
Total 0.95 0.5 0.2 0.03
Balance Error 0.15 0.0 0.0 0.02
The total sulfur output is smaller than the total input. This often is the case when
performing a material balance. There are two possible explanations:
The samples are not representative.
The process was not in a steady state. Sulfur was kept back in the system, forming
encrustations. During cleaning, the sulfur containing deposits left the kiln with the
clinker but were not catched representatively by the sampling procedure.
The total chloride output is smaller than the total input. Here the same problems arise as
described above. Normally the clinker does not contain more than 0.01% Cl due to the
high chloride volatility. However, if the chloride cycle becomes very large, values up to
0.05% Cl in the clinker have already been measured.
6.3.2.2 Form of Sulfur in the Raw Materials

Note that the raw material contains sulfidic sulfur (FeS2) due to the SO2 emission. The real
sulfate input to the kiln is therefore:
from raw mix to kiln 0.5 to 0.1 = 0.4% SO3
coal + 0.6% SO3
total input to the kiln 1.0% SO3

6.3.2.3 Comparison with Typical Inputs of Circulating Elements (see section 6.2.2)
Inputs by the raw materials:

SO3 = 0.4% (sulfates only!) o.k.
K2O and Na2O o.k.
Cl = 0.05% very high, bypass required
Sulfur inputs by the coal:

S = 2% high, but within the normal range
6.3.2.4 Alkali / Sulfur Ratio

%K 2O %Na2O %Cl 0.5 0.2 0.05
+ + +
A / S = 94 62 71 = 94 62 71 = 0.63
%SO3 1.0 *)
80 80
*)
total input to the kiln (see section 6.3.2.2)
The A/S ratio is outside of the desirable range of 0.8 to 1.5.
6.3.2.5 Enrichment of Circulating Elements in the Hot Meal
Comparison with tolerable concentrations of circulating elements (see section 6.2.2)
SO3 = 4.5% High, far beyond normal limit, problems
with encrustation
Alkalis: Slightly above normal, but highly desirable
0.66 4 + 0.3 =2.9% due to the large sulfur input!
Cl = 2% High, far beyond normal limit, problems
with encrustations
Comparison with diagram for maximum concentrations (see section 6.2.2)
SO3 = 4.5%
frequent blockages to be expected
Cl = 2%
6.3.2.6 Total Sulfur Volatility (SO3)

ccli 0.85
= 1 = 1 = 0.81
cHM 4 .5
> 0.7: Because of the low A/S ratio, a substantial portion of the sulfur is in form of
CaSO4 which leads to the high volatility of > 0.7.

7. MEASURES AGAINST BUILD-UP FORMATION
7.1 General Measures

Reduction of the Inputs of Circulating Elements
The most obvious measure against build-up formation is to reduce the input of
circulating elements. It is normally not possible to change the main raw materials.
However, sometimes a minor component that contains a substantial amount of
circulating elements can be replaced. In most cases the sulfur content of the fuel is
more easy to change. Please note that the fuel may also contain considerable
amounts of chlorides.
The circulating elements of the raw materials in the quarry deposits are often not
homogeneously distributed. In such cases prehomogenization of the raw materials
combined with selective quarrying helps to reduce peak inputs of circulating
elements into the kiln system.
Smooth Kiln Operation
In many cases the cycles of circulating elements are frequently increased by an
inadequate kiln operation applying to drastic changes in fuel feed and draft.
Also the best kiln operator cannot apply during his daily work of 8 h such a super
constant kiln operation. Therefore the author is of the strong opinion that only a fully
automatic kiln control will provide optimal results (cp. LINKman).
7.2 Intelligent Cleaning

Important: The more circulating elements are introduced into the kiln system the better and
more efficient cleaning methods have to be applied!
Preheaters have to be controlled and cleaned at least once a shift to remove
immediately possible build-ups! Therefore a experienced tower guard is needed.
On each platform enough cleaning slots are required to manually clean the riser duct and
the inlet chamber whenever it is necessary. At particularly critical spots additional slots
should be installed.
Critical locations where always material deposits are built shall be armed with air
blasters. Please note that mostly several blasters for one location must be installed since
their cleaning radius is rather small (< 0.5 m). The blasters should always shoot in the
direction of material flow (see Fig. 25).

Fig. 25 Application of Air Cannons

The riser ducts shall be protected by introducing the raw meal from the upper stage as
low as possible (Fig. 26). That way the circulating elements are able to condense at the
cooler meal.
Fig. 26 Protection of the Riser Duct Walls against Condensation of Circulating

Elements

The meal ducts should be equipped with pendulum flaps in order to avoid a hot gas bypass
through the ducts into the cyclones (Fig. 27).
Fig. 27 Example of a Pendulum Flape for the hot meal duct of the lowest
cyclone in a suspension preheater (Polysius)
Meal ducts have to have a sufficient inclination (> 55), a sufficient large diameter and
must not have any sharp bends.
The cyclone outlets for the hot meal should be sufficiently large. If frequent bridging of
the outflowing raw meal occurs, the outlet section shall be first equipped with air blasters
and, if this measure does not help, it must be enlarged.
The installation of coating neutral refractory at the transition chamber and the lowest
riser duct has produced good results in many applications and is state of the art.
Any false air entrance at the kiln inlet area and the lower preheater part must be avoided
(continuous supervision and maintenance). Otherwise false air is likely to increase the
formation of build-ups.
For efficient cleaning special high-pressure water pumps (e.g. type WOMA) shall be
applied. The application of this pump is dangerous and needs special training of the
operators. There is also a danger of destroying the refractory lining!

Adequate meal distribution boxes shall be installed, which distribute the raw meal over
the whole cross section. An example is given in Fig. 28.
Fig. 28 Meal Distribution Boxes by Polysius
Dead corners in riser ducts, cyclones and the transition chamber shall be avoided.
Cleaning should only be done if the pressure profile or a quick inspection indicate a need
for. Opening of the big access doors disturbs kiln operation and let enter cold air, which
favors build-up formation.

7.3 Measures against Chloride Problems
Important: With the clinker chlorine can only be withdrawn to a very limited amount (0.01 to
maximum 0.02% Cl). Therefore the measures against chlorine build-ups are
limited.
Reduction of the Volatility
There is little chance to decrease the chlorine and chlorine volatility as evaporation
takes place at a low temperature range (800 to 1000C).
Discarding of Dust (external cycle)
For long dry and wet kilns as well as for Lepol kilns the chlorides are effectively
withdrawn from the kiln by discarding the finest fraction of the total of the kiln dust.
For preheater kiln dust discarding is normally not very efficient as the accumulation of
chlorides in the dust is small. However, in case of a large internal chloride cycle this
measure could help to keep it at a lower level. The effect of the measure can be
calculated by a material balance.
Reduction of the Sulfur Cycle
Build-up formation is provoked by all circulating elements together. As the chloride
cycle cannot be effectively influenced, all possible measures should be taken to
reduce the sulfur cycle as it increases the tolerance threshold for chlorine.
Installation of a Kiln Gas Bypass
If the chloride input of a SP kiln is larger than 0.02 to 0.03% (loss free basis) a
chloride bypass must be installed. Whether the higher limit of 0.03% can be tolerated
depends on the expected sulfur cycle.
A rule of thumb to calculate the quantity of required bypass gases is:
% Bypass = % Cl input by the raw material (loss free) x 100
Example: % Cl input by the raw material = 0.05%

Required bypass size = 0.05 100 = 5%
The Bypass has to be installed at point of highest concentration in the gas of chlorine
and alkalis.
7.4 Measures against Sulfur Problems

Important: In contrast to chlorine sulfur can be withdrawn in high quantities with the clinker.
All measure aim to reduce the volatility in order to introduce the sulfur into the
clinker.
Reduction of the sulfur input (raw material and fuel)
Adjusting of the molar alkali / sulfur ratio between 0.8 and 1.2; best to a value beyond 1.
K2SO4 is little volatile, CaSO4 is highly volatile and can be withdrawn only in form of
double-salts or within the belite.
Keep the lime saturation factor as low as possible so that the sulfur can dissolve in the
belite.
Possibly modification of the raw mix in order to improve the burnability and the
granulation of the clinker (decrease of the silica ratio). These measures help to reduce
the required maximum temperature in the burning zone and to reduce the sulfur
volatilization from the clinker granules.
Minimal fluctuations in the chemistry and the quantity of the kiln feed so that constant
burning conditions can be maintained.

Avoid overburning! High sulfur content means a porous clinker. In this case not a high
litre weight is needed to produce low free lime! For quality control drop litre weight or
adjust regularly the rated value for the litre weight to the free lime values.
High sulfur cycles produce a dusty kiln atmosphere. Do not burn down the dust at all
costs in order to avoid overheating.
Sufficient excess air at the kiln inlet to avoid reducing conditions (2% O2). Please note
that the measurement of O2 at kiln inlet is a spot sample and does not necessarily
represent the whole inlet cross section. The 2% O2 can therefore be a guide value only.
Attention: Too much excess air produces high kiln inlet temperatures, which again
increase the sulfur volatility. Use carefully excess air! Do not tolerate any CO
at the kiln inlet, i.e. CO < 0.05%.
Introduce automatic kiln control (LINKman) in order to stabalize the kiln operation.
The burner should be directed parallel to the kiln axis and should not point to the kiln
charge in order to avoid local reducing burning conditions.
The burner itself should be of the latest design, which produces a short and stable flame
in order to have the shortest possible retention time of the kiln charge in the high
temperature zone and low kiln inlet temperatures are achieved. As a result the sulfur
volatility is minimized.
Apply secondary firing or precalcination to reduce the thermal load in the sintering zone.
This possibility is very limited when the kiln inlet temperatures are too high and reducing
conditions occur. In this case heavy sulfur build-ups or even build-ups from the first
clinker mineralization result.
Very good dispersion of all liquid fuels in order to achieve a short flame (enough viscosity
and atomizing pressure).
Solid fuels (also alternative fuels) should be enough prepared. The main fuel at the main
burner should have the fineness of coal! Use coarser fractions only limited in the main
burner (10 to 20% heat). Coarse particles make the flame longer and so increase the
sulfur volatility. Introduce additional coarse fuel fractions separate above the flame and
not directly into the flame.
If solid fuels (e.g. used tires) are burnt at the kiln inlet, lifters shall be installed in order to
keep the fuel pieces at the surface of the kiln charge and to avoid local reducing burning
conditions.
Permanent control of pressure fluctuations in the pneumatic pipes at the burner head.
Pulsation < 5 mbar! Avoid long transport pipe and keep the velocity > 30 m/s.
In case of extremely high sulfur inputs and a alkali deficiency at the same time kiln gas
bypasses are applied. But these are by far not as efficient as for chlorine problems.
Remark: Especially for long dry and wet kilns discarding of dust is a very effective
measure to withdraw the sulfur from the kiln. This does not hold true for SP
kilns as the accumulation of sulfur in the external cycle is small.
For SP kilns the installation of a kiln gas bypass may become necessary if
there are almost no alkalis to extract the sulfur from the kiln.

Shifting of the condensation area of the sulfur towards the kiln inlet and the first zone of
the rotary part.
This effect is achieved by increasing the A/S ratio and by avoiding high kiln
inlet temperatures.
Avoid overburning of the clinker, Fig. 29 demonstrates the strong influence of burning
zone temperature (indicated by the concentration of NOx on the accumulation of SO3 and
especially CaSO4 in the hot meal. Furthermore also the influence of the excess air is
shown.
Fig. 29 Enrichment of SO3 in the Hot Meal of the Lowest Cyclone Stage

7.5 Measures against Alkali Problems
Alkali Volatility
In absence of sulfur the alkali volatility is very high and creates problems in the
preheater. In such a case the raw mix can be sulfatizated by addition of gypsum.
Low Alkali Clinker
If low alkali clinker must be produced, all measures must be taken to increase the
alkali volatility, such as
reducing the sulfur input
producing a long and stable flame
applying hard burning (CaOfree < 1), if possible reducing the burnability by
increasing the silica ration
applying a minimum of excess air
chlorination of the raw mix either by burning chloride
containing solvents or adding CaCl2
In case of long dry or wet kilns the alkalis are withdrawn by discarding a fraction or the
total dust of the external cycle. In case of a SP kiln a (large) kiln gas bypass is required.

8. MATHEMATICAL MODEL TO SIMULATE THE CYCLES OF THE CIRCULATING
ELEMENTS
The cycles of the circulating elements in the kiln system can be simulated by a mathematical
model (Fig. 30). This model is used for bypass calculations and for simulating special effects
on the behavior of the volatile cycles.
Fig. 30 Mathematical Model to Simulate the Cycles of the Circulating Elements

in the Kiln System (by Weber)
Clinker Alkali from fuel
Ak
(1- )
1
1
(K-1) 2
K (K-1) + + A
2 1 k
(K-1)
[ (K-1) + +A
2 1 k ]V
Kiln feed Alkali valve (V)

Chapter 9
Kiln Control

Kiln Control and Operation
Stefan Gross
TPT 02/21185/E
1. INTRODUCTION ............................................................................................................. 570

2. CONTROL PARAMETERS............................................................................................. 571
2.1 Location of the control parameters and variables ............................................. 571
2.1.1 Wet kiln.......................................................................................................... 571
2.1.2 Lepol kiln ....................................................................................................... 572
2.1.3 Preheater kiln ................................................................................................ 572
2.1.4 Precalciner kiln .............................................................................................. 573
2.2 Burning Zone Temperature (BZT)..................................................................... 573
2.2.1 Pyrometer reading ......................................................................................... 574
2.2.2 NOx of the exhaust gases ............................................................................. 574
2.2.3 Kiln torque ..................................................................................................... 575
2.2.4 Secondary Air Temperature (SAT) or Tertiary Air Temperature (TAT).......... 576
2.2.5 Position of the dark feed (valid only for pure natural gas flames).................. 576
2.3 Back End Temperature (BET).......................................................................... 576
2.4 Oxygen control .................................................................................................. 577
3. CONTROL VARIABLES ................................................................................................. 578
3.1 Fuel rate to the kiln............................................................................................ 579
3.2 Feed rate to the kiln .......................................................................................... 579
3.3 Kiln draft ............................................................................................................ 579
3.4 Kiln speed ......................................................................................................... 580
4. CONTROL STRATEGY .................................................................................................. 580
4.1 General remarks ............................................................................................... 580
4.2 Possible kiln parameter conditions.................................................................... 581
4.3 Decision tree ..................................................................................................... 582
4.4 Counteractions .................................................................................................. 582
5. PRECALCINER CONTROL ............................................................................................ 587
5.1 Calcination temperature .................................................................................... 587
5.2 Oxygen control .................................................................................................. 588
5.3 Other control parameters .................................................................................. 588
6. GRATE COOLER CONTROL ......................................................................................... 589
6.1 Bed height and grate speed .............................................................................. 589

6.2 Subordinate control loops ................................................................................. 590
6.2.1 Cooler air flow................................................................................................ 590
6.2.2 Kiln hood pressure......................................................................................... 591
7. LEPOL GRATE CONTROL ............................................................................................ 592
7.1 Subordinary control loops ................................................................................. 592
7.1.1 Pressure control ............................................................................................ 592
7.1.2 Traveling grate speed control ........................................................................ 593
7.1.3 Pelletizer control ............................................................................................ 593
8. PROCESS PROBLEMS AND KILN UPSET CONDITIONS ........................................... 595
8.1 Cycling .............................................................................................................. 595
8.2 Coating Collapse............................................................................................... 596
8.3 Ring break out................................................................................................... 597
8.3.1 Clinker ring break out .................................................................................... 597
8.3.2 Sinter, middle, meal and mud rings ............................................................... 597
8.4 Hot meal rush (avalanche) ................................................................................ 598
8.5 Red spot on the kiln shell .................................................................................. 598
8.6 Loss of kiln feed ................................................................................................ 599
9. STARTUP AND SHUTDOWN PROCEDURES .............................................................. 600
9.1 Start up.............................................................................................................. 600
9.2 Shut down ......................................................................................................... 602
10. GENERAL CONSIDERATIONS .............................................................................. 603
11. CALCULATION OF THE RETENTION TIME .......................................................... 605

SUMMARY
Control Parameters
The most important control parameters in any type of kiln system are:
Burning Zone Temperature (as an indicator for the produced clinker quality)
Back End Temperature (as an indicator for the raw meal preparation, which is mainly
drying, heat up and calcination)
Oxygen concentration of the kiln exit gas (as an indicator for the efficiency of
combustion)
In case of precalciner systems: Temperature and oxygen concentration at precalciner
outlet
In case of a grate coolers: Bed height of the clinker layer on the grate
Control variables
The most important control variables that are adjusted to maintain the previous control
parameters in a certain allowed range are:
Fuel rate to the kiln
Material feed rate to the kiln
Draft in the kiln
Kiln speed
In case of precalciner systems: Fuel rate and draft in the precalciner
In case of a grate coolers: Grate speed
Objectives of kiln operation

The operation of a kiln system has to meet the following basic requirements:
Protection of equipment and personnel at all times
Consistently good clinker quality
Smooth and stable operation
Maximum thermal efficiency
Maximum production rate

1. INTRODUCTION
The different kiln systems can be distinguished by the moisture content of the raw material,
which is fed to the kiln system. The different process types are:
Wet process 30-40 % Moisture
Semi wet process 17-21 % Moisture
Semi dry process 10-15 % Moisture
Dry process < 1 % Moisture
Although each type of process has its own particular operation parameters, the raw meal
has to undergo similar chemical and physical reactions until the clinker finally leaves the kiln
system.
The basic reactions are:
Drying of the raw meal
Heat up of the material
Calcination
Formation of the clinker minerals
Cooling of the clinker
The main difference between the different types of process lies in the required time and
energy, each reaction needs to be completed.
Therefore the operating range of the different control parameters is varying from one kiln
system to the other. The same applies to the response time as well as to the frequency and
magnitude of any control variable adjustment.
Nevertheless certain basic rules are applicable for all types of kiln systems. They are
presented in the following and shall give an elementary concept of kiln operation.

2. CONTROL PARAMETERS
Kiln systems normally are equipped with a great number of instruments, sensors and
measuring devices.
Some of them are for information purpose only (e.g. if a flap is opened or closed), others to
detect dangerous situations (e.g. cyclone blockages), but only few of them are continuously
used for kiln operation.
Those measured values that are the most important for kiln operation are herein called the
control parameters. To keep the control parameters close to a certain target control
variables (or actuators) are manipulated.
The most important control parameters that are looked at during kiln operation are:
Burning Zone Temperature (as an indicator for the produced clinker quality)
Back End Temperature (as an indicator for the raw meal preparation, which is mainly
dryout, heat up and calcination)
Oxygen concentration of the combustion gas at the kiln inlet or the preheater (as an
indicator for the efficiency of combustion and the gas atmosphere inside the kiln)
In case of precalciner systems: Temperature and oxygen concentration of the
exhaust gas at the precalciner outlet
In case of a grate cooler: Bed height of the clinker bed
2.1 Location of the control parameters and variables

The location where the different control parameters are measured and some typical target
values are given below. The control variables, which are adjusted to keep the parameter
close to target, are shown also for the different types of kiln systems.
2.1.1 Wet kiln
Oxygen
(for coal)
BZT BET 0.5 - 1%
~1450 C ~800 C
Fuel rate Kiln speed Feed rate Kiln draft
Figure 1: Control parameters and variables of a wet kiln

2.1.2 Lepol kiln
Oxygen
(for coal)
BET 1.5 - 2.5 %
BZT ~1050 C
~1450 C
Fuel rate Kiln Feed rate Kiln

(H2O/RM) draft
Figure 2: Control parameters and variables of a Lepol kiln
2.1.3 Preheater kiln
Oxygen
BET (for coal)
BZT
1.5 - 2.5%
1450 C ~1050 C
Fuel rate Kiln speed Kiln draft Feed rate
Figure 3: Control parameters and variables of a PH kiln

2.1.4 Precalciner kiln
Oxygen after
PC 1 - 1.5 %
Oxygen at
KI (for coal)
BZT BET 2.5 - 3.5%
1450 C ~1150 C
Kiln draft
Fuel rate Kiln speed and PC draft Feed rate
Figure 4: Control parameters and variables of a precalciner kiln
2.2 Burning Zone Temperature (BZT)

The BZT determines, how well the clinker is burned and how complete the transformation
from C2S to C3S is. When the clinker is leaving the burning zone it should not have more
than ~1.5% and not less than ~0.5% of uncombined CaO or free lime (which roughly
corresponds to a liter weight of about 1250-1350 g/l).
The periodically measured free CaO (or the liter weight) is the main indicator to state, if the
actual BZT target is adequate or if it has to be increased or lowered.
Due to changes of the raw meal composition an alteration of the BZT target can be required
now and then, since burnability of the raw material changes with fluctuations of the LSF, SR,
magnesia and alkali content among other factors. A change in the flame characteristics,
which can result from the use of multiple fuels or varying combustion conditions has a similar
impact and therefore may also require an alteration of the BZT target.
Variations of the BZT target are made only over a long-term period, in the same way as the
free CaO (or liter weight) is measured normally only on a one or two hourly basis.
The actual BZT must be measured to determine, if a deviation between the temporary target
and the actual BZT exists.
Since kiln operation is a dynamic process, deviations are quite frequent and of a short-term
character.
Several indicators are used to determine the BZT and they are listed below.

Most do not explicitly express an absolute temperature but they give sufficient information to
establish, if the BZT increased or decreased compared to an earlier defined target. This
means they give the short-term trend of the control parameter BZT. This helps to estimate
the probable free CaO (or the liter weight) and must bridge the time until the actual clinker
analysis is available.
2.2.1 Pyrometer reading

The radiation of clinker and coating in the burning zone has a relation to its temperature.
Therefore the BZT can be determined by measuring the intensity or the color of this
radiation.
A rough temperature color correlation is given below (visible range):
Lowest visible red to dark red 475-650 C

Dark red to cherry red 650-750 C
Cherry red to bright red 750-850 C
Bright red to orange 850-900 C
Orange to yellow 900-1100 C
Yellow to light yellow 1100-1350 C
Light yellow to white 1350-1550 C
By measuring the color of the burning zone and not only the radiation intensity, the reading
is less influenced by disturbances like the dust load of the secondary air and others.
Therefore modern pyrometers and kiln cameras mainly measure the color and not the
intensity of radiation.
2.2.2 NOx of the exhaust gases

The NOx that is produced during combustion varies according the temperature of the flame
and of the gases of the burning zone. The hotter the flame is the more NOx is produced
during combustion and vice versa.
When BZT is deduced from the NOx content of the exhaust gases, the oxygen and CO
concentration in the exhaust gases have to be considered also, since CO reduces and
oxygen promotes the formation of NOx.
A further factor influencing NOx formation is the type of fuel burned. If the fuel mix changes,
the quantity of NOx produced may change too (especially when burning alternative fuels).

2.2.3 Kiln torque
The burning zone shifts upwards and gets longer, if the BZT rises.
A higher BZT produces more liquid phase in the clinker.
The diameter of the clinker balls increases and the granulation of the raw meal in the kiln
starts earlier. Therefore more and bigger clinker balls have to be moved inside the kiln when
the BZT is increasing.
At the same time, the material bed in the kiln is lifted up more, which moves the center of
gravity away from the kiln axis.
The different granulation of the clinker and the changing angle of repose increase the
required torque to turn the kiln, if BZT rises.
Figure 5: Torque of a hot and a cold kiln
This kiln torque is proportional to the current of a DC drive and proportional to the quotient
kiln power / kiln speed in case of an AC - drive or a hydraulic drive.
A changing coating situation in the burning zone strongly influences the kiln torque (as more
as lower the L/D ratio of the kiln). As a result the kiln torque can change significantly on a
long-term basis without any change of the BZT. Therefore the kiln torque signal in most
cases only serves as a short-term indication of the BZT. Short-term is defined as up to 12
hours. Torque is a very useful indicator of kiln conditions relative to its trend.

2.2.4 Secondary Air Temperature (SAT) or Tertiary Air Temperature (TAT)
Kilns with a grate cooler, which are equipped with a reliable SAT or TAT measurement
device, often use SAT or TAT as an indicator for the BZT.
The higher the BZT, the hotter the released clinker and the higher SAT or TAT (and vice
versa).
Nevertheless SAT and TAT are influenced by a number of other effects (clinker
granulometry, cooler operation, etc.), therefore not in all kilns can a clear correlation
between BZT and SAT or TAT be given.
2.2.5 Position of the dark feed (valid only for pure natural gas flames)
If the visibility inside the kiln does permit (little dusty, flame shape, etc.), the calcined hot
meal can be seen as a dark flush, when entering the burning zone. An area can be
determined, where this meal, floating like water starts to form liquid phases and where it
transforms to solid clinker balls.
The position of the area where this transformation takes place is an additional indicator of
the BZT.
If this area shifts downwards, the BZT is falling and vice versa.
Figure 6: Position of the dark feed
2.3 Back End Temperature (BET)

The term BET refers to the temperature of the exhaust gas at the kiln inlet in case of kilns
with a preheater and to the intermediate gas temperature (before the chain section), if the
kiln is equipped with chains. The BET depends on the type of system and the L/D ratio of

the kiln. It ranges from ~800 C in wet kilns (before the chain section) up to ~1200 C in AS
precalciner kilns.
If a reliable temperature measurement at the kiln inlet (kilns with preheaters) or before the
chain section (kilns with chains) is not available, frequently the preheater exit temperature
(kilns with preheaters) or the kiln inlet temperature (kilns with chains) is used.
The BET is an indicator of how well the hot meal is prepared, when entering the kiln (or
when passing the intermediate measuring point, in case of kiln systems with chains).
Before the meal is finally clinkerized in the burning zone, it has to undergo dryout,
calcination (release of CO2) and to a minor extend also reactions like structural
modifications and the formation of intermediate products like C12A7 or CA).
The higher the energy demand for intermediate reactions in the kiln, (which depends mainly
from the type of process), the more important is a proper control of the BET.
Hence the BET in a wet kiln is more critical (90-100% calcination missing) than in an air
separate precalciner kiln (~10% calcination missing).
The BET reacts as slower to any control variable adjustment as longer the kiln (higher L/D
ratio) and as more energy is consumed in the kiln for the intermediate reactions before
clinkerisation.
Ideally the material should always have the same degree of preparation, but due to the kiln
operation itself and normal fluctuations of the raw material properties, fuels, ambient
conditions and others, this is not the case.
Variations in the raw material are first observed through the BET, thereby indicating early the
future impact on the BZT.
If all other parameters remain constant and the BET drops, this is evidence that all reaction
zones shift downward and vice versa.
If the zones within the kiln shift either down or up, the BZT will decrease or increase
consequently afterwards also, if no counter measures are taken earlier.
2.4 Oxygen control

For a good combustion of the fuels a certain amount of excess air is required.
The quantity of excess air is determined by measuring the oxygen content of the exhaust
gases. Proper control of the oxygen level is crucial for good combustion and a high thermal
efficiency.
Typical oxygen ranges (at the kiln inlet) for the most common kiln systems are given in the
following table. The required oxygen level depends on the type of kiln system but also
strongly on the type of fuel being used (e.g. gas needs less excess air than coal or solid
AFR).
A further important influence to the oxygen level is the amount of sulfur, chlorine and alkali
within the kiln system. Therefore the higher oxygen levels within the given ranges are
chosen, if sulfur volatilization is a problem.

Gas Fuel oil Coal Petcoke
Solid AFR
Wet kiln 0.5 - 1.0 % 1.0 - 2.0 % 1.5 - 2.0 % 2.0 - 2.5 %
Lepol kiln 1.0 - 1.5 % 1.5 - 2.0 % 1.5 - 2.5 % 2.0 - 3.0 %
Preheater kiln 1.0 - 1.5 % 1.5 - 2.0 % 1.5 - 2.5 % 2.0 - 3.0 %
Inline PC kiln 2.5 - 3.0 % 2.5 - 3.0 % 2.5 - 3.5 % 2.5 - 4.0 %
Separate PC kiln 1.0 - 1.5 % 1.5 - 2.0 % 1.5 - 2.5 % 2.0 - 3.0 %
Table 1: Typical oxygen levels at the kiln inlet (dry gas)
In most kiln systems the required amount of excess air is slightly above the point where the
flame temperature would be highest due to sulfur, chlorine or alkali situation.
Therefore a slight reduction of the amount of excess air increases the flame temperature a
little.
Extremely low excess air (or oxygen) reduces the flame temperature, which lengthens the
flame (postcombustion) and lowers the fuel efficiency. Additionally sulfur and alkali
volatilization is enhanced, which may lead to build up and incrustation problems within the
kiln system.
Too much excess air (or oxygen) cools the flame too and leads also to a poor thermal fuel
efficiency due to the deteriorating heat transfer. Since the total gas flow is increased, the kiln
internal and external dust cycles also increase.
Hence lower oxygen levels are chosen, if sulfur volatilization is not problematic.
As a general rule applies, that at the minimum permissible oxygen level no notable CO is
formed (i.e. < 0.05 % at kiln inlet).
3. CONTROL VARIABLES
In the proceeding paragraph the main control parameters used for kiln control were
presented.
Those parameters must be kept in a narrow range closed to target, in order to run the kiln in
a smooth and efficient manner. Generally speaking the smaller the deviations of the
parameter from the target, the better the kiln control.
To keep the control parameters close to target it is generally necessary to make adjustments
to the control variables.
The main control variables are (excluding precalciner systems):

Fuel rate to the kiln
Material feed rate to the kiln
Kiln draft
Kiln speed
In the following the effect of any adjustment made to one of the four kiln control variables is
shown. To keep the normally relatively complex relations simple it is assumed, that the

adjustment is done only to one variable at a time (the others remain untouched) and that the
adjustment is so small, that none of the control parameters deviates far from target.
3.1 Fuel rate to the kiln

When increasing the fuel rate into the kiln:
BZT increases
BET increases
Oxygen concentration decreases
(A decrease of the fuel rate has the inverse effect).
More energy is introduced into the system, therefore the BZT and the BET rises. Oxygen
level drops, since part of the excess air is consumed for the combustion of the extra fuel.
Of course, the above applies only, if the combustion does not takes place under reducing
conditions.
3.2 Feed rate to the kiln

When increasing the material feed rate into the kiln:
BZT decreases
BET decreases
Oxygen concentration decreases
(A decrease of the feed rate has an inverse effect).
When more feed is introduced into the kiln system the temperature starts to drop first at the
back end. After the higher amount of feed has reached the burning zone, the BZT starts to
decrease also.
The extra feed releases additional H2O and CO2. The additional amount of gas from the
meal reduces the quantity of combustion air (and thus of excess air), since the amount of
exhaust gas remains (almost) constant. This leads to a decrease of oxygen in the exhaust
gas.
3.3 Kiln draft

When increasing the draft in the kiln:
BZT decreases
BET increases
Oxygen concentration increases
(A slight decrease of the draft has an inverse effect. If combustion takes place under
reducing conditions (too little excess air) the BZT drops again and BET rises (see chapter
2.4).
When the kiln draft rises, more (and slightly colder) secondary air is pulled into the kiln and
through the burning zone. This decreases the flame temperature and as a result the BZT
drops.
Although the amount of heat supplied to the kiln system is somewhat higher (more
secondary air), less heat is released in the burning zone (lower flame temperature results in
reduced heat transfer) thus the BET rises.

The higher kiln draft increases both, the oxygen concentration and the volume of the
exhaust gases.
3.4 Kiln speed

When increasing the speed of the kiln:
BZT decreases (temporarily)
BET decreases (temporarily)
Oxygen decreases (temporarily)
(A decrease of the kiln speed has the inverse effect)
When the kiln speed is increased, the meal advancement is faster. All reaction zones shift
temporarily downwards, which causes the BET and the BZT to drop.
Since more meal enters temporarily the drying and the calcining zone (same effect as an
increased feed rate), a higher amount of H2O and CO2 is released which leads to a drop in
the oxygen concentration.
Those effects are only temporary, since a higher kiln speed at a fixed feed rate reduces the
filling degree in the kiln. Once the kiln filling degree has been lowered in the entire kiln (all
material, that was in the kiln, when the speed change was made, has left the kiln), BZT, BET
and oxygen recover.
Remark:
In almost all kiln operation strategies the kiln speed depends on the kiln feed rate. The
objective is to keep the filling degree in the kiln constant. Thus the kiln speed is under
normal operation conditions is not an independent variable, but adjusted in tandem with the
kiln feed rate.
In certain upset conditions (especially the dreaded hot meal rush), the kiln speed has to be
reduced that much that a parallel reduction of the kiln feed would produce an overheating of
the back end. In these situations, a compromise between increased filling degree and BET
must be made (under this situation, the so-called cycling can occur).
4. CONTROL STRATEGY
4.1 General remarks

As seen before an adjustment to only one of the control variables has an impact on all three
control parameters.
To keep all control parameters within a certain range close to target it is normally necessary
to make adjustments to more than one control variable.
But in doing so, the correction of one particular control parameter can increase the deviation
of an other.
As a result a compromise must be found that reduces the weighted average of all
parameter deviations to a minimum.
The weight (degree of importance) of each particular control parameter depends on the type
of system and the particular kiln. In a precalciner kiln for example the BET is less critical
than in a wet kiln. Therefore the influence of the BET is less weighted in a precalciner kiln
than in a wet kiln.

The selection of the control variables used to bring one or more control parameter that are
out of range back to normal depends on the overall view of all control parameters.
In the following an example of a kiln operation strategy is presented, which describes the
most typical actions, which may be taken for all types of kiln systems. Certain modifications
of this strategy may be necessary in particular cases, since each kiln behaves differently and
has dissimilar preconditions.
A very important factor herein plays the ability to make positive changes to the control
variables. In many kilns a bottleneck exists, that does not allow any positive change to one
of the control variables, once nominal capacity has been reached (e.g. an ID fan that runs
already with maximum speed or with the damper completely opened). This consequently
may require an alteration of the presented strategy.
The purpose of this strategy is to understand the principles of kiln operation and may not be
directly applicable on all kiln systems.
4.2 Possible kiln parameter conditions

It is assumed that the three kiln parameters BZT, BET and Oxygen at kiln inlet can be
attributed to one of the following states (although transitions are fluid), which are:
Inside the desired range, which means o.k.

Above the desired range, which means high
Below the desired range, which means low
The desired range for the three control parameters typically is as followed:
BZT BET Oxygen

Wet kiln Target +/- 20 C Target +/- 10 C Target +/- 0.3 %
Lepol kiln Target +/- 20 C Target +/- 15 C Target +/- 0.5 %
Preheater kiln Target +/- 20 C Target +/- 15 C Target +/- 0.5 %
Inline PC kiln Target +/- 20 C Target +/- 20 C Target +/- 0.7 %
Separate PC kiln Target +/- 20 C Target +/- 20 C Target +/- 0.5 %
Table 2: Desired range of the different control parameters
Combining the three control parameters and their respective state in any combination, leads
to 3^3=27 possible conditions. These 27 basic conditions are shown graphically below in the
so-called decision tree.

4.3 Decision tree
Based on the 3 possible conditions of the 3 control parameters, the following decision tree is
built, which covers all 27 possible states.
Figure 7: Decision tree
4.4 Counteractions
For each of the possible 27 states a particular action (control variable adjustment) is
described, to bring the control parameters back to target, which is the desired state 14.
The counteractions apply under the assumption that no bottleneck exists.
Although it is not always explicitly mentioned, the magnitude of a possible deviation plays a
great role, since it influences the intensity of counteraction. Any adjustment to the control
variables must be more or less pronounced, according the degree of parameter deviation
from the target. Otherwise the correction of the kiln parameters is either insufficient or too
strong, which results in a weak response or in an overshoot of the parameter reaction
(creating a cycling effect).
In kiln systems, where the BET is of minor importance, mostly the BZT and the oxygen
establish the actions. This means a certain deviation from the BET target is temporarily
accepted, if this helps to maintain a high productivity and as long as the overall kiln condition
does not suffer. If allowed to continue for prolonged periods, elevated BET may result in
increased condensation of circulating elements in difficult to reach areas. This can cause
plugging and possible kiln stoppages and must be avoided.
The actions to be usually taken are (the actions to be taken lay more emphasis on a stable
kiln operation than on maximum productivity):

Case Condition Action Reason
1 BZT low When BZT is slightly low:

OXYGEN low 1 increase kiln draft increase oxygen and prepare for step 2, increase BET
BET low 2 increase fuel rate increase BZT and BET, bring oxygen back to normal
When BZT is very low:
1 increase kiln draft increase oxygen and prepare for step 2, increase BET
2 increase fuel rate increase BZT and BET, bring oxygen back to normal
3 decrease kiln speed increase BZT and BET even more
4 decrease feed rate maintain kiln filling degree constant
2 BZT low 1 increase kiln draft slightly increase oxygen, increase BET to prepare for step 2
OXYGEN low 2 decrease fuel rate increase oxygen even more, bring BET back to normal
BET o.k. 3 decrease kiln speed increase BZT
3 BZT low 1 decrease fuel rate increase oxygen and prepare for step 2, decrease BET
OXYGEN Low 2 decrease kiln draft decrease BET
BET High 3 decrease kiln speed increase BZT
4 BZT Low When BZT is slightly low:

OXYGEN o.k. 1 increase kiln draft increase BET, increase oxygen to prepare for step 2
BET Low 2 increase fuel rate increase BZT and BET
1 increase kiln draft increase BET, increase oxygen to prepare for step 2
2 increase fuel rate increase BZT and BET
3 decrease kiln speed increase BZT and BET even more
4 decrease feed rate maintain kiln filling degree constant and increase BET
5 BZT Low When BZT slightly low and oxygen slightly high, but still o.k.:
OXYGEN o.k. 1 increase fuel rate increase BZT
BET o.k. When BZT very low:
1 increase kiln draft slightly increase oxygen and prepare for step 2
2 increase fuel rate slightly increase BZT
3 decrease kiln speed increase BZT more
6 BZT Low When BZT is slightly low and oxygen is slightly high, but still o.k.:
OXYGEN o.k. 1 decrease kiln draft increase BZT and decrease BET
BET High When BZT is very low:
1 decrease fuel rate decrease BET, increase oxygen to prepare for step 2
2 decrease kiln draft decrease BET, bring oxygen back to normal
3 decrease kiln speed increase BZT

7 BZT Low When BZT is slightly low:

OXYGEN high 1 increase fuel rate increase BZT and BET, decrease oxygen
BET low When BZT is very low:
1 increase fuel rate increase BZT and BET, decrease oxygen
2 decrease kiln speed increase BZT and BET

OXYGEN high 1 increase fuel rate increase BZT, decrease oxygen
BET o.k. When BZT is very low:
1 increase fuel rate increase BZT, decrease oxygen
3 decrease feed rate to maintain kiln filling degree constant

OXYGEN high 1 increase fuel rate increase BZT, decrease oxygen
BET high 2 decrease kiln draft decrease BET and oxygen, compensate BET increase of step 1
1 increase fuel rate increase BZT, decrease oxygen
2 decrease kiln draft decrease BET and oxygen, compensate BET increase of step 1
4 decrease feed rate to maintain kiln filling degree constant
10 BZT o.k. 1 increase kiln draft increase oxygen, increase BET

OXYGEN low 2 increase fuel rate slightly compensate BZT decrease of step 1
BET low
11 BZT o.k. 1 increase kiln draft increase oxygen

OXYGEN low
BET o.k.
12 BZT o.k. 1 decrease fuel rate decrease BET, increase oxygen and prepare for step 2
OXYGEN low 2 decrease kiln draft decrease BET
slightly
BET high
13 BZT o.k. 1 increase kiln draft increase BET

OXYGEN o.k. 2 increase fuel rate slightly compensate oxygen increase and BZT decrease of step 1
BET low
14 BZT o.k. When this condition persists:

OXYGEN o.k. 1 increase kiln draft increase oxygen to prepare for step 2

BET o.k. 2 increase fuel rate increase BZT to prepare for step 3
3 increase feed rate increase production
4 increase kiln speed to maintain kiln filling degree constant
When this condition is only temporary:
1 No action
15 BZT o.k. When oxygen is slightly high, but still o.k.:

OXYGEN o.k. 1 decrease kiln draft decrease BET
BET high When oxygen is slightly low, but still o.k.:
1 decrease fuel rate decrease BET and prepare for step 2
2 decrease kiln draft decrease BET
16 BZT o.k. 1 increase kiln draft slightly increase BET, increase oxygen to prepare for step 2
OXYGEN high 2 increase fuel rate increase BET, compensate BZT decrease of step 1
BET low
17 BZT o.k. 1 decrease kiln draft decrease oxygen

OXYGEN high
BET o.k.
18 BZT o.k. 1 decrease kiln draft decrease oxygen and BET

OXYGEN high 2 decrease fuel rate decrease BET, compensate BZT increase of step 1
slightly
BET high
19 BZT high When BZT is slightly high:

OXYGEN low 1 increase kiln draft increase oxygen, increase BET
BET low When BZT is high:
1 increase kiln draft increase oxygen, increase BET
2 increase kiln speed decrease BZT
3 increase feed rate to maintain kiln filling degree constant

OXYGEN low 1 increase kiln draft increase oxygen, decrease BZT
BET o.k. When BZT is high:
1 decrease fuel rate increase oxygen, decrease BZT

OXYGEN low 1 decrease fuel rate decrease BZT and BET, increase oxygen
BET high When BZT is high:
1 decrease fuel rate decrease BZT and BET, increase oxygen
2 increase kiln speed decrease BZT and BET


OXYGEN o.k. 1 increase kiln draft decrease BZT, increase BET
1 increase kiln draft decrease BZT, increase BET

OXYGEN o.k. 1 decrease fuel rate decrease BZT
1 decrease fuel rate decrease BZT

OXYGEN o.k. 1 decrease fuel rate decrease BZT and BET
BET high When BZT is high:
1 decrease fuel rate decrease BZT and BET

OXYGEN high 1 increase kiln draft decrease BZT, increase BET
1 increase kiln draft decrease BZT, increase BET

OXYGEN high 1 decrease fuel rate decrease BZT
1 increase kiln draft decrease BZT, increase BET to prepare for step 2

OXYGEN high 1 decrease fuel rate decrease BZT and BET
BET high 2 decrease kiln draft decrease oxygen
When BZT is high:
1 decrease fuel rate decrease BZT and BET
2 decrease kiln draft decrease oxygen

The basic 27 counteractions apply to the kiln operation only and do not consider
precalcining, Lepol preheater or grate cooler operation.
If a precalcination system is installed, additional control parameters have to be held on
target.
The same applies for a grate cooler (which is the most common type of cooler and has the
highest operational adjustability) and a Lepol preheater.
5. PRECALCINER CONTROL
The purpose of precalcining is to shift the reaction of calcination from the kiln upstream into
a kiln external precalciner.
Although preheaters with secondary firing and air through precalciner belong to this group
also, they are not considered further, since the fuel rate to the secondary firing or to the air
through calciner is normally held constant and they do not provide particular control of the
calcining degree at the kiln inlet. Also the amount of energy spent for precalcination is limited
in those types (normally << 20 % of the total thermal energy).
The presented system is the air separate calciner, where the combustion air is not delivered
through the kiln, but through a tertiary air duct.
In these precalciners, the meal is calcined almost completely (up to 90 %) and more than
half of the total thermal energy (up to 60 %) is spent for this reaction.
It is the objective to calcine the meal as much and as uniform as possible.
This is the main operational advantage of an air separate precalciner, because it gives the
possibility, to control the preparation of the meal completely independent from the kiln.
Therefore the BET has little importance from an operational point of view.
5.1 Calcination temperature

In kiln operation the free CaO (or liter weight) determines the BZT target (control parameter),
which is held close to target mainly by means of the fuel rate to the main burner (control
variable).
Similarly in precalcining systems the calcination degree establishes the calcining
temperature (control parameter), which then is held, close to target by the fuel rate to the
calciner (control variable). This calcination temperature in calciner systems basically takes
over the function of the BET of the other kiln systems.
A normal desired calcination degree in a precalciner system is around 90 %, which
corresponds to a material temperature (or a calciner exit gas temperature) of 870-890C.
A higher calcination degree is obtained with a higher calcination temperature, but at
temperatures above 900C already liquid phase starts to form, which leads to build-ups in
the calciner and a restricted gas flow. Therefore excess temperatures are not desired.
From an operational point of view the following general rules exist to establish the calcining
temperature and calcination degree:
- The higher the calcination temperature (and calcination degree), the more uniform is the
calcination degree (since the gradient of the calcining curve decreases with increasing
calcination temperature) and the more stable is consequently kiln operation, but the lower is
the thermal efficiency of the precalciner kiln system, due to higher calciner and preheater
exit temperatures.

- The lower the calcination temperature (and calcination degree), the less uniform the
calcination degree (since the gradient of the calcining curve increases with decreasing
calcination temperature) and the more sensitive is the calcination degree to fluctuations
(disturbances) on rate and properties of precalciner fuel and kiln feed.
100%
90%
80%
70%
Calcination degree (apparent)
60%
50%
40%
30%
20%
10%
0%
760 780 800 820 840 860 880 900 920 940
Temperature
Figure 8: Typical calcining curve
5.2 Oxygen control

The oxygen at the calciner outlet is controlled through the calciner draft, in the same way as
the oxygen at kiln inlet is controlled by the kiln draft.
A normal target value for the oxygen after calciner (bottom cyclone exit) is 1 - 1.5 %.
In air separate calciner systems, the calciner draft is adjusted by the exhaust gas fan.
Adjustments to the exhaust gas fan also influence the kiln draft. Therefore any adjustment to
the exhaust gas fan normally implies also an adjustment of the tertiary air damper (or
depending on the design to the kiln orifice), in order to maintain the kiln draft constant.
Only in case of separate line calciners the calciner draft can be controlled independently
from the kiln through the additional exhaust gas fan.
5.3 Other control parameters

Beside the two principal control parameters temperature and oxygen and the two control
variables fuel rate into the calciner and calciner draft, in certain AS designs still more
adjustability is given through a staged introduction of fuel and / or meal and / or tertiary air.
These design features are implemented for different reasons (e.g. NOx reduction, better
ignition and burnout of low reactive fuels, refractory protection at a local hotspot, etc.). The
particular adjustment of them will not be discussed within this paper.

6. GRATE COOLER CONTROL
The grate cooler is not only the most widespread type of cooler, it is also the one with the
highest adaptability to the different process requirements.
6.1 Bed height and grate speed

The main objective of grate cooler operation is to recuperate the highest possible amount of
energy from the hot clinker, in order to achieve a high combustion air temperature and to
cool the clinker adequately.
Dependent upon the clinker granulometry a certain clinker bed height is found, at which the
efficiency of recuperation is best. The grate cooler operation should try to maintain the
thickness of the clinker layer on the grate constant at this optimum height.
Since this optimum thickness depends principally on the clinker granulometry, the bed height
of best performance can vary over time and may change the bed height target (control
parameter).
To determine the thickness of the clinker bed the following methods are used:
Back-pressure under the first grate, if the flow of the respective cooler fan is held
constant (most common method)
Level detecting systems, working with a radar distance measurement
Weighing systems that weigh the clinker load of a particular area, e.g. the clinker
load over the fixed inlet (rarely used)
The control variable to adjust the clinker bed thickness is the speed of the cooler grate (or
transport velocity of the clinker). By increasing the grate speed the clinker bed is moved out
faster and the bed height is lowered. A decrease of the grate speed has the inverse effect.
Normally this grate speed is controlled automatically (mostly in function of the backpressure
of one or several chambers under the first grate).
The loop, which is responsible for this control, basically has to fulfill two contrary tasks:
quickly speed up the grate during coating falls, ring break outs and all other
occasions, where the kiln releases rapidly a high amount of clinker (which requires
aggressive PID settings: Proportional , Integral )
maintain under normal operating conditions the clinker bed height (i.e. the
backpressure under the grate) as close to the desired setpoint as possible, without
oscillation or cycling (which requires soft PID settings: Proportional , Integral )
A controller, which fulfills these requirements, may have the following typical characteristics:
Smooth grate speed adjustment, if the pressure is within a certain predefined band
(e.g. 35 mbar +/- 2 mbar)
Aggressive grate speed adjustment, if the pressure is outside the predefined band

45
43
41
39
under grate pressure (mbar)
37
35 dp vs SPM
high limit
33
low limit
31
29
27
25
-6 -4 -2 0 2 4 6
grate speed change (SPM)
Figure 9: Typical grate speed controller characteristic (example)
If the cooler consists of more than one grate, the speed of the following grates is held
normally in a proportional ratio to the speed of the first grate. Thereby the grate speed ratio
of the following grates increases gradually, thus the clinker bed height decreases stepwise
from the first grate to the following ones.
6.2 Subordinate control loops
6.2.1 Cooler air flow

The flow of the cooler fans is normally automatically held constant and independent from the
backpressure of the grate and clinker bed.
The objective is to provide always the same amount of cooling air to the grate, independently
from minor fluctuations of the cooler throughput.
A constant airflow is a basic prerequisite to control the bed height of the clinker when the
grate speed control works with the backpressure under the first grate.
The specific airflow (amount of cooling air per square meter and second) and the
backpressure are highest at the first rows of the cooler and have a descending pattern
towards the cooler outlet.
This is due to the fact, that the cooling of the clinker and the recuperation of the heat is as
more efficient as greater the temperature difference between clinker and cooling air is.
Therefore more cooling air is used at the inlet, where the clinker is still hot.
A typical air distribution (with the respective backpressures) of a cooler is shown below (in
the example the cooler has 1 grate and 6 chamber aerated sections):

1.75 50
1.50
1.25
[Nm3/m2,s]
1.00
[mbar]
25
0.75
0.50
0.25
0.00 0
1 2 3 4 5 6
Compartments
Specific cooling air flowrate Undergrate pressure
Figure 10: Cooling air flow distruibution (example)
6.2.2 Kiln hood pressure

In a grate cooler, the amount of air used to cool the clinker is higher than the amount of air
required for combustion in the kiln (and calciner). The surplus air has to be taken out of the
cooler by a separate venting system, the so-called waste air fan.
The draft of the waste air fan has to be adjusted so that the pressure in the cooler and the
kiln hood is neither positive nor too negative.
Normally this pressure is measured in the kiln hood and held constant at a value of around
0.2 - 0.5 mbar. To decrease the kiln hood pressure (make more negative), the draft of the
waste air fan is increased and vice versa.
Most frequently an automatic control takes care of this.
A constant and slightly negative kiln hood pressure is important for safety reasons and for
the protection of the equipment. As the waste air fan counterbalances (pulls against) the ID
fan the kiln hood pressure should not be more negative than necessary under normal
operation conditions.
Nevertheless the kiln hood pressure frequently is set during start ups (cold kiln) to much
lower values ( 2.0 - 3.0 mbar), in order to hold the flame back (shorten flame) and to avoid
an overheating of the back end (avoid high BET due to long flame and overdrafting).

7. LEPOL GRATE CONTROL
A Lepol grate preheater basically consists of a traveling grate, which passes through two
compartments. The pellets on the grate, which were formed in the pelletizer (semi dry
process only), are first dried in the drying compartment and are then heated up further and
partly calcined in the hot compartment.
The hot gases from the kiln penetrate the bed on the grate in the hot compartment
downward, from where they are withdrawn by the intermediate fan, which controls the kiln
draft.
From there they are pulled through the drying compartment by the exhaust gas fan, which
controls the pressures in the drying compartment.
Adjustable gate
for bed height
control
p 0
Figure 11: Gas - and material flows of a Lepol preheater
7.1 Subordinary control loops
7.1.1 Pressure control

The pressure drop over the grate in the drying and in the hot chamber is typically in the
magnitude of 6 - 10 mbar, depending on the bed height on the grate and the void volume of
the bed.
The size of the compartments of the Lepol preheater is designed in a way that the pressure
drop over the grate in both compartments is normally equal. (Nevertheless in certain kilns,
where the dust load at the kiln inlet is high, the pressure drop over the bed in the hot
compartment is higher than the one in the drying compartment, due to a reduced void
volume through the dust).
Since the sealing in-between the two compartments is limited it is of utmost importance to
balance the pressures in between the two compartments underneath and above the grate in
a proper way.

This is the task of the exhaust gas fan which is usually adjusted in a way that the pressure
underneath the grate of the drying compartment is approximately equal to the pressure
underneath the grate of the hot compartment (if the pressure drop over the grate in both
compartments is equal then consequently also the pressure above the grate of the drying
compartment must be equal to the pressure above the grate of the hot compartment).
If the exhaust fan pulls too much, the pressure in the drying compartment underneath and
above the grate is more negative than in the hot compartment, which leads to a short-circuit
of hot gases from the hot compartment into the drying compartment and the exhaust gas
system.
This short-circuit of hot gases from the hot compartment into the drying compartment often
creates a bursting of the pellets (due to an excessive water vapor pressure in the pellets).
If the exhaust fan does not pull enough, the pressure in the drying compartment underneath
and above the grate is less negative than in the hot compartment, which leads to a short-
circuit of "cold" gases from the drying compartment into the hot compartment and a
recirculation of the gases from the drying compartment through the intermediate fan, which
reduces consequently the kiln draft.
7.1.2 Traveling grate speed control

In order to guarantee a good heat exchange between gases and pellets it is important to
maintain a certain bed height of the bed on the grate constant. The optimum bed height of
the pellets, which is basically a function of the bed resistance must be found empirically and
depends on the void volume of the bed (pellet size distribution)
The bed height is established by adjusting the passage wall between material feed chute
and drying compartment (see figure 11).
The level of the material in the feed bin is held constant by adjusting the material transport to
and from the pelletizer to the grate.
With increasing traveling grate speed the material feed rate increases and vice versa.
7.1.3 Pelletizer control

In semi dry kiln systems the pellets are formed from raw meal under addition of 10 - 15 %
water in the pelletizer, which is just a rotating dish.

Addition of water
1 - pellet nucleus
Raw 2 - pellet of medium size
meal
3 - circulation zone
4 - finished product
Figure 12: Pelletizer (semi dry process)
The most important properties of the pellets are size, uniformity, porosity, moisture content
and strength. These properties are influenced by the raw meal feed rate (throughput), the
ratio of water to raw meal (moisture content of the pellets), the angle of tilt of the dish, the
speed of rotation, the height of the rim of the dish and other parameters.
From an operational point of view the raw meal feed rate (throughput) and the ratio of water
to raw meal are the most important factors, which influence the properties of the pellets.
The influence of raw meal feed rate (throughput) and the ratio of water to raw meal is:
Raw meal feed rate (at a fixed water to raw meal ratio)
The higher the raw meal feed rate, the smaller the pellet size (reduced void volume), the
more porous the pellets and the less uniform the size distribution of the pellets (reduced
void volume)
The lower the raw meal feed rate, the bigger the pellet size (bursting of the pellets), the
less porous the pellet (bursting of the pellets) and the more uniform the size distribution
of the pellets
Ratio of water to raw meal (at a fixed raw meal feed rate)

The higher the ratio of water to raw meal, the bigger the pellet size (bursting of the
pellets) and the higher the strength of the pellet
The lower the ratio of water to raw meal, the smaller the pellet size (reduced void
volume) and the lower the strength of the pellet (breakage during transport)
The relation between pellet size, raw meal feed rate and ratio of water to raw meal (moisture
content of the pellets) is shown qualitatively in figure 12.
5
m
3
m tio o me
m
(ra ra
oi f w al)
4
to
st ate
m
u r
w
average pellet size
re
2
m
1
retention time in pelletizer m
raw meal througput

(m5>m4>m3>m2>m1)
Figure 13: Pellet size in function of raw meal feed rate and ratio of water to raw meal
In order to maintain a certain pellet size (example in figure 13), the ratio of water to raw meal
(moisture content of the pellets) has to be increased slightly with higher raw meal feed rates
(and vice versa).
Nevertheless the required change in the ratio of water to raw meal is small (several tenth of
a percent in moisture) and a proper granulation does not simply depend on the raw meal
feed rate and the ratio of water to raw meal, but is also strongly influence by the chemical
and mineralogical properties of the raw meal, the design and setting of the pelletizer and a
number of other factors.
8. PROCESS PROBLEMS AND KILN UPSET CONDITIONS

There are a number of unfavorable situations that require special attention and
counteractions. From the great number of possible problems and upset conditions the most
important ones and their respective counteractions are now presented.
8.1 Cycling
Cycling is a condition, where one or more kiln parameters (but mostly the BZT) oscillate with
great amplitude around the target. This is the so-called overshooting.
An improper kiln operation itself causes quite frequently cycling, especially if the
counteractions that are taken to bring back one or more control variables to target are too
strong. This is a well-known phenomenon in control engineering, a general problem that
exists also in other areas of control.

To inhibit a cycling of the kiln, the operation has to be smooth, i.e. any adjustment to the
control variables must be as big as necessary, but as small as possible. A counteraction
should consider not only on the actual condition, but also have a view to the past and future
changes, i.e. beside the actual value also the trend of a certain control variable has to be
considered.
To break a cycle, the kiln operation must start to counteract against one extreme of the cycle
already before it gets there. When the period of the cycle is known, the counteraction for the
following extreme should be taken already before the half of both extremes. This reduces the
amplitude of each following cycle and brings the kiln back into stable operation.
In cycles that are caused by the kiln operation itself, the periodicity normally is approximately
twice the retention time of the material in the kiln.
In chapter 11 two formulas to calculate the retention time are given. With the knowledge of
this periodicity it is possible to break a cycle early.
Beside those cycles, that can be attributed to the kiln operation other factors are also
responsible. The most common ones are fluctuations of the raw material and the fuels
properties. If those fluctuations have a periodic character, they may be counteracted in the
same way as explained above. If they are random, the operation never will be a (counter-)
action at time but always a reaction, to remedy what already happened.
Since cycling conditions have a strong negative impact on the productivity of the kiln, all
efforts must be made to achieve a stable kiln operation. Therefore one of the most basic
prerequisites are homogenous feed and fuel properties.
8.2 Coating Collapse

Coating is the protective layer of molten and resolidified liquid phases on the refractory in the
burning zone. Dependent on the properties of the raw materials, the refractory, the fuels and
the flame, this coating can be more or less pronounced.
A coating collapse is a situation, where large parts of this coating break away from the
refractory in big lumps, due to excessive weight, large temperature changes in the burning
zone (especially sharp temperature drops), fluctuations of the raw material properties,
inadequate operation and others.
A coating collapse is detected primarily through the kiln amps. A sharp increase in the average kiln
amps indicates that suddenly a higher amount of material has to be moved in the kiln.
Constant spiking of the kiln amps may also indicate an uneven loss of coating in one area of
the kiln.
In case a kiln camera is installed, the fallen coating can also be observed visually in the last
meters before the kiln outlet. These indications help to estimate the severity of coating loss
and the magnitude of counteraction that has to be made.
When coating falls out and the quantity is not too excessive, no adjustment is done to the
control variables. The BZT may drop slightly for a short period but normally quickly recovers,
as the coating was already closed to sintering temperature. A coating collapse results
normally in elevated free CaO values of the clinker, since it is difficult to clinkerise the large
pieces. For this reason no attempt should be made to clinkerise the collapsed coating, as
this results only in an overheating of the entire system.
If the coating collapse is more pronounced, special attention has to be paid to the cooler,
since an excessive amount of material may overfill the cooler and the clinker transport
systems or lead to an excessive clinker end temperature.
If a hammer crusher is installed a close look must be given, that the coating lumps do not
overload or block the crusher, otherwise the cooler grate speed has to be reduced.

When a cooler or clinker crusher overload is probable and the cooler grate speed has to be
reduced significantly, the kiln should be slowed down temporarily also, in order to decrease
the clinker output of the kiln and to avoid an overfilling of the cooler at the inlet. If the
situation is that severe, that the kiln has to be slowed down for a longer period (longer than
five minutes maximum, local equipment conditions must prevail), the fuel rate, the feed rate
and the kiln draft have to be reduced.
8.3 Ring break out

In the following the counteractions after a ring break out are described.
8.3.1 Clinker ring break out

A clinker ring, which forms at the kiln outlet, is an obstacle for the clinker causing the clinker
to dam up behind this ring. If this ring breaks out, an excessive amount of completely
sintered clinker is released from the kiln.
A sharp increase of the cooler bed height (for grate coolers), a smooth decrease of the kiln
amps and a smooth increase of the secondary air temperature indicate a break out of a
clinker ring. In most cases where a kiln camera is installed, the break out of a clinker ring
can be also observed visually.
For grate coolers: Due to the temporarily resulting higher output of the kiln, it is often
necessary to slow down the kiln, in order to avoid an overloading of the cooler which may
result in excessive clinker and waste air temperatures or a stalling of the cooler fans.
8.3.2 Sinter, middle, meal and mud rings

Each of the different ring types has its particular location and genesis (origin).
However they all have in common, that they dam up the material flow. Once a rings breaks
out, an elevated amount of only partly prepared material is released. This cools down the
kiln and shifts all reaction zones downwards. The kiln "pushes".
The immediate detection of ring breakouts is difficult. Since unsintered material does not
influence the kiln torque significantly, the kiln amps often do not show any change. Only if
the ring was big enough that it already constituted a considerable hindrance to the gas flow,
a small decrease in the pressure loss over the kiln may then be observed.
Because an immediate detection of a ring break outs is often difficult, only a sharp drop of
the BZT and an suddenly increased kiln output can give rise to the suspicion that a ring has
broken out earlier.
Kiln shell scanners will assist in the detection of ring formation. Likewise they can easily
indicate the gradual loss of a ring. Unfortunately they will not often respond fast enough to
indicate a ring collapse.
The counteractions to be taken depend on the size of the ring breakout and the general kiln
condition and are most likely according to cases 1-9 of the counteractions described
previously.

8.4 Hot meal rush (avalanche)
A hot meal rush or the so-called "avalanche" is a situation where the BZT drops that much
that calcined but unsintered material reaches the kiln outlet. This situation may be caused by
the kiln operation itself, by irregular feed advancement (e.g. ring break out, temporary
cyclone blockage) or by failures of the feed or fuel feeding systems among others.
Once the hot meal has reached the kiln outlet, every effort must be made to avoid that this
meal enters the cooler. Unsintered but hot meal can cause serious damage to the cooler
grate.
Manual operation of the cooler may also be required if the flush reaches the cooler before
corrective kiln actions can be initiated.
If the hot meal enters the cooler, a high percentage of it is blown back into the kiln. The very
dusty atmosphere in the kiln inhibits a proper heat transfer of the flame towards the burning
zone. Additionally the dust cools down the flame. Since nearly no heat can be recuperated
from the meal, the secondary air temperature decreases.
All of these factors negatively impact the heat transfer and lead to a further drop of the BZT
(but an increase of the BET).
Therefore the only, but absolute obligatory counteraction is to slow down the kiln
immediately, in order to avoid dust entering the cooler and to adjust the control variables
according case 1-9 of the counteractions described before.
Most likely the kiln speed has to be reduced that much that a parallel reduction of the kiln
feed is not possible to avoid an overheating of the back end. In these circumstances cycling
can occur.
8.5 Red spot on the kiln shell

A red spot is a locally limited area of the kiln shell that is overheated.
The shell is considered to be overheated, when the surface temperature exceeds 450C.
visible radiation on the shell surface starts at temperatures above approximately 475C that
is why it is called a red spot.
This normally comes from a worn out refractory and the lack of any coating in this area,
resulting in a too high heat transfer from the inside to the outside of the kiln, due to the
insufficient insulation.
High shell temperatures cause a warping of the kiln shell, which consequently inhibit proper
installation of the refractory and lead to a premature failure of the bricks, due to excessive
mechanical forces. Therefore they must be avoided at all times.
There are several possible reasons for a hot spot.
When the bricks are already thin and the protective coating layer breaks away,
the remaining refractory does not provide sufficient insulation. Hot spots created
by this mechanism can appear from one moment to another.
When a coating collapse occurs, the coating breaks off a layer of the refractory
and weakens the brick. This may occur also as a secondary effect, when the
refractory starts to crack due to the thermal shock the brick suffers, when it is
exposed suddenly to the full heat of the burning zone (thermal spalling).
All type of mechanical brick failures, where parts of the refractory fall out and
leaves the shell partly or completely unprotected.
A coating loss and a local overheating of the shell may be caused by:

A too hard burning practice, which melts away the coating and washes out the
lining. The generally hot kiln transfers more heat to the shell than usual and a hot
spot may occur in those areas, where the bricks were previously weak.
Improper alignment of the burner pipe with respect to the kiln axis.
Incorrect adjustment of axial or radial dampers, especially when too much radial
air create a too broad and bushy flame.
Snowman build-up on burner pipe which deflects air flow (or fuel) at the burner tip
towards the bricks and the kiln shell.
Poor kiln feed chemistry control resulting in periods of heavy over-burning
followed by periods of under-burning.
If the hot spot is not too big (less than ~1 m2) and if it is in the burning zone, where the
coating is, the kiln operator first of all should try to rebuild a protective coating layer.
A common practice is to maintain the kiln warm but to permit the calcined meal to enter
periodically the affected area (e.g. by variation of the kiln speed). This cools down the
concerned hot spot slightly and the meal starts to solidify and to stick over the damaged
area. Nevertheless, special attention has to be paid to avoid a real cool down of the burning
zone, since a cold kiln does not form coating and the following heat up during recuperation
ruins the coating and the affected kiln shell even more.
In certain cases the same effect of moving the reaction zones up and down, which favors the
build up of coating can be obtained by a periodic change of the temperature profile of the
flame. The temperature profile is changed by regularly moving the burner into the kiln (which
makes the flame longer) and out (which makes the flame shorter). Alternately the primary air
settings (ratio of axial / radial air) can be changed, although this is rather not recommended.
With the correct chemistry a very slightly overheated kiln will build and maintain coating
whereas a cold kiln cannot easily build or sustain coating.
An external cooling of the kiln shell in the damaged area through a fan promotes the
formation of coating also. Thereby the shell temperature is reduced locally, which decreases
also the temperature inside the kiln. The meal or already formed clinker will solidify more
likely on the cooled area than in the hotter neighborhood.
If the mentioned measures are not effective and the situation does not improve after a
certain lapse (not more than 4 h), the kiln must be shut down.
If the hot spot is bigger than ~1 m2 it is almost impossible to recover a normal shell
temperature through the build-up of coating. In most case the refractory is affected already
that strongly that a further detoriation of the lining through mechanical weakness is likely.
There can be also nothing done, if the red spot occurs in an area, where no coating is
formed.
The common practice is then to shut down the kiln immediately and to replace the damaged
refractory sections completely.
A red spot located under or adjacent to a tire is cause for immediate action. Often plant
management will mandate an immediate kiln shut down.
8.6 Loss of kiln feed

A loss of feed is a situation where the kiln feed devices like feeders, auxiliary transport or
extraction equipment lead to a reduction or a total loss of the kiln feed.
Since no heat is absorbed by the missing feed, the system starts to overheat, which may
lead to equipment failures due to overtemperature.

Therefore the fuel rate and the kiln speed have to be reduced significantly (i.e. up to 80%).
As a guideline, if the normal kiln feed cannot be restored within less than 10 minutes, the kiln
has to be shut down. Tower exit temperature must be maintained below safety set points for
cyclone dip tubes, downcomer, ID fan along with any additional downstream equipment. The
actual temperature profile must be the guide to how fast and how severe any required
adjustments are made.
9. STARTUP AND SHUTDOWN PROCEDURES

The start up and the shut down of a kiln system are the most critical phases during kiln
operation. Most serious equipment failures and accidents, which are caused by the
operation itself, happen usually during these two periods. Therefore special caution and
attention has to be paid to the kiln system before and during start up and shut down.
9.1 Start up
When starting up a cold kiln, the general rules that have to be considered are (list not
comprehensive):
Make sure all equipment is released and cleared from maintenance people
All vessels and risers have been inspected
All process measurement instruments are available
Ensure that all inspection doors and openings are closed
The flame is supervised after lighting the burner (by a flame detector or by eye)
The draft in the kiln is sufficient, but not excessive (overheating of the back).
The heat up of the kiln is according a determined rate, which is given mainly by
the kiln dimension and the type of refractory (normally ~50-70C/h at the
beginning, 60-90C/h when the shell is already warm). A preheating curve is
recommended. Various curves can be used depending upon several factors
including type and amount of refractory replacement within the kiln system.
The kiln is periodically turned (jogged) 100 degrees, in order to avoid unbalanced
heating of the kiln shell. The jogging schedule should be aligned with the
preheating curve. The kiln should be turned continuously once the BET has
reached 750C (in case of heavy rain even earlier).

Figure 13: Example of a heat up curve (4 stage PH kiln)
Occasionally introducing small amounts of kiln feed during the preheat will help
the bricks to set during heat up (especially, if the bricks are just installed). This
will also help to prevent brick migration during periods of continuous kiln rotation.
In case of grate coolers: make sure the first rows of the grate are covered with
clinker, run the first cooler fans already during heat-up.
In case of grate coolers: decrease (more negative) kiln hood pressure to
maintain a short flame and to avoid overdrafting of the system
In case of grate coolers: start all cooler fans before feeding the kiln. Adjust the
total cooler airflow to ~2-2.5 Nm3/kg cli. Increase the cooler airflow proportionally
with the kiln feed. Maintain also during start up the descending pattern of the
specific air distribution, which is highest at the inlet and lowest at the outlet

(graph of 7.2.1.). In many cases, the first fans reach nominal air flow already at
70 - 80% of the nominal kiln feed rate.
In case of grate coolers: Automatic control of the cooler fans using the above
setpoint guidelines, including the cooler waste air fan is desired before feeding
the kiln. This will ensure sufficient air volumes for combustion.
In case of cyclones: make sure all pendulum flaps are released and all shutoff
gates are opened.
In case of cyclones: make sure the draft is sufficient when starting to feed in
order to guarantee, that the meal is not falling directly through the risers.
Shock blasters must be online and set in the automatic mode, manual pulsing
can still be initiate should the operator require it.
Careful and close observation of the tower temperature and pressure profile
must be maintained during this very critical phase of kiln operation.
9.2 Shut down

A typical shut down procedure of a kiln is (list not comprehensive):
In case of a short-term shut down: Shut off all fuel to the system and reduce the
draft in the kiln to a minimum (avoid overheating of the back and maintain the
burning zone hot). However in some plants a maintenance flame is re-
established in order to help maintain kiln temperature.
In case of a long-term shut down: maintain a certain draft in the kiln but avoid
overheating of the back (cool the system as fast as possible).
Turn the kiln with the auxiliary drive for 30 minutes (in case of heavy rain fall even
longer). Later on turn the kiln gradually less frequent by 100 degree, in order to
avoid unbalanced cooling of the kiln shell. Unless requested for a kiln relining: do
not run the kiln empty.
In case of grate coolers: shut of the grate drive to maintain the first rows of the
grate covered with clinker; leave the first cooler fans running.
In case of cyclones: make sure all pendulum flaps are tightened and all shutoff
gates are closed.

10. GENERAL CONSIDERATIONS
When operating a kiln the following general principles have to be kept in mind, which are the
fundamental guidelines for proper operation.
Protection of personnel and equipment
It is mandatory in kiln operation to avoid dangerous situations, where personnel

may be injured or equipment may fail, even if a production loss or a poor quality
results.
Any abnormal situation such as, but not limited to hot spots on the kiln shell, red
grates in the cooler, uncontrolled hot meal rushes or fuel spillages must be
prevented or remedied as soon as possible. The physical damage to the
equipment that may result from an acceptance of these abnormal conditions is
out of all proportion to the possible loss of clinker production that may occur from
the resulting repair works.
This applies even more, if the health or life of personnel is concerned.
Constantly good clinker quality
The aim of clinker manufacturing is to produce a clinker that is neither

overburned nor underburned, since both extremes are detrimental for the cement
strength. The quality, most frequently expressed as free lime content or the liter
weight, should vary only in a small range.
Constant clinker quality is not only significant with view to the properties of the
final product cement, but also for the following milling process (grindability,
granulometry, gypsum addition, etc.).
Smooth and stable operation
A smooth and stable kiln operation is a condition, where only very small
adjustments have to be made to the kiln system.
Smooth operation is a prerequisite for a constant and uniform clinker quality.
Running a kiln in a smooth and stable manner prolongs the lifetime of the
refractory as it enhances the stability of the coating in the burning zone. Both
directly improve the overall productivity, since fewer kiln stops are necessary for
rebricking works.
A stable kiln operation must always be preferred to a temporary maximum
production that cannot be sustained.
Therefore the same emphasis that is put to push the kiln feed up should be
applied to reduce it, if this is necessary.
By respecting this rule, the highest long-term productivity will be achieved.

Maximum thermal efficiency
For economical and environmental reasons it is naturally the interest of every

plant, to produce each ton of clinker with a minimum amount of fuel. Overburning
the clinker, poor combustion, and unstable operation among many others
prevents high fuel efficiency.
Maximum production rate
The highest average production rate is achieved, when the kiln runs stable. Any
upset condition (especially cooling of the burning zone) must be avoided.
Hence the focus must not lay only on a temporary production increase, if this
provokes a cycling kiln or carries the risk of weakening (cooling off) the burning
zone, since these situations imply a dramatic decrease of the feed rate and lead
to a loss of overall productivity.
The feed should be increased only, if the kiln can be held fundamentally stable at
a certain high production rate over some time.

11. CALCULATION OF THE RETENTION TIME
The knowledge of the retention time of the material in the kiln is of a great importance for kiln
operation.
When making adjustments to the control variables, especially during cycling conditions, the
retention time is in most kiln systems a guide for the reaction time of the kiln, since both are
almost the same.
The retention time of the material in a dry rotary kiln can be determined mathematically in
the following ways:
According Duda:
1.77 l
t= F
d n
According Labahn:
l 1
t=
d n tan
With: t: Time in minutes

l: Length of kiln in meter
: Angle of repose in degrees: 35-40 for clinker 0-50 mm
: Slope of kiln in degrees (normally 1 2.5 )
: Angle of the moved material in the kiln
(tan must be calculated from sin = sin / sin )
d: Internal diameter of the kiln in meter (in-between bricks)
n: Number of revolutions per minute
F: Factor, which equals 1, if the kiln has a constant diameter

Example:
PH Kiln with the following data:
l: 60 m, : 35 , : 3 % = 1.72 , d: 3.6 m, n: 1.7 1/min, F: 1
According Duda:
1.77 60 35
t= 1 = 59.7 min
1.72 3.6 1.7
According Labahn:
60 1
t= = 59.6 min
3.6 1.7 tan 3.00
with = 3.00 from sin = sin 1.72 / sin 35

PC Kiln with the following data:
l: 70 m, : 35 , : 3.5 % = 2.00 , d: 5.6 m, n: 2.0 1/min, F: 1
According Duda:
1.77 70 35
t= 1 = 32.6 min
2.00 5.6 2.0
According Labahn:
70 1
t= = 32.6 min
5.6 2.0 tan 3.49
with = 3.49 from sin = sin 2.00 / sin 35

Chapter 10
Heat Balance

Heat Balances of Kilns and Coolers and Related Topics
Th. Richner / A. Obrist

PT 99/14496/E (update of VA 8/5180/E)
1. INTRODUCTION ............................................................................................................. 611

2. OVERVIEW OF COMPLETE HEAT BALANCE PROCEDURE ..................................... 611
2.1 Why to do a Heat Balance? .............................................................................. 611
2.2 How to Proceed................................................................................................. 612
3. BASIS OF BALANCE AND REMARKS REGARDING EXECUTION ............................ 613
3.1 Balance Limits................................................................................................... 613
3.2 Guidelines for Test Duration ............................................................................. 615
3.3 Kiln Operation ................................................................................................... 615
3.4 Kiln Data............................................................................................................ 615
4. HEAT BALANCE CALCULATIONS ............................................................................... 617
4.1 General Remarks .............................................................................................. 617
4.2 Determination of Clinker Production ................................................................. 619
4.3 Heat from Fuel .................................................................................................. 625
4.4 Burnable Components in Raw Material............................................................. 626
4.5 Loss due to Incomplete Combustion................................................................. 627
4.6 Heat of Formation ............................................................................................. 627
4.7 Heat due to Partly Decarbonized Material ........................................................ 628
4.8 Heat of Evaporation .......................................................................................... 630
4.9 Sensible Heat.................................................................................................... 631
4.10 Heat Loss due to Radiation and Convection..................................................... 635
5. HEAT BALANCES .......................................................................................................... 644
5.1 General ............................................................................................................. 644
5.2 Examples of Heat Balances of Various Kiln Systems ....................................... 646
6. SPECIAL PART .............................................................................................................. 648
6.1 Influence of Reference Temperature ................................................................ 648
6.2 Heat of Formation ............................................................................................. 651
6.3 Radiation Heat Transfer .................................................................................... 655
6.4 Convective Heat Transfer ................................................................................. 656
6.5 Effect of Thermal Improvements ....................................................................... 660
6.6 Heat Transfer in Preheaters and Coolers and Improvement Potential ............. 662
7. TEST QUESTIONS ......................................................................................................... 667

8. LITERATURE.................................................................................................................. 668
9. SYMBOLS AND UNITS .................................................................................................. 669

SUMMARY
A heat balance is an efficient tool to measure the actual state of a kiln system. It allows a
better assessment of the heat consumption and reveals the potentials for improvements
(regarding the thermal aspects).
The principle of a heat balance is rather simple: we select our system limits for the balance
and measure all inputs and outputs.
After the necessary measuring campaign we need to calculate the various heat items, e.g.:
Heat of fuel combustion
Combustibles in raw meal or exhaust gas
Heat of formation
Heat of evaporation
Sensible heats of all gas and mass flows
Radiation and convection heat
The heat balance shows clearly how the heat is spent among the individual items. This
information is therefore most suitable to detect abnormal operating conditions or potential for
improvements. As a next step we can decide which practical measures should be taken in
order to achieve improvements in the thermal energy consumption.
When improvements are realized we can often take into account that the saving of fuel is by
a factor of 1.3 to 1.5 higher than the primary improvements on the balance item
(multiplication factor), as long as we consider the high temperature zone.
Considerable improvements can be realized by reducing shell losses, false air inleaks, heat
exchange in certain preheater types and in clinker coolers.

1. INTRODUCTION
Heat balances on a kiln system can offer extremely useful information on the thermal
performance of the system. Heat balances show where or how the fuel heat is consumed,
based on the simple principle of:
input = output
Unnecessary energy losses can be easily detected.
The principle aim of this chapter is to serve a practical guide for doing heat balances on
cement kilns. It can be used as working paper and does neither require special
thermodynamic knowledge nor supplementary literature or tables.
For the more interested reader some special aspects are treated in a separate chapter.
The principle of heat balance may be easily transferred to other systems such as
preheaters, coolers and drying systems. Therefore the use of this chapter can be extended
to other systems than cement kilns.
In this chapter, only SI units are used, which means that heat is always given in kJ (kilo
Joule). Conversion calculations within the chapter will no more be required. Incidentally this
may also contribute to the consequent use of SI units.
2. OVERVIEW OF COMPLETE HEAT BALANCE PROCEDURE
2.1 Why to do a Heat Balance?

Various reasons or circumstances may cause a need for a heat balance measurement. The
following situations may justify a heat balance:
Performance test
Recording of kiln performance before/after a modification
Unusually high heat consumption or abnormal kiln operational data
Kiln optimization campaign
It may be self explaining that an extensive heat balance also costs money, especially if a
large number of people are involved. Therefore the costs may be put in relation to the
obtainable benefits. A potential improvement of say 100 kJ/kg of heat consumption on a
3000 t/d kiln means a savings in the order of US$ 200000 per year (at current fuel prices).
In such a case it may be worthwhile to invest some money for a detailed investigation
including a complete heat balance.
Although the specific heat consumption proper could also be determined by measuring
nothing but fuel heat and clinker production, a complete heat balance does offer
considerably more information and security. The consistency of the measured data is proved
much better, and the balance shows clearly where the heat is consumed. A heat balance is
obviously a very efficient tool for assessment of thermal efficiency.

2.2 How to Proceed
A heat balance does not only mean calculation of heat balance items. The complete
procedure usually includes the following steps:
2.2.1 1st Step: Preparation

The extent of works to be done depends on the completeness and reliability which is
desired. A careful planning and preparation is recommendable. The following basic items
must be clarified:
What has to be measured (kind and location of measuring / sampling points)?
Duration of test?
Frequency of measurements (continuous recording, spot measurements, time intervals,
etc.)?
Under above preconditions the number of people required and the necessary measuring
equipment can be determined. Temporary equipment may usually become necessary,
whereas the existing permanent instrumentation should be carefully checked and calibrated.
The following list may be used as checklist for a test preparation:
People available for test period?
Does everybody know what to do at what time?
Necessary logsheets for manual recordings ready?
Data recording system (electronic, pen recorder, etc.) available?
Flow of information among test team?
Calibration or checks of instruments and scales done (flowmeters, orifice plates,
venturis, dp-cells, thermocouples, etc.)?
Temporary measuring equipment available? Complete? Correct span? Functionning
properly?
For example:
pitot tube
U-tube manometer / electronic manometer
mobile thermometer
radiation pyrometer
gas analyzer / Orsat
sampling equipment for gas
Sampling procedure for solids (e.g. meal, dust, clinker) clarified?
Analyzing facilities?
2.2.2 2nd Step: Execution

An important precondition for a good test is a steady kiln operation. The test should only be
started if the system has reached a constant equilibrium state. During the test, variations of
operating parameters should be avoided.
For the measuring techniques reference is made to the corresponding chapter.

It is recommended to check completeness and reliability of measurements already during
the test, afterwards missing or uncertain information may create problems at the final
evaluation.
2.2.3 3rd Step: Evaluation

This step is the main scope of this chapter. Evaluation of data means to establish a heat
balance calculation according to the principle
input = output.
2.2.4 4th Step: Discussion

A heat balance as such must bring some practical conclusions otherwise it would be only of
academical interest. The following items may be considered:
Acceptable (normal) heat consumption?
Are heat balance items normal for given kiln systems?
Measures in order to improve heat economy of the system?
The last item can become quite an extensive work and the economical feasibility must be
considered as well. Such subjects, however, are beyond the scope of this chapter.
Note: Summarizing, it is obvious that the 3rd step Evaluation is only a limited part in the
whole context.
3. BASIS OF BALANCE AND REMARKS REGARDING EXECUTION
3.1 Balance Limits

In many balance reports the boundary for the balance is not shown expressively. As long as
standard cases are considered it may sometimes seem evident how the balance limits
were selected. Nevertheless it is advantageous to indicate always clearly where the
boundary for the balance is. This will avoid misunderstandings and reduce error possibilities.
For that purpose a (simplified) flowsheet is required. In this scheme the boundary must be
marked by a line which envelopes the system under consideration. When a measuring
campaign in a plant is planned such definitions should be made already in the planning
phase, i.e before the test.

Fig. 1 Examples of Balance Boundaries
Kiln + Cooler + Preheater Cooler only
boundary
boundary
Raw Mill
Kiln only
boundary
boundary
The boundary generates various cutpoints with ingoing or outgoing solid- and gas streams.
Likewise radiation and convection heat crosses the boundary.
Each cutpoint means a certain item in the heat balance because it represents a heat flow
either into or out of the system.
By this measure the system is clearly defined and nothing can be forgotten.
Basically, any shape of the boundary could theoretically be chosen. To give an extreme
example: The boundary could even cut a rotary kiln at half length! However, the boundary
must be selected according to practical considerations. This means that the cutpoints which
are generated must be:
easily accessible for reliable measurements
of practical interest in the whole context.

3.2 Guidelines for Test Duration
A long test duration would allow for a good accuracy but the available time is limited by
practical considerations.
As a rough guideline the minimum test duration should be about ten times the material
retention time in the complete kiln system:
test duration = 10 x retention time
Other factors like regularity of kiln operation also influence the test duration. If the process is
very unsteady longer times should be envisaged.
On a normal suspension preheater kiln the retention time of the material is in the order of
one hour. In contrast, on a lime shaft kiln retention times above 24 hours may occur.
Therefore the necessary test durations for these two cases must be completely different.
Although test durations must be set individually the following list may be used as rough
guideline:
Type of Kiln Test Duration (hours)

for Heat Balance
Suspension preheater kiln with precalciner 12
Suspension preheater kiln without precalciner 12 to 24
Long dry / wet kiln 24
3.3 Kiln Operation

During the test, the kiln must run at constant and steady conditions. Changing of setpoints
should be avoided whenever possible.
Interruptions have to be logged. If serious problems occur, the test has to be extended or
even postponed. Therefore it is often worthwhile to plan a certain time reserve.
From a theoretical point of view a proper balance can only be made if the system runs at
steady conditions. E.g. during heating-up heat is stored in the system and there is no
balance between input and output (input > output).
3.4 Kiln Data

In order to facilitate the final discussion it is usually necessary to collect the main data of the
system such as:
a) System
process
type of kiln
nominal capacity
type of preheater / precalciner
type of cooler
supplier
year of commissioning
fuel and firing system
type of burner nozzle
dust reintroduction system
dimensions of main equipment (sizes, inclinations, etc.)

data on fans, derives, etc.

b) Operation
various operating data (rpm, kW, temperature and pressure profiles along kiln
system, grate speed, undergrate pressures, etc.)
electric power readings (before / after test)
chemical analysis of raw meal, dust(s) and clinker, LSF, SR, AR, etc.
Above data are not necessarily required for heat balance calculations proper, but they
should be included in a complete balance report in order to describe the system and to give
more information on its performance.
4. HEAT BALANCE CALCULATIONS
4.1 General Remarks
4.1.1 Symbols and Units
A m2 area
CR W/m2K4 radiation constant
cP kJ/kg C specific heat (at const. pressure),
or kJ/Nm3 C specific heat capacity
CV kJ/kg calorific value
D m diameter
g m/s2 gravity constant
h kJ/kg heat content (specific)
or kJ/Nm3
or kJ/kg cli
L m length
m kg mass
or kg/kg specific mass
mf kg/h mass flow
Qf kW heat flow (1 kW = 1 kJ/s)
t C temperature (Centigrade)
T K temperature (Kelvin)
v m/s velocity
w kg/kg water content
Greek Letters
W/m2K heat transfer coefficient
- emissivity (for radiation)
W/m C heat conductivity
kg/m3 density

Dimensionless Numbers
Nu Nusselt number (for heat transfer)
Pr Prandtl number
R kiln feed (raw meal) / clinker-ratio
Re Reynolds number
Indices
conv convection
rad radiation
tot total
o ambient condition or zero condition
Conversion Factors
Length 1 inch 0.0254 m
1 ft 0.3048 m
Area 1 sq. ft 0.092903 m2
Volume, Volume Flow 1 cu.ft 0.028316 m3
1 cu.ft/min 1.699 m3/h (actual m3)
Mass 1 lb. 0.45359 kg
1 short ton 907.185 kg
(USA)
Pressure 1 bar 105 N/m2
1 mm H2O- 9.806 N/m2
Col.
1 atm. 1.013 bar
Energy 1 kJ 1000 J
1 MJ 1000 kJ
1 kWh 3600 kJ
1 kcal 4.187 kJ
1 BTU 1.055 kJ
Temperature Conversion C= 5/9(F - 32)
K= 273.15 + C
Heat Flow 1 kW 1000 W = 1 kJ/s
1 kcal/h 1.163 W
1 BTU/h 0.29307 W
Specific Heat 1 kcal/kg C 4.187 kJ/kg C = 4187 J/kg C
1 BTU/lb F 1 kcal/kg C = 4.187 kJ/kg C
Heat Transfer Coeffic. 1 kcal/m2h C 1.163 W/m2 C
1 BTU/ft2h F 5.678 W/m2C
Standard Conditions for Standard 0C and 1 atm (1.013 bar)
Gases Conditions
273.15 p(bar)
Nm3 =act.m3
273.16 + t(C) 1.013bar

4.1.2 Reference Temperature
We will set the usual reference temperature to 20C i.e. the sensible heat of mass flows at
20C becomes zero.
All heats of transformation such as combustion, evaporation, formation are also based on
20C reference.
4.1.3 Input / Output

Whether a heat item represents an input or output is determined by the direction of mass
flow, according to:
into boundary = input
out of boundary = output
This is a simple and useful convention (not a natural law).
If heats of transformation occur within the boundary the net heat effect is normally used as
criterion:
heat producing process = input
heat consuming process = output
A heat producing process is e.g. the fuel combustion (exothermic).
4.1.4 Reference Quantity

The heat balance is referred to 1 kg of clinker produced. This requires a general calculation
step (division by clinker production) which is not shown in the following formulas in order to
maintain a more simple presentation.
4.2 Determination of Clinker Production
4.2.1 General
The clinker production during a test is a key figure for all the following calculations. We will
refer all flows and heats to 1 kg of clinker produced.
Determination of clinker production can be made by direct weighing of the production which
is by far the best method. If this can be done the following chapter 4.2.2 can be ignored. But
in some cases no direct measurement is possible. Therefore, an indirect determination may
be necessary as shown in the following.

4.2.2 Indirect Determination
It is well known, that from approx. 1.56 kg raw material (not kiln feed!) 1 kg clinker can be
produced. Starting from this fact it seems to be easily possible to calculate the clinker
production. However, the above factor of 1.56 if of limited practical help because it can be
superimposed by dust return and depends on some other influences.
The best method in such a case is to establish a mass balance for the system. Similar to a
heat balance a boundary for a mass balance can be defined. Thus the clinker production
becomes:
Kiln Feed
Coal Ash Dust Losses

(Exhaust, gas,
Bypass etc.)
Dust Insufflated Clinker
Boundary for mass balance
+ kiln feed
+ coal ash
Clinker = + dust insufflated all calculated on L.o.I. free basis!
- dust losses

This principle is quite simple. The following additional remarks can be given:
The boundary for the mass balance does not necessarily have to coincide with the heat
balance boundary.
The mass balance boundary must cut the kiln feed measuring point (because the mass
flow is known at this point).
The balance is always made on a loss-on-ignition-free (L.o.I.-free) base (no balance for
the entire quantity including L.o.I. can be made since part of it is converted to gas, which
is not included here).

Regarding only the dust loss from the exhaust gas there are two basic possibilities:
a) Dust Flow crosses mass balance boundary
mass balance is influenced by dust flow

note that is not significant whether or how the dust is returned (outside the
boundary)!
b) Dust flow does not leave mass balance boundary (internal dust return)
mass balance is not influenced by dust flow

internal dust return must be steady, i.e. no storage being built up (silos!)
Two practical examples illustrating above two cases are shown in the following.

4.2.3 Examples
a) External Dust Recirculation (outside of boundary)
Fig. 2 Mass Balance / External Dust Recirculation
D=11t/h
Dust Flow boundary for mass balance
during comb.
Operation
Feed Exhaust
Raw
Material
Dust Flow during
direct Operation
Silo
Ash=1.8t/h
Raw Mill
KF=204t/h
BpD=2.1t/h Cli=?t/h
mf (dry) L.o.I. mf (1-L.o.I.)

[t/h] [-] [t/h]
Kiln Feed (KF) 204.0 0.357 131.17
Coal Ash (Ash) 1.8 0 + 1.80
Dust in Exhaust (D) 11.0 0.315 - 7.54
Bypass Dust (BpD) 2.1 0.070 - 1.95
Clinker (Cli) 0 123.48 t/h
Note that the kiln feed data (L.o.I.) may slightly change when going from combined operation
to direct operation or vice versa!
Kiln Feed / Clinker = R = 204.0 / 123.48 = 1.652 (all included)

b1) Internal Dust Recirculation (inside of boundary)
Fig. 3 Mass Balance / Internal Dust Recirculation
Stack
Exhaust boundary for mass balance

Feed
No Feed
during direct OP.
Dust Return
D=11t/h
Silo
Ash=1.8t/h
Raw Mill
KF=204t/h
BpD=2.1t/h Cli=? t/h
During direct operation the following calculation method applies:
mf (dry) L.o.I. mf (1-L.o.I.)

[t/h] [-] [t/h]
Kiln Feed (KF) 204.0 0.357 131.17
Coal Ash (Ash) 1.8 0 + 1.80
Bypass Dust (BpD) 2.1 0.070 - 1.95
Clinker (Cli) 0 131.02 t/h
Kiln Feed/Clinker = R = 204.0/131.02 = 1.557

Note the difference to example a)!
Incidentally the example b1) is shown for comparison to example a) and does by no means
represent a superior solution of the dust return!
The following example b2) shows a very similar way of calculation to example b1).

b2) Internal Dust Return (wet kiln)
Fig. 4 Mass Balance / Internal Dust Recirculation
boundary for mass balance
D=8.6t/h
Coal Ash=0.8t/h
KF= 42t/h Slurry (37% H2O)

= 26.48% t/h dry Solids
Cli=?
mf (wet) mf (dry) L.o.I. mf (1-L.o.I.)

[t/h] [t/h] [-] [t/h]
Kiln Feed (KF) 42 26.46 0.360 16.93
Coal Ash (Ash) - 0.8 0 0.8
Clinker 0 17.73 t/h

4.3 Heat from Fuel
4.3.1 Fuel Firing

Fuel can be introduced at the following locations:
Kiln firing
Secondary firing or precalciner
Burnable components in raw meal
Generally the heat from fuel is calculated:
h = m CV (kJ/kg cli)
h = heat (kJ/kg cli)
m = specific fuel consumption (kg/kg cli or Nm3/kg cli)
CV = calorific value (kJ/kg fuel or kJ/Nm3 fuel)
For the calorific value CV only the low (or net) value is used within the Holcim Group (for
conversion see chapter Combustion Engineering).
Heat balance referring to high (or gross) heat value are sometimes encountered. But here an
important note must be made: the use of high heat values is based on a quite different
method of considering heat losses. Therefore those balances cannot be compared directly
to our type of balance.
4.3.2 Coal Firing

For coal firing it is essential that the CV and the consumption (m) refer both to the same
conditions. The normal convention is to refer to the state as fed to the kiln burner.
The state as fed to burner may sometimes be different from the state as analyzed.
Differences can occur because of dust addition in coal mill, loss of volatile matter or just
simply by changed moisture conditions.
If the actual CV at the given coal moisture w is not known it can be calculated from the value
CVo at the moisture wo:
CV = [(1-w) / (1-w0)] (CV0 + w0 r) - w r

where: r = 2450 kJ/kg
= heat of water evaporation at reference temperature 20C
w, wo = weight fractions of water

Examples:
a) Coal, CV at 5% moisture = 26500 kJ/kg
CV at 1% moisture (as fed to burner) = ?
CV = [(1-0.01) / (1-0.05)] (26500 + 0.052450) - 0.012.450
CV = 27464 kJ/kg
If above coal (1% moisture) is fired at a specific rate of 0.1230 kg/kg cli:
h = 0.1230 27464 = 3378 kJ/kg cli

b) If the CV for dry matter is known (CVdry) the general formula can be simplified (for wo =
0):
CV = (1 - w) CVdry - w r
CVdry = 28 MJ/kg, moisture 2%
CV = 28 (1 - 0.02) - 0.02 2.45 = 27.39 MJ/kg
4.4 Burnable Components in Raw Material
4.4.1 Organic Matter

Organic matter can be present in the raw meal e.g. in form of oil shale or exceptionally even
in form of free crude oil.
Such material is partly volatilized in the preheating zone and leaves the system partly as
unburnt light hydrocarbons. If the latter are actually measured and considered in the total
heat balance it is recommended to determine the calorific value of the raw meal (not directly,
but by extracting its organic matter). Thus the heat input becomes:
h = CVlow R (kJ/kg cli)
where:
CVlow = calorific value due to organic matter, referred to 1 kg raw meal
R = kiln feed / clinker - factor
If light hydrocarbons in the exhaust gas are not measured nor considered in the total heat
balance anyway the following approach will produce better results:
Determine the organic content only (org. C, measured by burning in pure O2 to CO2). Then
consider exclusively the organic C (the rest can be neglected). The approximate heat input
becomes:
h = C R 33000 kJ/kg (kJ/kg cli)
where
C = organic carbon content in raw meal
Above mentioned is only an approximation but often used due to its simplicity.
Example:
R = 1.6 kg/kg cli, C = 0.2% carbon
q = 0.002 1.6 33000 = 106 kJ/kg cli

4.4.2 Inorganic Matter
In certain cases residues of non-oxidized pyrite (FeS2) can occur in the raw meal. The heat
input becomes:
h = S R 12930 kJ/kg (kJ/kg cli)
where
S = weight fraction of sulfur (expressed as S!) from pyrite in raw meal
Example:
R = 1.6 kg/kg cli, 0.05% S (from pyrite) in raw meal
h = 0.0005 1.6 12930 = 10 kJ/kg cli
The practical heat effect in the above case is only marginal.
4.5 Loss due to Incomplete Combustion

If unburnt gases such as CO, H2, CH4 occur in the exhaust gas an additional heat output
occurs.
The loss can be calculated to:
h = m (CO 12640 + H2 10800 + CH4 35840) (kJ/kg cli)
3
m = specific gas quantity (Nm /kg cli)
CO, H2, CH4 = volume fractions in exhaust (referred to wet gas)
In most cases only CO is measured and then the calculation becomes:
h = m CO 12640 (kJ/kg cli)
Example
m = 1.50 Nm3/kg cli (SP kiln)
CO = 0.2% (ref to wet)
q = 1.50 0.002 12640 = 38 kJ/kg cli
4.6 Heat of Formation

The heat of formation takes into account all main reactions which occur when raw meal is
transformed to clinker, as follows:
RAW MEAL
Heat of dehydration of clay (endothermic)
Heat of decarbonation of CaCO3 + MgCO3 (endothermic)
Heat of formation of clinker (exothermic!)
CLINKER
In most of the practical cases it is sufficient to assume a constant value of
h = 1750 kJ/kg cli
The value mentioned before represents a heat output since the overall reaction is
endothermic (heat consuming). The main contribution to the overall value comes from the
decarbonation of CaCO3 (approx. 2100 kJ/kg cli).
The heat of formation may naturally have some variations from one raw meal to the other.
But due to the narrow range which is specified for the cement clinker composition no major
deviations (of say more than +/- 50 kJ/kg cli) have to be expected in normal cases.

If the CaO in clinker does not originate from CaCO3, but from raw material sources
containing less CO2 than the CaO balance (such as coal ash or partly decarbonized raw
mix) the heat of formation would theoretically change. However, this problem can be easily
avoided by the following definition:
All mass flows (as kiln feed, dust losses, coal ash) which carry a certain amount of CaO
not coming from CaCO3 are considered as heat flows according to their non-
carbonatic CaO content (see chapter 4.7).
By following the above definition it is not necessary to make alterations at the heat of
formation. An abnormal situation can simply be treated by creating additional inputs or
outputs, but not by changing the standard heat of formation.
The procedure is based on the idea of an ideal clinker burning from pure carbonates. If
those ideal conditions do not apply we set corrections in the corresponding heat inputs or
outputs. Such principles are well known from calculation of bypass CaO-losses.
Additional information on the heat of formation is given in chapter 6.2.
4.7 Heat due to Partly Decarbonized Material

Decomposition of carbonates in an essential heat effect in a cement kiln. Usually one thinks
of the two extreme cases of either complete presence of carbonates (CaCO3, MgCO3) or
complete absence of carbonates (e.g. clinker). But there can also be cases in between
which influence the heat balance.
Material streams can transport non-sensible heat due to partly decarbonized material (CaO)
or non-carbonatic CaO. The following possibilities exist:
1) raw meal (containing e.g. CaO free or non carbonatic, CaO-bearing minerals)
2) exhaust gas dust (partly decarbonized)
3) bypass dust (largely decarbonized)
4) coal ash
The heat can either be negative or positive according to the direction of flow (ingoing or
outgoing).
If item 2) (exhaust gas) is considered in the balance it is essential to consider also item 1)
(raw meal)! In many cases kiln dust (which contains some free CaO) is returned to the kiln
but not necessarily at exactly the same rate as the outcoming dust (e.g. on a system
combined with raw mill, alternating between combined and direct operation). Even if dust is
returned the net heat effect of item 1) minus item 2) may not automatically be zero!
If the CaO and CO2 is known from chemical analysis the non carbonatic part CaOnc can be
calculated:
CaOnc = CaO - (56 / 44) CO2
Strictly speaking, above formula is only correct if no other carbonates than CaCO3 occur. If
MgCO3 is present in form of carbonates the carbonatic MgO has to be taken into account:
CaOnc = CaO - 56 [(CO2/44) - (MgO/40)]
Thus the heat is calculated to
h = m CaOnc 3150 (kJ/kg cli)
m = spec. mass (kg/kg cli)
CaOnc = non-carbonatic CaO, expresses as weight fraction

Above formula does not consider the small possible influence of MgCO3 decarbonation heat.
If the assumption can be made that all MgO occurs in non-carbonatic form (e.g. after a heat
treatment in the 700C range) the following improved formula can be applied:
h = m (CaOnc . 3150 + MgO . 2710) (kJ/kg cli)
Examples
1) Raw Meal and Kiln Dust
(dust returned, measurements refer to balance limits, normal raw mix)
Raw Meal Kiln Dust

R = 1.65 kg/kg cli m = 0.09 kg/kg cli
CaO = 42.3% CaO = 43.5%
MgO = 0.9% MgO = 1.1%
CO2 = 34.0% CO2 = 30.5%
CaOnc = 42.3-56 [(34/44) - (0.9/40) CaOnc = 43.5-56 [(30.5/44) - (1.1/40)

= 0.3% = 6.2%
h = 1.65 0.003 3150 h = 0.09 0.062 3150

= 16 kJ/kg cli (input) = 18 kJ/kg cli (output)
In the above example the net heat effect is virtually zero and may be completely neglected.
But this may not be used as a general rule as shown by the next example.
2) Raw Meal and Kiln Dust

(similar to example 1), but raw mix containing a major proportion of non-carbonatic CaO)
Raw Meal Kiln Dust

R = 1.60 kg/kg cli m = 0.09 kg/kg cli
CaO = 43.2% CaO = 43.5%
MgO = 0.9% MgO = 1.1%
CO2 = 30.0% CO2 = 30.5%
CaOnc = 43.2-56 [(30/44) - (0.9/40) CaOnc = 43.5-56 [(30.5/44) - (1.1/40)

= 6.28% = 6.2%
h = 1.60 0.0628 3150 h = 0.09 0.062 3150

= 317 kJ/kg cli (input) = 18 kJ/kg cli (output)
Although non-carbonatic CaO occurs seldom, it may largely contribute to the heat balance
(in the latter example a net effect of 299 kJ/kg cli!). The practical problem, however, is to
determine the CaOnc with sufficient precision.

3) Bypass Dust
m = 0.04 kg/kg cli (quantity of bypass dust)

CaO = 56.2%
MgO = 1.0%
CO2 = 1.8%
CaOnc = 56.2 - (56/44) * 1.8 = 53.9%
q = 0.04 * 0.539 * 3150 + 0.01 * 2710 = 95 kJ/kg cli
(output)
4) Coal Ash
m = 0.02 kg/kg cli (quantity of ash)

CaO = 21%
MgO = 2%
q = 0.02 (0.21 * 3150 + 0.02 * 2710) = 14 kJ/kg cli
(input)
4.8 Heat of Evaporation

If water is evaporated within the balance limits the heat of evaporation becomes
h = m 2450 (kJ/kg cli)
m = evaporated water (kg/kg cli)
Above items means an output. A large source of water evaporation is usually the slurry feed
to a wet kiln.
Although the calculation of heat of evaporation is obviously simple a few notes are given:
Only the free water of the kiln feed is considered (the hydrate water is already included in
the heat of formation!).
Water evaporation can also be caused by water injection into preheater, kiln, cooler, etc.
The water evaporated from the fuel must not be included (this effect is already included
in the net heat value, provided it has been determined as described in paragraph 4.3).
Examples
a) Wet Kiln: slurry water content = 35%
R = 1.56 kg/kg cli (dry)
m = 0.35 / (1-0.35) 1.56 = 0.84 kg/kg cli
h = 0.84 2450 = 2058 kJ/kg cli
b) Water Spray into Planetary Cooler

m = 0.05 kg/kg cli
h = 0.05 * 2450 = 123 kJ/kg cli

4.9 Sensible Heat
4.9.1 General
Generally the sensible heats are calculated as follows:
h = m cp (t - 20C) (kJ/kg cli)
m = specific mass (kg/kg cli or Nm3/kg cli)
cp = average specific heat (kJ/kg C or kJ/Nm3 C)
t = temperature of m (C)
Above formula uses a reference temperature of 20C, i.e. sensible heats of material and gas
flows at 20C are zero.
The value h (kJ/kg cli) can either be positive (if t > 20C) or negative (if t < 20C).
Whether h represents an input or an output (see 4.1) depends only on the direction of the
flow m:
if m = entering boundary h = input
if m = leaving boundary h = output
Basically the value m can be expressed as kg or Nm3. It is then logical that the cp values
used must also refer to same unit (per kg or per Nm3).
For convenience we use the following convention:
for solid flows unit = kg
for gaseous flows unit = Nm3 1)
1)
The application of the unit kg for gases would also be thinkable and would even
facilitate the mass balance calculations. But since Nm3 is well established in the cement
industry and since the cp-diagrams refer to Nm3 only, Nm3 is used in this chapter.
The necessary cp values are to be found in the diagrams Fig. 5 and 6 (for solids and fuels)
and Fig. 7 (for gases).
The same basic mode of calculation of sensible heat can be adapted for all material and gas
flows crossing the boundary such as:
Raw meal
Clinker
Dust(s)
Exhaust gas, bypass gas, waste gas, etc.
Cooling air, false air, etc.

Fig. 5 Cp of Solids

Fig. 6 cp of Liquids and Fuels
Temp. cp of Liquid Water
0C 4.22 kJ/kg C
50C 4.18 kJ/kg C
100C 4.22 kJ/kg C
Cp of Fuels
2.7
2.5 Ref. 20C

g /l
0.8 k
g/l
2.3 0.9 k
g /l
1.0 k
OIL
2.1
kJ / kg C
1.9
1.7
1.5 lati les

40% Vo
30%
1.3 COAL 20%
10%
1.1 0%
0.9
0.7
0 50 100 150 200
Temperature [C]

Fig. 7 cp of Gases
4.9.2 Calculation of cp of Mixtures

The cp values are usually tabulated for pure substances. If a mixture is present it may
become necessary to calculate the cp starting from the given composition. This is better
shown in a practical example (instead of a general formula):
Example:
Exhaust gas of suspension preheater kiln at 360 C
Measured by Orsat (dry):
CO2 = 27.2% (CO = O)
O2 = 4.3%
N2 = 100 - 27.2 - 4.3 = 68.5%
From moisture determination:
H2O = 0.08 Nm3/Nm3 wet

What is the cp value of the mixture (wet gas)?
Dry Gas Wet Gas cp (pure cp x Vol. Frac.

(Vol %) (Vol %) (kJ/Nm3C) (kJ/Nm3C)
CO2 27.2 25.0 1.92 0.480
O2 4.3 4.0 1.37 0.055
N2 68.5 63.0 1.31 0.825
H2O 0 8.0 1.55 0.124
Total 100% 100% - 1.484 1.48
Result: cp of mixture = 1.48 kJ/Nm3C (at 360C)

The above value is typical for a normal SP kiln. We will always find values in the 1.5
kJ/Nm3C range.
4.10 Heat Loss due to Radiation and Convection
4.10.1 General
In practice it is quite convenient to treat both, radiation and convection heat transfer
together. Although the physical laws of these two heat transfer phenomenas are different
they are usually given as a total.
The determination of total heat transfer coefficients by using simplified working diagrams will
be accurate enough. Therefore the theory how to calculate the values is treated in separate
chapters (6.3 and 6.4). The user of this chapter does usually not need to go into details of
radiation and convection heat transfer theories. In addition, the calculation from the complete
theory would hardly increase the practical accuracy, since all these theories contain a
certain inaccuracy.
4.10.2 Radiation Heat Transfer

Radiation heat transfer depends on the surface temperature and the emissivity (0...1).
Typical values for :
Type of Surface Temperature C

Rough oxidized steel 100 1
Rough oxidized steel 400 0.9
White lime paint (on steel) 400 0.8
Aluminium paint 100 0.2 to 0.4
Aluminium, rolled sheet metal 100 to 500 0.08
Aluminium, polished or bright foil 100 to 500 0.04
In most cases an value has to be set at the pyrometer used for the measurement
(depending on the type).
For pyrometers giving readings for a constant emissivity = 1. Fig. 8 can be used for
correction.

The pure radiation heat transfer coefficient rad (W/m2C) is shown in Fig. 9.
Note that the temperature axis in Fig. 9 means temperature difference to ambient (same
convention as for the following Fig. 10 and 11 - see paragraph 4.10.3 and 4.10.4).
Radiation heat transfer calculation does usually not create special problems as long as the
radiating surfaces face freely towards ambient. A particular situation occurs on a planetary
cooler where certain directions of radiation are shielded. That means not all areas are able
for radiation towards the ambient.
Here, it would be obviously wrong to consider the total of the tube surfaces as radiation
area. As a first approach the outer enveloping cylinder area (enveloping all tubes, see
sketch) can be taken as reference area.

Fig. 8 Relation between Emissivity and True and Apparent Surface
Temperature

Fig. 9 Radiation Heat Transfer Coefficient

4.10.3 Convection Heat Transfer
Two different types of convection can be distinguished:
Free convection (occurs by natural thermal draft, at low wind velocities)
Forced convection (occurs at high wind velocities)
In practice an intermediate region between these two extreme cases can be encountered,
depending on the wind velocity v (m/s). The resulting conv can be taken from Fig. 10.
Fig. 10 Convection Heat Transfer Coefficient
The diagram Fig. 10 refers to a normal kiln diameter range from 3 m to 6 m.

Note that the temperature axis in Fig. 10 means temperature difference to ambient, not
temperature itself.

Influence of diameter
Free convection (v = 0 m/s):
The free convection coefficient is not depending on the diameter (or height) of
equipment. If there is no wind Fig. 10 can be applied for any dimension.
Forced convection:
In contrast to the free convection the forced convection does depend on the
diameter. For this reason the application range of Fig. 10 is limited to 3 m...6 m,
but only if high wind velocities occur. At low wind velocities no diameter limits
exist.
For very small tubes (or gas ducts) the following corrections of conv (increases)
can be made:
For 2 m + 15%
For 1 m + 30% only at high wind velocities (min. 2 m/s)
For 0.5 m + 50%
4.10.4 Total Heat Transfer Coefficient (Radiation and Convection)

For the determination of the total heat transfer coefficient tot (W/m2C) two cases are
possible:
a) Standard Case
= 0.9 (rough oxidized steel surface)
This particular case is rather simple. Direct use of Fig. 11 is possible.
(Go into diagram with temperature difference and read out the total heat transfer coefficient.)
b) Non-standard Case
differs clearly from 0.9 and/or correction of convective heat transfer coefficient is
necessary.
Make use of Fig. 9 and 10.
tot = rad + conv
(Caution: never add free convection + forced convection, Fig. 10 does already include the
overlapping of both effects.)

Fig. 11 Radiation and Convection Heat Transfer Coefficient (total)

4.10.5 Total Heat Flow
The total heat flow from radiation and convection heat transfer is calculated according to:
Qf = tot A (t - t0) (W)
Qf = heat flow (W); 1kW = 1000W
tot = rad + conv = total heat transfer coefficient (W/m2C)
A = shell area (m2)
t = shell temperature (C)
t0 = ambient temperature (C)
From the heat flow Qf the specific loss can be calculated:

h = Qf (kW) / mf (t/h) 3.6 (kJ/kg cli)
mf = clinker production (t/h)
3.6 = conversion factor, because of unit (t/h) for mf
If the temperature of a shell area is not approximately constant it is necessary to subdivide
the area into individual sections.
On a rotary kiln we may e.g. select about 10 (or more) individual cylindrical areas, each of
them having the surface:
Ai = Di Li (m2)

Example:
Rotary kiln
Clinker production = 1900 t/d = 79.17 t/h
4.4 m x 67 m
Average emissivity: = 0.9
Ambient temperature = 20C
Average wind velocity v = 1 m/s (nearly free convection!)
Shell temperature profile as indicated in the following table:
Length Element Dia- Element Surface Qf

1) 2)
Position Length meter Area Temp. total (heat flow)
2 2
(m) (m) (m) (m ) (C) (-) (W/m C) (W)
0 - 5 5 4.40 69.1 230 0.9 22.7 331000
5 - 10 5 4.40 69.1 360 0.9 32.9 774000
10 - 15 5 4.40 69.1 310 0.9 28.6 574000
15 - 25 10 4.40 138.2 220 0.9 22.1 611000
25 - 35 10 4.40 138.2 330 0.9 30.3 1299000
35 - 45 10 4.40 138.2 260 0.9 24.9 825000
45 - 55 10 4.40 138.2 290 0.9 27.1 1011000
55 - 60 5 4.40 69.1 270 0.9 25.6 442000
60 - 67 7 4.40 96.8 220 0.9 22.1 428000
Total 67 m 926.1 m2 6295000 W
Total heat loss = 6295 kW

6'295 3.6
Specific heat loss = = 286 kJ/kg cli
79.17t / h
1)
A = . D . L (cylinder)
2)
radiation plus convection (see Fig. 11)

5. HEAT BALANCES
5.1 General
The results for heat inputs and outputs as calculated according to the preceding paragraph 4
are summarized in a balance sheet.
Table 1 (complete kiln system) and Table 2 (cooler only) are given as examples. They may
also be used as checklist for completeness of own measurements and calculations.
Rest-item:
A real balance (from complete date) will always include an item rest, where all
measuring errors (no measurement is 100% accurate) and non-considered items are
included. The amount of the rest item should not exceed 3% of the total heat input.
Heat consumption:
The actual heat consumption is the total of fuel input(s) which is not exactly the same
as the total of heat inputs. If there are more than one fuel input, it is advantageous to
indicate the actual (true) heat consumption separately.
Cooler balance:
On a cooler, reliable measurement of secondary air heat is virtually impossible.
Therefore, this value is determined by balance calculation and no rest item is given.
The thermal efficiency of the cooler is usually defined as:
= heat of secondary (and tertiary) air (%)
heat of hot clinker
The % values in the balance sheet, however, are based on the total of inputs. This
can cause a slight difference from the percentage of above .

Table 1 Heat Balance of Cement Kiln System
(General case)
- All referred to 1 kg clinker
- Reference temperature = 20C
- Ambient temperature = ...... C
- Production = ...... t/h
- Specific heat cons. = ...... kJ/kg cli
Specifications Temp. Heat

(kg/kg cli), (C) (kJ/kg (%)
(Nm3/kg cli) cli)
(kW) etc.
INPUT
Fuel combustion
- primary firing -
- secondary firing -
Burnable matter in kiln feed -
Raw meal: sensible heat
Fuel: sensible heat
Primary air: sensible heat
Cooler air: sensible heat
CaO (non-carbonatic) in kiln feed -
Total of inputs - 100%
OUTPUT
Heat of formation - -
Water evaporation:
- kiln feed -
- water spray(s) -
Exhaust gas:
- sensible heat
- dust CaO-loss -
- unburnt gases (CO, etc.) -
Cooler: - waste air sensible heat
- middle air sensible heat
- clinker exit sensible heat
Bypass losses: - sensible heat
- dust sensible heat
- dust CaO-loss -
- unburnt gases (CO, etc.) -
Radiation and Convection: kW
- preheater kW
- rotary kiln kW
- cooler kW
- tert. Air duct kW
Rest
Total of outputs - 100%

Table 2 Heat Balance of Clinker Grate Cooler
All referred to 1 kg clinker

Reference temperature = 20C
Ambient temperature = ... C
Specifications Temp. Heat

(C) (kJ/kg (%)
cli)
INPUT
Hot clinker
Cooling air
Total input 100%
OUTPUT
Secondary air (incl. dust)
Tertiary air
Middle air
Waste air
Clinker outlet
Water spray
Radiation and convection
Total output 100%
5.2 Examples of Heat Balances of Various Kiln Systems

Heat balances are given in Table 3 of various kiln systems. While comparing these it has to
kept in mind, however, that kiln systems cannot only be judged based on these heat
balances. The heat consumption of kiln system may depend on many items, those of major
influence are:
Kiln size, i.e. production rate
Heat loss due to radiation and convection
Secondary air quantity and temperature
Operating conditions of kiln
Furthermore related systems have also to be taken into account. Consider a grate preheater
kiln (Lepol kiln):
Its specific heat consumption may almost be as low as that of a 4-stage preheater kiln.
However the grate preheater kiln, the semi-dry type, requires that additional heat is spent for
material drying, whereas the exhaust gas of the 4-stage preheater kiln may be utilized to dry
up to 8% water content of the raw material.

The table 3 gives heat balances of three different kiln systems.
Wet kiln:
The wet kiln has a production capacity of approximately 3000 t/d. The slurry water
content is 38%.
Lepol kiln:
The heat balance of the Lepol kiln is of a comparatively small unit. The nodule
moisture content is 12%. In general grate preheater kilns of the semi-dry type have a
heat consumption only slightly higher than the 4-stage preheater kilns.
Dry suspension preheater:
The heat balance of the 4-stage preheater kiln is typical for a unit in the 2000 to 3000
t/d range.
Table 3 Heat Balances of Wet, Grate Preheater and Preheater Kilns

WET PROCESS SEMI-DRY DRY PRE-
(LEPOL) HEATER
(4-STAGE)
kJ/kg cli % kJ/kg % kJ/kg %
cli cli
1. INPUT
FUEL
from sensible heat 25 0.4 15 0.4 13 0.4
from combustion 5560 96.7 3343 97.6 3150 97.7
RAW MEAL
from sensible heat 25 0.4 30 0.9 54 1.7
from sensible heat of water 71 0.2 17 0.5 - -
COMBUSTION AIR
from sensible heat of all the air 67 1.2 20 0.6 6 0.2
supplied (prim. sec.)
TOTAL INPUT 5750 100 3425 100 3223 100
2. OUTPUT
Heat of formation 1750 30.4 1750 51.1 1750 54.3
Evaporation of water from raw 2370 41.2 506 14.8 13 0.4
meal
Exhaust gas sensible heat 754 13.1 314 9.2 636 19.7
Dust sensible heat 25 0.4 21 0.6 18 0.6
Incomplete combustion (CO) - - - - - -
Clinker exit temperature 59 1.0 50 1.5 63 2.0
Cooler exhaust gases 100 1.7 276 8.1 423 13.1
Losses due to radiation and 540 9.4 452 13.2 297 9.2
convection
Water cooling (Recupol inlet - - 42 1.2 - -
chute)
Rest 152 2.6 14 0.4 23 0.7
TOTAL OUTPUT 5750 100 3425 100 3223 100

6. SPECIAL PART
6.1 Influence of Reference Temperature

The reference temperature for a heat balance is usually set to 20C. At this temperature all
sensible heats become zero. This is quite practical because the normal ambient temperature
is always near 20C.
In addition the choice of the reference temperature has various other consequences which
are normally not mentioned expressively. As long as the calculation procedure according to
this chapter is applied we will not run into any practical problems. But if we want to go a bit
deeper inside the matter we are soon confronted with some special questions as will be
shown in the following.
6.1.1 Heat of Transformation

By the term heat of transformation we summarize here all heat effects by chemical
reactions and changing of the physical state (combustion, heat of formation, evaporation). In
order to understand the meaning of the reference temperature on these heat effects the
following diagram is shown. It gives an example of an (ideal) isothermal process, including
upheating and cooling.
Fig.12
Temperature
heat of transformation
at true reaction
temperature
True reaction
temperature
n
d ow
up
g
in
ling
at
he
c oo
Starting Ending
point point
Reference
temperature 20C
heat of transformation
at reference temperature
heat content
We note from the diagram that the heat of transformation at reference temperature and at
true reaction temperature are not identical. The reason is the different specific heat content
of the starting material and the reaction products. It is also obvious that any other level for
the reference temperature would basically change the heat of transformation.

The heat of transformation at a reference temperature 20C should therefore be seen in its
proper sense: It is the heat effect under the convention, that we start at 20C and end
exactly at those 20C. The real process (heating up, transformation, cooling) has no
influence on the result, regardless of the true temperatures. This is a direct consequence of
the principle of energy conservation.
Therefore, we may take into account that all heats of transformation which we use in a heat
balance are based on reference temperature 20C. Such values should not be confused with
the heat effect at the true temperature of transformation or any other temperature.
The items of interest are:
Calorific value (combustion heat):
The CV must also be based on 20C reference. The error which occurs when
choosing a reference of e.g. 0C is fortunately so small that it is within the normal
accuracy of a CV determination.
Heat of formation:
Regarding the heat of decarbonation only we may sometimes find values for the true
reaction temperature (decarbonation, in the 800C range). Such values shall not be
used for a heat balance at 20C reference, because this would produce an essential
error!
Heat of evaporation of water:
Note that we must use the evaporation heat at 20C (2450 kJ/kg) in the balances and
not the value at say 100C (2260 kJ/kg). The effects of upheating of water or vapor
are automatically considered by the normal balance calculation procedure (items for
sensible heats).

6.1.2 Specific Heats
The specific heats used in this chapter are average values which are consequently based on
20C reference temperature. We may illustrate this fact by an upheating process starting at
20C as shown in the following graph:
Fig.13
heat content (kJ/kg)
inclination
P = av.spec.heat (kJ/kg C)
h
sensible
heat content
P
0 Temperature
20C t - 20C t
The average specific heat can be graphically explained as the inclination of the straight line
from Po to P. It is obvious that this inclination depends from the reference point Po at 20C.
At the point P the tangent to the heat curve is also shown by a dotted line. The inclination of
this line is the actual specific heat at the temperature t. This value is different from the
average specific heat. Moreover, it does not depend from any reference point.
As a practical consequence we should never use specific heats from any source in the
literature without checking what they really mean. In most cases actual values are given and
not averages. Therefore, an integration or averaging would be required before we can use
them for calculation of heat contents. Special caution should be made in the high
temperature range where actual value and average value may be considerably different! In
the 20 to 200C range the practical differences, however, are rather small.

6.2 Heat of Formation
6.2.1 General
The heat of formation expresses the theoretical heat required for producing 1 kg of clinker.
The following steps are taken into account:
Approx. Heat Effect

Temperature
Step 1 Expelling of hydrate water (+ 500C negative (heat
transformation of clay minerals) consumed)
Step 2 Decarbonation of MgCO3 (first) 850C negative (heat
and CaCO3 (CaCO3 CaO + consumed)
CO2)
Step 3 Formation of clinker minerals 1400C positive (heat
(C2S, C3S, C3A, C4AF) produced)
The net heat produced by the overall reaction is negative i.e. heat is consumed (listed as
output).
The heat formation is defined at 20C. In other words it expresses the theoretical amount of
heat to transform raw meal at 20C to 1 kg clinker at 20C (if no heat losses would occur). It
is therefore also considered as the theoretical minimum heat required for burning clinker.
The heat of formation is not at all depending on the way of the actual reaction (i.e.
temperature versus time). The only precondition of course is that the reactions (step 1, 2, 3)
have really been completed.
Although the basic principle of the heat of formation seems quite simple a few conventions
(or definitions) are necessary here:
1) The hydrate water is expelled is considered as vapour (therefore no additional heat of
evaporation has to be introduced into the balance!).
2) As a standard case the assumption is made that the CaO is present as CaCO3 in the
raw meal (in non carbonatic CaO occurs we do not change the heat of formation, but add
heat inputs according to paragraph 4.7).
3) Burnable components in the raw mix are considered separately as inputs (see paragraph
4.4.2).
6.2.2 Calculation of Heat of Formation
6.2.2.1 Introduction
It is difficult to determine the heat of formation directly be calorimetric measurements.
Therefore calculation methods have been developed in order to take into account the
varying influences of raw mix properties.
Although formulas are given the following they should be applied with care. As long as the
raw mix does not clearly deviate from the standard value of 1750 kJ/kg the calculated
figures must not necessarily give more accurate results than the 1750 kJ/kg.
Even extensive formulas for heat of formation contain a certain incertitude due to the partial
lack of accurate thermodynamic standard values (for the minerals which are involved).

6.2.2.2 Calculation Formulas
The following formulas are partly based on the work of H. zur Strassen (Lit. 1) however a few
recent adaptations have been made:
We will use the following notation:
CaO, MgO, SiO2, Fe2O3, Al2O3 = Clinker analysis, expressed as weight fractions
(kg/kg cli)
H2O (kg/kg cli) = Hydrate water in raw material, referred to kg clinker (not to raw meal)
The following general formula for the heat of formation applies:
h = CaO 3200 + MgO 2710 - SiO2 2140 - Fe2O3 250 + hRes (kJ/kg cli)
The last contribution hRes depends on the Al2O3 and H2O (hydrate) content. The calculation
of hRes depends on the information which is available on hydrate water and/or type of clay.
Usually, the more information is known the better the result will be.
Case Hydrate Type of Formula for hRes (kJ/kg cli)

No. Water Clay
(H2O)
1 not known not known hRes = Al2O3 . 1720
2 known not known hRes = Al2O3 . 120
+ H2O . 5520
3 not known known hRes = (Al2O3)K . 2220
+ (Al2O3)M . 1310
+ (Al2O3)I . 1640
4 known known hRes = (Al2O3)K . 1400
+ (Al2O3)M . 620
+ (Al2O3)I . 760
+ H2O . 2450
In case 3 and 4 a distinction has to be made from which type of clay the Al2O3 originates.
The indices K, M, I denote:
K = Kaolinite
M = Montmorillonite
I = Illite
Note on the formulas for hRes:
No. 1:
For general purposes, if no information on hydrate water available.
No. 2:
Takes into account the actual hydrate water content. It is also a very good approach if the
hydrate water does not only originate from clays but for example from Ca(OH)2.
Note that the coefficient for H2O (hydrate) is more than twice the normal heat of
evaporation 1)

Nos. 3 + 4:
Take into account the type of clay. (Hydrate water must originate from clay only, otherwise
take No. 2).
When the CaO is introduced into the formula it is not required to know whether the CaO
results from CaCO3 or not. If any non carbonatic CaO is entering into the system we will take
into account a balance heat input by definition (see paragraph 4.7) and therefore the
calculated heat of formation is not affected.
1)
The older formula from H. zur Strassen did allocate only 2450 kJ to H2O, but instead
930 kJ to the Al2O3, which made it difficult to extrapolate it for other hydrates than from clay.
It is certainly more logical to refer the heat to the H2O, which is actually responsible for the
heat consumption.
Examples
1) - Clinker composition:
CaO = 64.8 %
MgO = 1.2 %
SiO2 = 22.6 %
Fe2O3 = 3.3 %
Al2O3 = 5.2 %
SO3 = 0.52% TiO2 = 0.29%

Na2O = 0.08% Mn2O3 = 0.03%
K2O = 1.20% L.o.I = 0.54%
- No information on hydrate water in raw meal available
h= 0.648 3200 + 0.012 2710 - 0.226 2140 - 0.033 250
+ 0.052 1720 = 1704 kJ/kg cli
2) -Clinker composition:
CaO = 65.2 %
MgO = 1.2 %
SiO2 = 22.9 %
Fe2O3 = 3.0 %
Al2O3 = 5.0 %
- Raw material = 1.1% hydrate water

R = 1.57 kg/kg 0.017 kg hydrate water/kg clinker
h= 0.652 3200 + 0.012 2710 - 0.229 2140 - 0.03 250
+ 0.050 120 + 0.017 5520 = 1721 kJ/kg cli

6.2.2.3 Sulfatization Effects
The SO2 can react with the alkalis K2O or Na2O (but also with CaO). For the heat effect the
following (simplified) reaction is considered as typical:
Na2O + SO2 + O2 Na2SO4 + heat
(molar heat effects of K2O and Na2O are of similar size!)
The overall degree of sulfatization reactions is usually not known. For the heat effect the
following lower and upper boundaries can be calculated:
lower boundary = - 8370 [(SO3)cli + (SO3)D - (SO3)R] (kJ/kg cli)
upper boundary = - 10800 [(Na2O)R - (Na2O)cli]
- 7120 [(K2O)R - (K2O)cli]
+ 8370 [(SO3)R - (SO3)cli] (kJ/kg cli)
where SO3, Na2O, K2O are expressed as kg/kg cli (referred to clinker!).
The indices denote: cli = in clinker
D = in dust
R = in raw meal
An average value can be introduced into the heat balance. Preferably this item is just listed
under the item heat of formation (under output). Its sign is then negative, i.e. the outputs are
reduced.
6.2.2.4 Lime Kilns

The formula for heat of formation of burnt lime is quite similar to that of cement clinker.
The following main reactions are taken into account:
CaCO3 CaO + CO2
MgCO3 MgO + CO2
CaO + 2 SiO2 C2S (dicalciumilicate)
Unlike on cement clinker the CaO in the lime product does not exclusively consist of non
carbonatic CaO, but some residual CaCO3 is also present. Therefore the non carbonatic
proportion CaOnc has to be calculated first:
CaOnc = CaOtot - CO2 (56/44)
The heat of formation becomes:
h = CaOnc 3150 + MgO 2710 - SiO2 . 2100 (kJ/kg lime)
CaOnc, MgO, SiO2 denote the weight fractions in lime product.
If the lime is present in lump form and the SiO2 does originate e.g. from coal ash it is unlikely
that C2S mineral is formed. In such cases only the SiO2 which comes from the limestone
feed has to be considered in the formula.

6.3 Radiation Heat Transfer
The basic radiation formula is
Qf = CR A [ (T/100)4 - (T0/100)4 ] [W]
Qf = heat flow [W]
CR = radiation constant = 5.67 W/m2K4
= emissivity (0...1)
A = radiating area (m2)
T = absolute temperature of radiating surface [K] (T = 273+t)
To = absolute temperature of ambient [K]
Strictly speaking above formula is only valid for cylinders of infinite length which radiate
towards an ambient which is far away from the cylinder. No obstacles may shield off the
radiation.
Though its simplifications this formula may be taken as a good approach for most cases.
By dividing Qf by T - To and A, the -value (heat transfer coefficient) can be obtained:
C T 4 T0 4
= R
2
(W / m C )
T T0 100 100
Or alternatively the following formula, which is identical to the latter comes out:
3
4 T 1 T 2
=CR 1 +
2
(W / m C )
100 100 4 T
T + T0
T= = average temperature (K)
2
T =T T0 = temperature difference (K or C)
From the second formula it becomes evident that (at small temperature differences)
increases with the third (not fourth!) power of the average temperature. In addition, the -
value does converge towards a finite value at zero temperature difference.
For practical evaluation of the (radiation) Fig. 9 (see paragraph 4.10.2) is available. It is
based on a constant ambient temperature of To = 293 K (20).
If the true ambient temperature is actually not exactly 20C one commits a slight error.
Usually this can be neglected. In some extreme or exceptional cases the above formulas
can be used for calculations.
Since the emissivity is often not known precisely and does also depend on the temperature
it is usually not worthwhile to do small corrections on the -values obtained from Fig. 9.

6.4 Convective Heat Transfer
6.4.1 Free Convection

Free convection occurs due to density differences between hot air (at the shell surface) and
ambient air. A natural updraft causes a convective air movement.
For vertical plates or horizontal cylinders the following relationship applies:
Nu = 0.13 (Pr . Gr) 1/3 (-)
where: Nu = D / = Nusselt number (-)
Pr = cp / = Prandtl number (-)
D 3 g 2 T
Gr = = Grashof number (-)
2 T0
(W/m2K) heat transfer coefficient
D (m) characteristical dimension *
(W/mK) heat conductivity
cp (J/kg K) specific heat
2
g (m/s ) gravity constant = 9.81 m/s2
(kg/m3) gas density
T (K) temperature difference (T - To)
(kg/ms) dynamic viscosity
To (K) absolute ambient temperature
* for a cylinder with diameter d D = (/2) d
Application range: Pr Gr > 109
As long as there is a minimum temperature difference of a few degrees and the diameter
range is over 1 m, above limitation does not affect the calculation.
A mathematical transformation of the basic equation reveals that the free convection heat
transfer does not at all depend on the characteristical dimension! It follows:
1 1
c 2 g 2 3 T 3
=0.13 p
T
(W / m 2K )
0
At T = 0 the free convection becomes zero (which is different from the behavior of the
radiation heat transfer!).
The numerical values for cp, , , must be taken from tables for air at the average
temperature between surface and ambient (use SI-units only).
Note that the density depends also on the barometric pressure and therefore the result will
depend on the altitude above sea level ( p2/3). As a general guideline the convective heat
transfer drops by about 8% per 1000 m of altitude.

By using numerical approximations for the properties of air the following relationship has
been developed (dimensional equation):
1.4 (0 T)1/3 (W/m2K)
o (kg/m3) = density at ambient temperature
3
(kg/m ) = density at average temperature
Though its simplicity the latter formula covers the temperature range from 0...500C with an
accuracy of better than 1%! This is more than enough for practical purposes.
At sea level and at 20C ambient it follows
1
3

T
1.6 (W / m 2K )
T
1+ 2 T
0
Above relationship is also an excellent numerical approximation of the curve for v = O in Fig.
10 (see paragraph 4.10.3), which is actually based on complete computer calculations out of
properties for air.

Table Properties of Air at Pressure = 1 bar
Temp. cp Pr
3 3 3 6 6 2 6 2
C kg/m kJ/kg K 10 /K 10 W/mK 10 kg/ms 10 m /s 10 m /s 1
-180 3.8515 1.071 11.701 9.0 6.44 1.67 2.18 0.77
-160 3.1258 1.036 9.320 10.9 7.85 2.51 3.37 0.75
-140 2.6391 1.021 7.758 12.7 9.20 3.48 4.71 0.74
-120 2.2867 1.014 6.659 14.6 10.49 4.587 6.30 0.73
-100 2.0186 1.011 5.846 16.4 11.72 5.806 8.04 0.72
-80 1.8073 1.009 5.219 18.16 12.89 7.132 9.96 0.72
-60 1.6364 1.007 4.719 19.83 14.02 8.567 12.0 0.71
-40 1.4952 1.006 4.304 21.45 15.09 10.09 14.3 0.71
-20 1.3765 1.006 3.962 23.01 16.15 11.73 16.6 0.71
0 1.2754 1.006 3.671 24.54 17.10 13.41 19.1 0.70
20 1.1881 1.007 3.419 26.03 17.98 15.13 21.8 0.70
40 1.1120 1.008 3.200 27.49 18.81 16.92 24.5 0.69
60 1.0452 1.009 3.007 28.94 19.73 18.88 27.4 0.69
80 0.9859 1.010 2.836 30.38 20.73 21.02 30.5 0.69
100 0.9329 1.012 2.684 31.81 21.60 23.15 33.7 0.69
120 0.8854 1.014 2.547 33.23 22.43 25.33 37.0 0.68
140 0.8425 1.017 2.423 34.66 23.19 27.53 40.5 0.68
160 0.8036 1.020 2.311 36.07 24.01 29.88 44.0 0.68
180 0.7681 1.023 2.209 37.49 24.91 32.43 47.7 0.68
200 0.7356 1.026 2.115 38.91 25.70 34.94 51.6 0.68
250 0.6653 1.035 1.912 42.43 27.40 41.18 61.6 0.67
300 0.6072 1.046 1.745 45.91 29.20 48.09 72.3 0.67
350 0.5585 1.057 1.605 49.31 30.90 55.33 83.5 0.66
400 0.5170 1.069 1.485 52.57 32.55 62.95 95.1 0.66
450 0.4813 1.081 1.383 55.64 34.00 70.64 107 0.66
500 0.4502 1.093 1.293 58.48 35.50 78.86 119 0.66
600 0.3986 1.116 1.145 63.5 38.30 96.08 143 0.67
700 0.3577 1.137 1.027 67.8 40.87 114.3 166 0.69
800 0.3243 1.155 0.932 71.3 43.32 133.6 190 0.70
900 0.2967 1.171 0.852 74.3 45.65 153.9 214 0.72
1000 0.2734 1.185 0.786 76.8 47.88 175.1 237 0.7

(actual cp, not average)
6.4.2 Forced Convection

Forced convection occurs at comparatively high wind velocity and dominates the convective
heat transfer, i.e. the free convection is suppressed.
The calculation of forced convection is depending on many factors, such as:
Wind velocity
Direction of the wind
Velocity distribution and flow obstacles
Uniformity of wind
Reynolds number (depends on kiln diameter).
There are a few more influence factors than in case of the free convection. Generally
speaking the calculation of forced convection heat transfer contains more possible sources
of error than free convection.

As a simplification, we will assume a cylinder in a non disturbed flow of a constant velocity v
(at 90 against kiln axis).
Fig. 14
For air the following formula apply:

Nu = 0.0239 Re 0.805 for Re = 40000...400000
Nu = 0.00672 Re 0.905 for Re > 400000
vD v D
Re= = =Re ynoldsNumber

D
Nu= =Nusselt Number

The properties , , have to be taken at average air temperature.
There are other formulas in use which can give different results, the above formulas are
preferred due to their simplicity. In any case there is always a incertitude from the mode of
calculation itself.

Two main factors determine the -value:
Velocity v
Average temperature (between surface and ambient)
In addition the does also depend on the diameter D. If the equation for high Reynolds
number (Re > 400000) is solved for , the following relationship is obtained:
D-0.095

proportional
This means that the value does not much depend on D! Therefore it is possible to work
with constant values within a certain diameter range.
This actually the base of Fig. 10 (see paragraph 4.10.3) which is calculated for a common
range from 3...4...6 m diameter.
6.4.3 Free Convection Plus Forced Convection

If the convection is clearly dominated either by free or forced convection the final
determination of the representative does not give any problem, since the higher value has
to be taken.
If the two values are of the same order they must be combined with an appropriate method.
It would be certainly wrong to add the two values. A better approach is the square addition:
tot = 2free + 2forced

It is also valid if either the free convection or the forced convection dominates.
Fig. 10 is based on the above method.
6.5 Effect of Thermal Improvements

According to the actual condition of an existing kiln system (heat consumption, heat balance,
other operating data) we can envisage an optimization campaign. From the thermal point of
view we can take certain measures in order to reduce the specific heat consumption. For
example:
Better insulation of rotary kiln or preheater/precalciner in order to reduce radiation losses
(except the sintering zone).
Improvement of the cooler efficiency (optimization of grate cooler operation or installation
of highly efficient internal equipment in planetary or rotary cooler).
Reducing of false air inleaks at kiln seals or at preheater.
Reducing of internal dust circulations in cooler, kiln or preheater (improves the internal
counter-current heat transfer).
Modification of raw mix in order to decrease the sintering temperature which in turn will
also reduce heat losses. A further effect of such a measure can be the reduction of
internal dust circulations due to better clinker granulometry (see above).
Modification of raw mix in order to decrease the heat required for decarbonation, e.g. by
making use of non-carbonatic CaO raw material sources. This possibility, however, is
very rare and often not feasible. (Note the basic difference to the measures which tend to
decrease the sintering temperature or increase the proportion of melting phase!)

The above measures are just a few typical examples. When one goes into such items, an
important phenomenon will appear soon: The so called loss multiplication factor for thermal
losses (or savings). What does this mean?
If a saving (or loss) in the high temperature zone in a kiln of say 100 kJ occurs the possible
gain in fuel heat consumption will not be 100 kJ but rather 130 to 150 kJ. That means that
the primary saving (in terms of heat balance item) will be multiplied by a factor of up to 1.5.
At the first glance the above principle seems to be contradicious because it would violate the
principle of heat balance or the energy law. However, what really happens is a differential
change of more than only one heat balance item. To illustrate this fact we start from an
example where we have reduced the shell radiation losses by 100 kJ/kg cli.
The following differential balance situation occurs:
Fig. 15
Exhaust gas heat = hoex - 40 kJ/kg

(The values ho
denote the situation
before modification)
Fuel heat=
hof - 140 kJ/kg
boundary
Radiation heat = hor - 100 kJ/kg
The corresponding multiplication factor for the above case is:

multiplication factor = - 140 kJ/kg = 1.40
- 100 kJ/kg
The above fact does result from the thermal behavior of the system and can be verified by
simulation models (not by a simple balance only).
A factor in the 1.4-range is quite typical for the situation in the high temperature zone (above
800C) of a cement kiln. The main effect of a change in this zone will be a corresponding
change at the exhaust gas, but also other minor effects will occur (e.g. at cooler losses).

A differential balance of heat can still be applied according to:
Input Change Output Change

fuel = - 140 kJ/kg radiation = - 100 kJ/kg
exhaust = - 40 kJ/kg
total = - 140 kJ/kg total = - 140 kJ/kg
The principle of loss multiplication does not only apply for radiation in the high temperature
zone but also for the heat which is recuperated in the clinker cooler.
Regarding the false air inleaks the corresponding deterioration of heat consumption is often
under-estimated. If false air inleaks into the high temperature zone it does not only cause a
heat loss because this air must be heated up to the exhaust gas temperature of the kiln
system! It actually causes much more losses than what would result from such a simplified
calculation approach. As a rule of thumb we may consider the following two main effects in
order to come to a realistic result:
Heating up of false air at the temperature of the high temperature zone which can be set
approximately to 800C (end of HT-zone).
Multiplication of the above heat requirement by a loss multiplication factor.
The above calculation is a rough approach. By more sophisticated simulation models we find
e.g. that a false air inleak into the high temperature zone of 0.08 Nm3/kg cli can cause an
additional heat consumption of 100 kJ/kg cli. In contrast, the inleak in the low temperature
zone (e.g. air lift on SP-kiln) is much less critical.
6.6 Heat Transfer in Preheaters and Coolers and Improvement Potential

Normally we are considering a cement kiln as an equipment for burning cement clinker. As
an essential feature we must be able to generate a high sintering temperature in the order of
1450C (material temperature).
But a kiln system is of course much more than a generator of sintering temperature. It is also
a system of heat exchangers which allows for burning at low heat consumption. Generally
speaking we will find two essential heat-exchanging systems on any cement kiln:
a) raw meal preheater
b) clinker cooler
Low heat consumption is only possible if the above two heat exchangers perform optimally.
How can we get optimum heat exchange performance? Form the basic theory it is known
that even in case of an exchanger which is perfectly insulated against ambient temperature
influence, three important conditions are required for optimum heat exchange:
1) Optimum heat transfer rate (here: from gas to solid) high specific contact or surface
area, high (W/m2C).
2) The two heat exchanging streams must flow in counter-current manner, or at least in an
arrangement which has similar characteristics to a counter-current system (e.g. overall
arrangement of a 4-stage cyclone preheater).

3) If we want to recover heat from a flow 1 completely into a flow 2 the flow 2 must
have at least the same heat equivalence as flow 1:
(flow 2) x (cp2) (flow 1) x (cp1) [kW/C] or [kJ/kg cli C]
flow: [kg/s] or [kg/kg cli]
cp: [kJ/kg C]
In the above equation the cp values are considered as constant (approximation).
Graphically this means that the heat characteristic curves of the two heat flows must be in a
certain relationship as shown here:
Temperature [C ]
hot end
1 2
w w
Flo Flo
maximum
allowable
inclination range
2 for Flow 2
Flow
cold end
Heat Flow [kW]
What are the practical consequences for a cement kiln? We consider:

A) Raw meal preheater
B) Clinker cooler
A) Raw meal preheater
1) Heat transfer rate:
Optimum heat transfer rate and optimum specific surface (small particles) and
distribution has been almost achieved in the cyclone suspension preheater. After
every cyclone stage material and gas come to almost complete temperature
approach and there is usually little to improve on that.
Improvement are rather to be done where the heat exchange takes place in the
rotary part itself. Especially on wet and long dry kilns the kiln internal fittings are
essential for good heat exchange.
2) Counter-current principle:

Counter-current flow in its proper sense does not exist in case of a cyclone
suspension preheater. Instead, we have usually four co-current heat exchanging
units, but the overall arrangement acts as counter-current system. To reach an
ideal state one would have to apply an infinite number of cyclone stages.
Practically the common arrangement of 4 or 5 stages can be considered as
sufficient.
True counter-current preheaters are shaft preheaters or preheating in long rotary
kilns. Such preheaters would theoretically be ideal. Practically they are less
efficient because of distribution problems and backmixing effect (internal material
circulations) and comparatively high losses to outside (in case of a rotary kiln).
3) Equivalence of heat flow characteristics:
A general feature of any raw meal preheater is the surplus of heat input by the
hot gases. After the calcining step the hot gas has a temperature of approx.
850C and the specific quantity is always above 1.3 Nm3/kg cli even in case of
an optimum kiln system (4-stage SP kiln). This amount of heat is too much, it
cannot be used completely for preheating of raw meal ( 1.56 kg RM/kg cli) up to
calcining temperature ( 800C). Therefore a certain amount of waste heat will
occur even in an ideal case. Theoretically we should not exceed 1 Nm3/kg cli
for ideal recuperation. Practically this cannot be realized, not even on modern kiln
systems which produce comparatively little exhaust gas.
Graphically we have the following heat situation in a raw meal preheater:
Fig. 16
Temperature [C ]
~850 C
t Ga s
aus ~800 C
Exh
al
Me
w
Ra Preheating Calcining
50 C
Heat transferred (or heat content) [kJ/kg cli]
Above diagram is simplified but typical for any preheater.

Because of the heat surplus of the exhaust gas it is not possible to achieve an ideal
recuperation even at perfect counter-current heat transfer (e.g. infinite number of cyclone
stages). The exhaust gas will always give a certain residual heat content.

Practically this means that all kind of improvements on the preheater have a limited
potential. In the example of a 4-stage preheater kiln we can add a fifth stage which causes a
reduction of about 100 kJ/kg cli in fuel heat consumption. More than 5 stages will bring only
marginal economical point of view. *) Even at 5 stages we may check if the necessary
investment and the (possible) increase of pressure drop can be justified by the local cost
structure.
*)
Instead of constructing more than 5 conventional cyclone stages one would prefer
here non-conventional preheaters, such as the cross-suspension-preheater (two strings
with cross flow of raw meals).

B) Clinker Coolers
1) Heat transfer rate:
On a planetary or a rotary cooler we have the possibility to increase the heat
transfer rate by installation of efficient internal equipment (tumblers, lifters) which
increase the active heat transfer area by better moving and scattering of the
clinker. If a cooler has worn out internal equipment or equipment of inadequate
design we may realize a considerable potential for improvements. Improvements
may also result from a more uniform clinker granulometry (less internal dust
circulations).
On a grate cooler we find quite a different situation. The real problem is not the
heat transfer rate between a piece of clinker and the cooling air but rather the
uniform air distribution through the clinker bed. Also here we may realize a
considerable improvement (thick bed operation, mechanical modifications at inlet
zone etc.).
2) Counter-current principle:
There is an obvious difference between planetary/rotary coolers and the grate
cooler:
- planetary / rotary counter-current flow
- grate cooler almost cross-current flow
The grate cooler has a nearly cross-current performance and has therefore, from
its principle, a limited heat recovery potential. Improvements are possible when
air recirculation (of hot waste air) to the first grate section is applied in order to
increase the heat content of the secondary air. Another quite different measure is
the increasing of the bed thickness in order to come more towards a counter-
current-like exchange (similar to a shaft cooler). The disadvantage is an increase
of the cooling air pressure.
3) Equivalence of heat characteristic
On a clinker cooler we would theoretically require approx. 0.77 Nm3/kg cli of
cooling air in order to transfer the clinker heat completely to the secondary air
(under perfect counter-current heat exchange).
The practical figures are usually above 0.8 Nm3/kg cli. Compared to a raw meal
preheater we have an inverse situation: The heat from the hot clinker could
theoretically be completely recovered (under ideal conditions)! Of course, we
know that the common, practical figures are often below 70%. The latter fact
illustrates that from the basic principle there is still a considerable heat potential
which is not used for reducing the kiln heat consumption.
In this context it is also important to note that the practical efficiency of any type
of clinker cooler increases with higher cooling air quantity. As a consequence we
should draw as much cooling air as possible through the cooler and therefore
avoid or reduce excessive primary air quantities or false air inleaks at the kiln
hood or kiln seals. The improvement does not come from the heat transfer proper
but rather from the improved heat characteristic (air/clinker ratio).

7. TEST QUESTIONS
1) Give an example where it can be worthwhile to execute a complete heat balance on a
kiln.
2) Summarize all important measurement points which are needed for doing a complete
heat balance on a suspension preheater kiln.
3) What is the usual value (or range) for the specific heat (kJ/Nm3 C) of exit gas of an SP-
kiln at 350C?
4) What is the sensible heat content of 1 kg clinker at 1450C, expressed as kJ/kg cli?
5) What is the standard value (or range) for heat of formation (kJ/kg cli) for clinker burning?
Which heat effects are included in above value?
6) Determine the heat transfer coefficient (W/m2C) for the total heat transfer by radiation
plus free convection. The temperature of the kiln shell section is 200C (ambient = 20C,
= 0.9).
7) When has the forced convection heat transfer (instead of free convection) to be
considered?
How is the above value affected in case of smaller dimensions (say 1 m instead of 5 m
diameter)?
8) If the shell losses in the calcining zone can be reduced by 50 kJ/kg cli, what will be the
approximate saving of fuel heat (kJ/kg cli)?
9) What is the approximate fuel heat which can be saved through the installation of a fifth
cyclone stage on a suspension preheater kiln?
What would be the approximate amount of false air reduction (Nm3/kg cli into the high
temperature zone) in order to achieve a similar fuel heat saving?

8. LITERATURE
1) Zur Strassen, H.
Der theoretische Wrmebedarf des Zementbrandes
ZKG 10 (1957), Vol. 1, p. 1-12
1) Jakob, M.
Heat transfer, Vol. I (1949), p. 529
2) Hilpert, R.
Wrmeabgabe von geheizten Drhten und Rohren im Luftstrom
Forsch.-Ing.-Wes., Vol. 4 (1939), p. 215-224
3) Gygi, H.
Thermodynamics of the cement kiln, third industrial symposium on the chemistry of
cement
4) Eigen, H.
Beitrag zur Thermodynamik des Drehofens
Tonindustrie-Zeitung 82 (1958), No. 16, p. 337-341
5) Frankenberger, R.
Beitrag zur Berechnung des Wrmebergangs in Zementdrehfen
Dissertation, Technische Universitt Clausthal (1969)
6) Khle, W.
Untersuchung ber die ussere Wrmeabgabe von Drehfen durch Strahlung und
Konvektion
Zement-Kalk-Gips, Vol. 6, 1970, p. 263
7) VDZ Unterlagen
Wrmetechnische Berechnungen
Verein Deutscher Zementwerke E.V., Dsseldorf
8) VDI-Wrematlas
Berechnungsbltter fr den Wrembergang
VDI-Verlag GmbH, Dsseldorf
9) Barin, I. and Knacke, O.
Thermochemical properties of inorganic substances
Springer Verlag, Berlin, Heidelberg, New York
10) Elkajer, P. (FLS)
Die Bestimmung des Wrmeverbrauches mit vierstufigem Zyklonvorwrmer durch
Aufstellung eines mathematischen Modelles
Zement-Kalk-Gips, Vol. 2, 1980
11) Gardeik, H.O.
Berechnung des Wandwrmeverlustes von Drehfen und Mhlen
Zement-Kalk-Gips, Vol. 2, 1980
12) Rother, W.
Ausfhrung von Rohmehl-Wrmetauschern unter Bercksichtigung heutiger
Kostenfaktoren
Zement-Kalk-Gips, Vol. 2, 1982, p. 66 ff.

9. SYMBOLS AND UNITS
A m2 area
CR W/m2K4 radiation constant
cP kJ/kg C specific heat (at const. pressure),
or kJ/Nm3 C specific heat capacity
CV kJ/kg calorific value
D m diameter
g m/s2 gravity constant
h kJ/kg heat content (specific)
or kJ/Nm3
or kJ/kg cli
L m length
m kg mass
or kg/kg specific mass
mf kg/h mass flow
Qf kW heat flow (1 kW = 1 kJ/s)
t C temperature (Centigrade)
T K temperature (Kelvin)
v m/s velocity
w kg/kg water content
Greek Letters
W/m2K heat transfer coefficient
- emissivity (for radiation)
W/m C heat conductivity
kg/m3 density
Dimensionless Numbers
Nu Nusselt number (for heat transfer)
Pr Prandtl number
R kiln feed (raw meal) / clinker-ratio
Re Reynolds number
Indices
conv convection
rad radiation
tot total
o ambient condition or zero condition
Conversion Factors
Length 1 inch 0.0254 m
1 ft 0.3048 m
Area 1 sq. ft 0.092903 m2
Volume, Volume Flow 1 cu.ft 0.028316 m3
1 cu.ft/min 1.699 m3/h (actual m3)
Mass 1 lb. 0.45359 kg
1 short ton 907.185 kg
(USA)

Pressure 1 bar 105 N/m2
1 atm. 1.013 bar
Energy 1 kJ 1000 J
1 MJ 1000 kJ
1 kWh 3600 kJ
1 kcal 4.187 kJ
1 BTU 1.055 kJ
Temperature Conversion C= 5/9(F - 32)
K= 273.15 + C
Heat Flow 1 kW 1000 W = 1 kJ/s
1 kcal/h 1.163 W
1 BTU/h 0.29307 W
Specific Heat 1 kcal/kg C 4.187 kJ/kg C = 4187 J/kg C
1 BTU/lb F 1 kcal/kg C = 4.187 kJ/kg C
Heat Transfer Coeffic. 1 kcal/m2h C 1.163 W/m2 C
1 BTU/ft2h F 5.678 W/m2C
Standard Conditions for Standard 0C and 1 atm. (1.013 bar)
Gases Conditions
2.73.15 p(bar )
Nm 3 =act .m 3
273.16 + t (c ) 1.013bar

Chapter 11
Main Fans

Main Fans
Authors: W. Zeller, Th. Richner, D. Brassel

PT 99/14500/E
1. Design and efficiency of fan impellers ........................................................................ 674

1.1 Fan impeller types .................................................................................................... 674
1.2 Selection criteria....................................................................................................... 675
1.3 Fan applications in the cement industry ................................................................... 677
2. Fan performance curves............................................................................................... 678
2.1 System Resistance Curve ........................................................................................ 678
2.2 Fan curves................................................................................................................ 680
2.2.1 Fan equations ................................................................................................... 681
2.2.2 Adjusting fan performance curves..................................................................... 682
3. Flow control ................................................................................................................... 685
3.1 Damper control......................................................................................................... 687
3.2 Radial inlet vane....................................................................................................... 691
3.3 Speed control ........................................................................................................... 693
3.3.1 Hydraulic transmission with fixed speed motor ................................................. 695
3.3.2 Speed-controlled electric motors....................................................................... 695
4. Possible problems with fans ........................................................................................ 696
4.1 Vibrations ................................................................................................................. 696
4.1.1 Variable speed operation .................................................................................. 699
4.1.2 Thermal effects ................................................................................................. 699
4.1.3 Hot shutdowns .................................................................................................. 699
4.2 Material build-up....................................................................................................... 700
4.2.1 Kiln exhaust fan build-up................................................................................... 702
4.2.2 Recommendations against build-up.................................................................. 702
4.3 Erosion ..................................................................................................................... 705
4.3.1 Erosion types .................................................................................................... 705
4.3.2 Improper duct connection.................................................................................. 705
4.3.3 Effects of impeller speed and wheel inlet velocity............................................. 706

4.4 Wear protection ........................................................................................................ 706
4.4.1 Protection of parts subjected to abrasion.......................................................... 706
4.4.2 Deflection of abrasive particles ......................................................................... 707
4.4.3 Liner materials................................................................................................... 709
4.5 Bearings ................................................................................................................... 710
5. Fan Capacity Adjustment ............................................................................................. 710
5.1 Fan capacity too low................................................................................................. 710
5.2 Fan capacity too high ............................................................................................... 711
6. Troubleshooting ............................................................................................................ 712
7. Start-up of fans .............................................................................................................. 715
8. Fan impeller arrangement and connections ............................................................... 716
8.1 Assembly.................................................................................................................. 716
8.1.1 Overhung assembly (Fig. 8.1)........................................................................... 716
8.1.2 Center hung assembly (Fig. 8.2)....................................................................... 717
8.2 Foundations.............................................................................................................. 718
8.3 Connections ............................................................................................................. 720
8.3.1 Inlet connections ............................................................................................... 720
8.3.2 Outlet connections ............................................................................................ 722
9. Information Sources...................................................................................................... 724

1. DESIGN AND EFFICIENCY OF FAN IMPELLERS
Fans are essential components of the cement manufacturing process and merit therefore
particular considerations with respect to
design and efficiency of the impeller
fan size and operating point
flow control
reliability
Main fans in a Cement Plant are found as
Kiln ID Fan
Raw Mill Fan
Cooler Exhaust Gas Fan
Kiln Dedusting Fan
Altogether these fans consume between 30 and 50% of the plants total electrical energy.
1.1 Fan impeller types

There are four basic blade forms used in industrial induced draft service:
backward airfoil blades
backward curved blades
backward inclined blades
radial straight blades
Fig. 1.1 shows design and efficiency of these impellers.

Fig. 1.1: Design and efficiency of impellers
TYPE EFFICIENCY APPLICATION

backward for clean gas

airfoil blades > 84 % applications
(dust < 50 g/m3)
for gas with a dust

backward concentration
curved blades up to 82 % < 100 g/m3
Backward for gas with a

inclined up to 80 % dustload
blades up to 100 g/m3
radial straight for gas with a high

blades 60 - 75 % dustload
(dust > 100 g/m3)
1.2 Selection criteria

It is of great importance that critical process equipment, such as fans, is selected on the
basis of proven ability in order to provide maximum reliability rather than on an efficiency
rating.
In situations where more than one blade form will meet a performance requirement, it then
becomes necessary to select the one form that will be most overall cost-effective.
For the selection process the supplier should provide the operating and service manual for
the equipment type being considered.
The operating and service report should include all occurrences that require fan
maintenance.

To assist in fan type selection, there are at least four important points to be considered.
Fan efficiency:
Because many of the higher efficiency fans can only achieve their stated efficiency within
a narrow operating range, a true energy evaluation must consider the actual operating
point and alternate operating points on a time basis.
Many systems include a built-in safety factor, which results in reduced efficiency when
operated at constant speed with damper regulation.
Continuous operation:
Any type of equipment will require maintenance. In blade form selection, blade build-up
(cp. section 4.2) and erosion (cp. section 4.3) have the greatest effect on fan operation.
Build-up on the wheel results in reduced performance. There is an increased tendency
for material to build up on blades as the blade angle is tilted back from radial. This build-
up can accumulate to the degree that it restricts and alters air passages, reducing both
efficiency and performance.
Mechanical design:
All fan rotors are subject to centrifugal force. Depending on blade form and angle,
different types of stress occur in a blade. The radial Blade is in tension, while bending
and tensile stresses act on the backward inclined/Airfoil design. Bending stresses are
more subject to fatigue stresses.
Equipment costs

1.3 Fan applications in the cement industry
The following table shows an overview of fan applications in the cement industry.
Dust load Blades Max. max. Flow Rotor Stator Efficiency
Location [g/Nm3] mainly Temp speed regulation protection protection
used [C] [rpm] [%]
Coal Mill
Filter exhaust < 0.15 F/C/A 150 1800 VC/ILD (WP) - 70 - 85
Separator / cyclone < 100 R/F/C 150 1200 ILD HSWP WP( HSWP) 55 - 75
exhaust
Raw Mill
Filter exhaust < 0.15 F/C/A* 300 1800 VC/ILD/VS - - 70 - 85
Separator / cyclone < 100 F/C 300 1200 ILD/VS (HW/HWSP) WP 65 - 75
exhaust
Cement Mill
Filter exhaust < 0.15 F/C/A 200 1800 VC/ILD - - 75 - 85
Separator / cyclone < 100 F/C 200 1200 ILD (WP/HSWP) WP 65 - 75
exhaust
Dry Process
Preheater exhaust < 100 R/F/C 450 1200 ILD/VS (WP) - 55 - 75
Kiln line filter < 0.15 F/C/A* 350 1200 VC/ILD/VS - - 70 - 85
exhaust
Semi-dry process
Recirculation fan < 20 R/F 450 750 ILD/VS HSWP WP 60 - 70
Kiln line filter <0.15 F/C/A* 450 1800 VC/ILD/VS - - 70 - 85
exhaust
Cooler
Filter exhaust <0.15 F/C 450 1200 VC/ILD/VS (WP) - 70 - 88
Filter exhaust 0.5 F/C 450 1200 ILD/VS WP (WP) 70 - 80
Cyclone exhaust <5 R/F 450 1200 ILD/VS HSWP WP 65 - 75
Recirculating < 15 R 450 750 ILD HSWP WP(HSWP) 60 - 70
Cooler F.D. fans Traces F/C/A 50 2200 VC/VS - - 75 - 85
* in case of bag filter
Abbreviations:

a. Blade Type
R: radial blade
F: backward inclined flat plate
C: backward curved plate
A: backward airfoil blade
b. Flow regulation (cp. section 3)
VC: Vane control set at the fan inlet
ILD: Inlet louver damper set on the inlet box
VS: Variable speed
c. Rotor and stator protection against abrasion (cp. section 0)
WP: bolted or welded wear plates
- mild steel
- quenched steel
- wear-resistant steel
HSWP: bolted mild steel wear plates with hard surfacing
- chromium carbide
- tungsten carbide
(WP/HSWP): optional
2. FAN PERFORMANCE CURVES
2.1 System Resistance Curve

The System Resistance Curve is the relation between pressure and volume flow for the
given system. For a system in which only air is moved and the geometry of the ductwork is
constant, the pressure drop is proportional to the square of the flow rate.
In most of the applications around the cement plant, this curve is constantly changing
because of the changing system conditions. For instance, the cooler under-grate fan system
resistance curve depends not only on the geometry of the ductwork and cooler, but also on
size distribution and thickness of the clinker bed on the grate above the compartment.
The system resistance curve of a roller mill system is dependent on the geometry of the
ducts, the amount of material being transported by the gas, the composition of the gases
and the speed of the classifier. Because of these variables, it is essential to understand that
a system resistance curve can vary significantly from the slope of the curve illustrated in e.g.
Fig. 2.3.
Principally the curve is derived as described next. Fig 2.1 schematically shows a system.

Fig. 2.1: System with fan
Total pressure (dynamic and static part) at S:

psuc = ps ps with ps = pressure loss due to friction of the suction
side
( )
= f V& 2
Total pressure (dynamic and static part) at P:

ppress = pp pp with pp = pressure loss due to friction of the pressure
side
( )
= f V& 2
Pressure difference pfan, which has to be produced by the fan, consists of a static and
dynamic part and is a function of the gas velocity or gas flow.
( )
p = ppress - psuc = = pstat + pdyn = f(w2) = f V& 2

2.2 Fan curves
The fan performance curve is derived from laboratory data when the flow conditions to and
from the fan are ideal. Since these conditions seldom exist in cement plant fan locations.
The fan curve data cannot be considered reliable when applied to field conditions. Because
of this uncertainty of the prevailing conditions, we find that engineers specify oversized fans
during the plant design stage.
For example, typical factors that can effect the performance of an under-grate cooler fan
include a dirty inlet screen; a structural column too close to the inlet; a silencer at fan inlet or
the influence of an adjacent fan. Likewise, similar factors influencing an induced-draft fan
include asymmetrical inlet duct which effects streamline distribution to the fan or heavier
than normal blade wear pads that restrict blade passage geometry.
A typical fan curve (see Fig. 2.2) shows the quantity of air on the horizontal axis and the fan
static pressure and fan power plotted on the vertical axis. The conditions of density and flow
are at the inlet of the fan.
The actual operating point of the fan will be the intersection of the fan curve with the
system resistance curve (cp. section 2.1).
The efficiency of a fan is best close to its design point of operation. The farther off these
ideal conditions, the lower the efficiency. The operating points of equal fan efficiency are
located on oval shaped curves around the maximum point.
Fig. 2.2: Fan characteristics and efficiencies by different fan speed
The fan curve depends on the impeller speed and the physical properties of the gas (cp.
section 2.2.2)

The pressure which is produced by the fan equals the pressure difference between the total
pressure at the fan outlet and the fan inlet:
p = ppress - psuc = = pstat + pdyn
For most fans in cement plants, it is satisfactory to assume that the fan's pressure is the
difference between the static pressure at the inlet and outlet of the fans.
The fan power is the power at the fan shaft. For most plant applications where the fan has
constant speed and is direct coupled, this can be assumed to be 96 % of the power
consumed by the drive motor when the motor is fully loaded. Since power factor and motor
efficiency vary with motor load, it is difficult to measure fan shaft power input without a
kilowatt meter. However, near full motor load conditions, the amp reading is a good indicator
of shaft power. For fans with variable speed drives, especially hydraulic or magnetic clutch
drives, the efficiency of the drive can be very poor so the power at the fan shaft cannot be
assumed to be a linear function of the power consumed by the motor.
2.2.1 Fan equations

Power requirement:
V& p 10 3
N= [kW]

N : Required power [kW]
V& : Gas flow [m3/s] (not [Nm3/s])
p : Total increase of pressure in fan [Pa]
: Fan efficiency [-]
Depending on the blade shape of the impeller, the power requirement to draw the gas
through a given system (and therefore for determined pressure losses) can vary in a
relatively wide range.
As the required fan power is given by the equation above, it is obvious that considerable
savings can be achieved with the most efficient impeller.
Total pressure increase:

Formula for a rough calculation of the total pressure p = ppress - psuc = pstat + pdyn:
0 .6 n 2 d 2 2
p [Pa]
3600
: Density [kg/m3]
n : Fan speed [rpm]
d : Impeller diameter [m]
Remark: p is just a rough guideline and depends very much on the blades shape and the
rotor efficiency.

Example: Effect of fan efficiency
Plant: 4-stage SP kiln, 3000 t/d clinker
Requirement for kiln ID-fan: V& = 115 m3/s at 350C

p = 6000 Pa
two efficiencies available: 1 = 0.75
2 = 0.85
V& p 10 3 115 6000 10 3
N1 = = = 920 kW
1 0.75
V& p 10 3 115 6000 10 3

N2 = = = 812 kW
2 0.85
Power saving = 108 kW = 0.86 kWh/t cli

Assuming an operation time of 7500 h/year and an energy price of
0.05 US$/kWh the yearly saving will amount to 40500 US$!
2.2.2 Adjusting fan performance curves

It was mentioned above that the fan performance curve is given at specific conditions of
density and fan speed. Most often, it is necessary to correct the fan curve for density and
speed other than the predicted conditions. Fig. 2.3 shows the influence of density changes
on the fan performance.
Indices: 1: reference conditions (or original fan curve characteristic)

2: actual conditions
Gas density correction:

A correction of the density is necessary if:
temperature changes
the chemical composition of gas changes
the altitude changes (height above sea level).
pstat 1 T
= 1 = 2
pstat 2 2 T1
Volume flow and efficiency of the fan are unaffected by gas density changes.

Remark: Calculation of actual density
M p 273
=
22.4 1013 T + 273
density [kg/m3]
M molecular weight of gas [kg/kmol]
p actual pressure [mbar]
T actual temperature [C]
Ambient pressure, depending on the altitudes can be calculated by:

p = 1013 exp[ 0.001255 h]
h altitude above sea level [m]
Density Molecular weight

Gas
[kg/Nm3] [kg/kmol]
O2 1.429 32
CO2 1.964 44
N2 1.250 28
Air 1.292 29
H2O 0.804 18
Fan speed correction:
n
Volume flow: V&2 = V&1 2
n1
2
n
Static pressure: pstat 2 = pstat 1 2
n1
3
n
Power requirement: N2 = N1 2
n1
(for 1 = 2)

Fan wheel dimensions correction:
3
d b
Volume flow: V&2 = V&1 2 2
d1 b1
2
d
d1
5
d b
Power requirement: N2 = N1 2 2
d1 b1
whereas d impeller diameter [m]
b impeller width [m]

Fig. 2.3: Influence of density changes on system resistance and fan performance
curves
3. FLOW CONTROL
Whether a fan ever will operate at a high efficiency is to a large extent already decided when
the fan size is selected.

Since the efficiency of the fan has a maximum at a specific operating point the fan ought to
be sized so that it can operate at such conditions most of the time. However, since the fan
performance needs to be flexible to meet variable requirements, the fan size will, to some
extent at least, be designed for the maximum requirement. Reduced requirements have to
be met by fan control dampers or variable speed control.
Besides the efficiency of the fan itself, the type of flow control has the strongest influence on
the overall power consumption.
Damper control generally results in higher power consumption and so more if the damper is
installed at the fan outlet. Therefore the damper should always be installed at fan inlet. Inlet
vane control may be satisfactory from 100 % down to about 70 % of maximum flow, but the
power demand becomes high when the flow is reduced further. Speed control is virtually
ideal.
Fig 3.1 shows different arrangements for inlet and outlet dampers.
Fig. 3.1: Design possibilities for fan dampers
Inlet box damper

parallel blades Diffuser
Inlet box damper

oposed blades
(not recommended)
Outlet damper
horizontal blades
(not recommended)
Radial inlet
vanes
Outlet damper
horizontal parallel blades
(not recommended)
Outlet damper
vertical opposed
blades
Outlet damper
vertical parallel
blades
Figure 3.2 shows how fan power consumption is affected by the different flow regulation
methods.

Fig. 3.2: Comparison: Types of Flow Regulation
3.1 Damper control

A parallel blade inlet damper is preferred over either outlet damper or an opposed-blade inlet
damper. The parallel-blade inlet damper pre-spins the incoming air in the direction of wheel
rotation, resulting in lower energy consumption in the regulation range of 100 - 80% of
maximum flow. Fig. 3.3 shows the parallel inlet damper arrangement.

Fig. 3.3: Parallel inlet box damper

The inlet box damper influences the fan curve (see Fig. 3.4).
Fig. 3.4: Flow regulation by inlet box damper

Fig. 3.5 shows the operating point at the intersection of the system resistance curve and the
fan curve. By reducing the airflow, the fan theoretically produces a pressure above 80 mbar.
Since the system behaves like the system resistance curve, the damper induces a pressure
loss of about 40 mbar.
Fig. 3.5: Flow regulation by outlet damper

3.2 Radial inlet vane
The radial inlet vane mounted direct at the fan inlet pre-spins the incoming air still better in
the direction of wheel rotation, resulting in a wider range of stable regulation (100 - 70%) and
less energy consumption. Fig. 3.6 shows the parallel inlet damper arrangement.
Fig 3.6: Radial inlet vane
Their use is mainly recommended in connection with over hung arrangement fans with low
rates of dust, thus limited to applications after filters or in clean air, e.g. for cooler under-
grate fans.
This arrangement is normally more costly.

Fig. 3.7: Flow regulation by inlet vane damper

3.3 Speed control
Variable speed controlled fans have more fan characteristic curves but only one
characteristic curve of the duct system (see Fig. 3.8).
It is obvious that the variable speed drive is the most efficient type of fan control since no
additional resistance for e.g. damper is built into the system. On the other hand a variable
speed control is more capital intensive than a damper control.

Fig. 3.8: Flow regulation by speed control

3.3.1 Hydraulic transmission with fixed speed motor
Hydraulic transmission in connection with a fixed speed motor can be an option for speed
ranges from 100% down to 85% of maximum speed, but the energy efficiency becomes low
when the flow is further reduced.
3.3.2 Speed-controlled electric motors

Flow control by variation of the fan speed is most efficient with regard to energy savings and
permits also the reduction of wear on the fan wheel.
DC motors have limitations in high speed and power (roughly 1000 rpm for 1500 kW,
600 to 800 rpm for 2000 kW motor) and require a lot of maintenance work (motor
ventilation circuits, carbon brushes etc.), so their use is not recommended any longer.
AC motors with slip recovery. These motors have a limited speed range down from 100 -
30%. Its cost increases with the width of the range.
Synchronous motors with AC variable frequency control
Both types of AC motors are well suited for high power (500 to 5000 kW).
Fan equations at variable speed:
n
Volume flow: V&2 = V&1 2
n1
2
n
n1
3
n
Power requirement: N2 = N1 2
n1
(for 1 = 2)

4. POSSIBLE PROBLEMS WITH FANS
4.1 Vibrations
Main reasons for fan vibrations are build-ups of process material (refer to section 4.2) and
wear/erosion on the fan impeller (refer to section 4.3/0). A moderate level of vibration can be
tolerated from a mechanical design point of view, although it certainly reduces the bearing
lifetime to a certain extent. Therefore it should be tolerated only to avoid additional main
equipment downtime.
If vibrations are excessive, fan impeller balancing, cleaning or replacement is required. For
balancing special vibration detectors are used on site or the wheel has to be shop balanced.
General Machinery vibration severity as per ANSI S2.41 (Fig. 4.1 a) for use as a guide in
judging vibration as a warning of impending trouble)
1. Rigid support
The fundamental natural frequency of the machine/support system is higher
than the operating speed
excellent 0. to 2.54 mm/s vibration velocity (Peak)
good 2.55 to 6.35 mm/s
alarm 6.36 to 12.7 mm/s
shutdown > 12.7 mm/s
2. Flexible support
The fundamental natural frequency of the machine/support system is lower
than the operating speed
excellent 0. to 3.81 mm/s vibration velocity (Peak)
good 3.81 to 10.16 mm/s
alarm 10.17 to 19.1 mm/s
shutdown > 19.1 mm/s
Vibration severity criteria (10 Hz to 1 kHz) per ISO 2372 (Fig. 4.1 b)
1. Large machines with rigid foundations whose natural frequency exceeds
machine speed
good 0.0 to 2.54 mm/s vibration velocity (Peak)
allowable 2.55 to 6.35 mm/s
just tolerable 6.36 to 15.84 mm/s
not permissible > 15.84 mm/s
2. Large machines operating at speeds above foundation natural frequency
good 0 to 4 mm/s vibration velocity (Peak)
allowable 4 to 10 mm/s
just tolerable 10 to 25.4 mm/s
not permissible > 25.4 mm/s

If the tolerable vibration levels are exceeded, the fan must be shut down, which usually
results in costly production losses. To extend the periods between shutdowns due to
vibration, the use of automatic balancing devices, mounted on the fan shaft may be
considered. Depending on their size, they are capable of automatic compensation of a
certain rotor unbalance. For manufacturers of such devices, refer to the information source
at the end of this report.
Fig. 4.1 a: Vibration severity chart (ANSI S2.41)

Fig. 4.1 b: Vibration severity chart (ISO 2372)

4.1.1 Variable speed operation
All electrical variable speed drive systems can generate harmful harmonics that result in
torque pulsation. Such harmonics can be predicted and filtered, but often at high cost. For
fans with variable speed control it must be verified by the motor supplier that the pulsating
torque do not cause excessive vibrations.
4.1.2 Thermal effects

Some typical problem areas are:
Due to expansion joint problems, forces due to thermal expansion of ducts are
transmitted to the fan housing, resulting in damage of the housing or interference
between wheel and housing. It can also cause excessive force on foundation bolts that
sometimes can result in foundation cracks.
Rapid temperature changes in a system require proper design of the wheel-to-shaft fit.
This will assure that looseness and resulting vibration sensitivity will be avoided.
4.1.3 Hot shutdowns

Thermal shaft set is a usual concern on centre-hung fans above 120C when shut down in
the hot condition. It is generally agreed that this thermal bowing of the shaft occurs due to
uneven thermal gradients across the wheel and shaft assembly at hot shutdown. This results
in small asymmetrical distortions, often sufficient to cause excessive unbalance forces
during start-up.
In some cases the resulting vibrations are within acceptable limits. If so, it is usually found
that the unbalance forces disappear after 12 to 36 hours of operation.
The preferred, but expensive solution is to have an auxiliary drive to slowly rotate the wheel
and shaft assembly, which should be engaged immediately after hot shut-down to avoid the
undesirable thermal distortion. Auxiliary drives are typically designed to maintain a minimum
speed (40 to 60 rpm) as the fan slows down. They are not intended for use in starting the fan
rotor from a dead stop.
Most hot gas fans work well without auxiliary drive, since hot shutdowns are infrequent and
of short duration and the natural draft of the chimney keeps the rotor in slow motion for some
time.
The tendency is to install hot gas fans without auxiliary drives. Often it is also the case that
variable speed drives have a turndown ratio of 10:1, which corresponds normally to less than
70 rpm.

4.2 Material build-up
A well maintained fan is generally a reliable piece of equipment. However, in case of
material build-ups within the fan, serious operational limitations could occur, such as
vibrations.
Possible reasons for material build-ups are:

Mineralogical composition of the material
Burning conditions
Duct/fan arrangement
High temperatures
Inappropriate blade shape and angles
High fan speed
Dew point problems
After balancing, the rotor of a fan will still have a certain residual imbalance, the value of
which depends on the balance quality grade. The unbalance force can be calculated with the
formula:
F = m 2 eper [N]
m rotor mass [kg]

angular velocity 2n/60 [s-1]
n rotor speed [rpm]
eper permissible residual specific unbalance [m]
Industrial fans often handle dust-laden gases and thus dust deposition on the impeller may
occur, increasing the initial unbalance. Asymmetrical wear on the impeller has the same
effect. These additional unbalances can be considerable compared to the permissible
unbalance according to the balance quality grade. If the resulting vibration exceeds the
tolerable limits, the fan has to be shut down to clean the wheel. See Fig. 4.1 c for
permissible unbalance.

Fig. 4.1 c: Balance quality grades

4.2.1 Kiln exhaust fan build-up
One idea about the causes of build-up is that some particles are "sticky" at temperatures
above 300C and begin to build up on the rotor surfaces. The impact energy of the particles
striking the rotor surface (especially at an angle of 90) is also converted to heat and results
in additional softening. Other particles with a higher softening point are caught in the sticky
material and increase the coating layer thickness. The originally soft build-up gets harder
under the influence of heat and pressure (from centrifugal force and the impact of other
particles).
4.2.2 Recommendations against build-up

The fan rotor should be designed for the smoothest possible flow lines to reduce the
impact energy of dust particles. Backward curved and airfoil are the best blade forms.
Airfoil blades must be designed carefully to prevent material from getting inside and
regular inspection of the fan blades is mandatory.
Backward curved blades must be inclined enough to prevent the "hard" build-up on the
front surface and radial enough to prevent the "soft" build-up on the back-surface.
Fans should be designed for low gas and particle velocity at the fan inlet. This reduces
the impact energy of particles against the rotor and can be achieved by
1) Double inlet instead of single inlet fans
2) Larger diameter / lower speed fans
3) If possible, the peripheral speed at the rotor inlet opening should be limited to
about 76 m/s, and wheel inlet velocity should not exceed 38 m/s
4) The fan shaft should be oversized to reduce unbalance effects.
The design critical speed (considering a bearing oil film thickness and a build-up thickness of
25 mm on all leading surfaces of the rotor blades) should be at least 1.25 times the
operating speed of the fan.
Suppliers generally do not recommend to spray water directly onto a hot fan wheel, but in
practice it has been done over years with success. The rapid cooling effect removes build-up
effectively, however, the strength properties of the wheel material may be affected adversely
if the injection rate is too high.
Several different techniques are used to clean the fan, none with guarantied success
however:
Sand blasting by addition of sand to the gas flow for short periods (Fig. 4.2)
Acoustic resonance (Fig. 4.3)
Compressed air injection (Fig. 4.4)
Steam injection (Fig. 4.4)
The following measures are performed below operational fan speed or even require a fan
stop.
Knocking with pneumatic hammer
High pressure water spray (up to 300 bar)
H2O injection (Fig. 4.4) (rather delicate; it has to be carefully dealt with)

Fig. 4.2: Sand cleaning device

Fig. 4.3: Acoustic cleaning device
Fig. 4.4: Compressed air, H2O or steam Cleaning Devices

4.3 Erosion
4.3.1 Erosion types

It has been established that maximum erosion occurs when the angle of incidence between
particle and surface is between 20 and 40 degrees. This erosion, referred to as "ductile", is
thought to be the consequence of microscopic melting, which occurs when sharp-edged,
hard particles scratch the surface.
Particles striking at an angle of about 90 degrees to the surface erode according to another
mechanism. The kinetic energy of the particle creates a stress at the contact surface that
can exceed the elastic limit of the material, thus forming a surface crack. This type of erosion
is called "fragile".
4.3.2 Improper duct connection

Figure 4.5a shows a duct arrangement encountered on a raw mill system. Due to the uneven
material distribution one side of the double-inlet impeller wore out much faster.
Figure 4.5b shows the recommended modification.
Fig. 4.5a, Fig. 4.5b

4.3.3 Effects of impeller speed and wheel inlet velocity
As mentioned earlier, erosion is proportional to the square of wheel inlet velocity and to the
second to third power of the relative gas velocity w1 at the rotor inlet. With a given rotor size
this velocity is proportional to rotor speed [rpm].
As a rough guideline we can expect that the erosion will vary with the speed to the power of
3. A large fan will therefore resist erosion better then a smaller one, operating in the same
conditions, due to lower fan speed, larger surfaces.
4.4 Wear protection
4.4.1 Protection of parts subjected to abrasion
4.4.1.1 Direct protection

Increased thickness of wear parts
This should only be done when abrasion is very light (ID behind electrostatic
precipitators)
Direct hard surfacing on impeller by deposits by electrode or fusion projection
This process is very efficient if the deposits are well chosen and properly applied.
However, the thickness of the deposit is limited. The base structure of the impeller can
be affected (dilution of hard surfacing and significant addition of energy) and the
mechanical characteristics of the steel can deteriorate, especially after several
maintenance operations.

Glued-on ceramics
Their hardness is very high and they are very resistant to ductile abrasion
Temperature must be limited and the ceramics tend to come off at the blade end (high
centrifugal forces). Application on static parts is easier and more reliable.
4.4.1.2 Protection by wear plates

Corners or mild steel plates (or steel of the same grade as the base structure) added on
to the blade and on the center plate. This protection is applicable only where abrasion is
low. If the protection area is not wide enough, the base plate can also be attacked.
Wear-plates in abrasion resistant steel. This process is applied particularly on flat-bladed
impellers. The wear-plates are fitted on with countersunk screws. Replacement is easy,
but the solution can be insufficient if abrasion is very high.
Mild steel wear plates with hard surfacing by electrode or hardsurface spray coating.
This solution has the double advantage of easy replacement and high resistance to
abrasion when the type of deposit is well chosen. Moreover, hard surfacing on site is
easily carried out with no risk for the base structure.
4.4.2 Deflection of abrasive particles
4.4.2.1 Deflector plates on impeller (Fig. 4.6)

Fig. 4.6a shows the fan inlet and the rotating impeller. A, B and C are flow lines for the gas
and lines 1, 2 and 3 represent the trajectories for particles of varying size. Line 1 refers to a
very small particle that closely follows the gas flow line. With increasing particle size the
trajectories 2 and 3 deviate from the gas flow lines. The particles hit the back-plate of the
impeller and erode it.
Fig. 4.6b depicts the passage of two different particle sizes between two blades. Line 1 is
the trajectory of a small particle, line 2 of a coarser one. Most particles will hit the blade
during their passage through the impeller.
With suitably located deflector plates (Fig. 4.6c), the particle trajectories can be altered so
that no particles will hit the blades. The location and direction of the deflectors depends on
the particle size distribution, density of gas and particle, and fan speed and size.

Fig. 4.6: Wear protection by particle deflection

4.4.3 Liner materials
Due to the need for ductility in the structural members of the wheel, the wheel itself is usually
not capable of high resistance to erosion. It is therefore necessary to install liners with a
higher hardness classification to provide sufficient protection against wear.
Material hardness is an indication of its resistance to erosion. A very good liner material is
chromium carbide with an average hardness of 600 Brinell.
Figure 4.7 shows erosion test results of nine different materials.
Fig. 4.7: Erosion Test on Fan (Solyvent-Ventec)
1) Alloy of tungsten and nickel-chromium. Flame spray coating.

2) Alloy of nickel, chromium and cobalt. Flame spray coating followed by fusion.
3) Alloy of tungsten, cobalt carbides and nickel-chromium. Flame spray coating followed by
fusion.
4) Special chromium cast iron. Special electric arc welding.
5) Chromium cast iron with chromium carbides. Electric arc welding.
6) Chromium cast iron. Semi-automatic electric arc welding.
7) Ceramic powder containing basically aluminium oxide. Flame spray coating.
8) Same as 7, but of different hardness.
9) Ceramic tiles containing basically aluminium oxide. Glued to the blades.
Carbon steel sheets with a protective layer of chromium carbide are normally available in
standard sizes and various thicknesses. Cutting discs or cutting plasma can cut them to the
necessary form and size.

The liner fixation can be done by bolting or by welding the carbon steel base plate to the
existing structure of the wheel or casing.
4.5 Bearings
Bearing problems are usually caused by operating conditions and not by deficiencies in the
bearings. By following regular operating and maintenance procedures, many bearing
problems will be avoided.
Most problems are due to hot bearings. For hints see section 5.
5. FAN CAPACITY ADJUSTMENT

Fan capacity adjustments in the field are practically limited to the speed adjustment and this
within a limited range: ~ 0 - 15 % for flow increase and ~ 0 - 30 % for flow reduction.
Other capacity adjustments require mainly dimensional modifications of the fan wheel, which
have to be carried out in the workshop. Those are more costly and critical to execute.
5.1 Fan capacity too low

Important variables that are hampered by insufficient gas flow are e.g.
kiln production
separator efficiency
mill performance (throughput, drying capacity, mill venting)
Possible remedies:
Design changes in the system (reduce false air, reduce pressure drop)
the fan consumes no additional energy!
reduce/rectify false air inleaks
reduce unnecessary pressure drops caused by e.g.
not fully open dampers
improper design of inlet/outlet connections
no turning vanes in bends (refer to section 8.3)
dust deposits in ducts and bends
too narrow ducts (appropriate air speed must however be maintained to avoid settling
of dust)
Changes in fan design, speed
Speed increase
volume flow is directly proportional to fan speed
fan absorbed power rises with the third power of fan speed
fan works with lower than design efficiency
normally only possible with V-belt drives
sound emission rises with fifth power of speed
higher wear, if fan handles dust-laden gas
increased sensitivity to rotor imbalance caused by dust deposits on blades
speed increase is limited by mechanical strength of rotor
speed increase is limited by fan critical and resonant speed

Fan wheel diameter increase
practically limited to wheels where the blades do not extend to the full diameter of
the rotor sideplates (plates can then be welded-in to increase the effective fan
wheel diameter)
volume flow rises with the third power of wheel diameter
fan absorbed power rises with the fifth power of wheel diameter
Replacement of rotor with inherent low efficiency by one designed for high efficiency
e.g. replacement of radial-blade wheel by backward-inclined or backward-curved
blade wheel.
limitations with regard to diameter, width (and speed, if wheel is directly coupled
to motor via flexible coupling) of the new wheel exist if fan housing and motor are
to be kept.
New fan
Can be designed for optimum performance at the new operating point
5.2 Fan capacity too high

May be due to e.g. oversized fan, decline in production, process changes
Possible remedies:
Flow reduction by damper
Widely used solution, but inefficient and expensive (fan energy consumption)
Flow reduction by inlet vane damper
More energy-efficient than damper, but can be recommended only for flow regulation
purposes, not for permanent use at lower capacity (expensive, may be difficult to fit into
existing equipment)
Changes in fan design, speed
Lower speed
volume flow decreases proportional to fan speed
absorbed power decreases with third power of fan speed
lower sound emission
lower wear, if handling dust-laden gas
lower sensitivity to rotor imbalance caused by e.g. dust deposits on blades
Rotor width reduction
applicable if fan must deliver the same pressure as before, but at lower volume
flow
Rotor diameter / width reduction
volume flow decreases with the third power of rotor diameter
volume flow decreases proportional to rotor width
fan delivery pressure decreases with the square of rotor diameter
absorbed power decreases with the fifth power of rotor diameter
absorbed power decreases proportional to width
lower sound emission
To maintain a good efficiency, it is not sufficient to decrease the diameter only, but
also to adapt the rotor width.
For a permanent capacity reduction this is a good solution
If the old transmission is kept, speed reduction is normally reversible, i.e. fan capacity can
be increased again if necessary, whereas a reduction in rotor diameter / width are not.

6. TROUBLESHOOTING
Problem Check for

Noise 1. Squealing V-belts, due to misalignment or improper tensioning
2. Defective bearings, or bearing seal rubbing
3. Misalignment of bearing seal
4. Misaligned housing-shaft seal
5. Foreign matter in fan housing
6. Rubbing of shaft seal, wheel to inlet piece, or wheel to housing
7. Heat flinger is contacting guard
8. Coupling failure
9. Untreated expansion joints
Poor 1. Incorrect fan rotation
performance 2. Wheel is off-center, poor inlet piece fit-up allows recirculation of air
3. Fan speed too low/high
4. Poor duct design, installation of elbow or turning vanes could
remedy problem
5. Inlet damper installed backwards (counter-rotation)
6. System resistance is excessive compared to design requirements
(partially closed damper may be the cause)
7. Density may be different from design density
High bearing 1. Defective bearings caused by electrical arc due to improper
temperature grounding of nearby welding
2. Over-lubrication
3. Improper lubrication or contaminated lubricant
4. Lack of lubrication, cooling fluid, or circulation
5. High ambient temperatures or direct exposure to sunlight
6. Misalignment
7. Excessive thrust loading
8. High vibration
9. Inadvertently exchanged bearing caps (mismatched)
10. Bearing race turning inside housing
11 Moisture in bearing
12. V-belts too tight
13. Improper location ; not enough room for free axial movement of
floating bearing in its housing at elevated temperatures)
14. Heat flinger missing
Excessive 1. Motor improperly sized for fan wheel WR2
starting time
2. Inlet dampers not closed during start-up
3. Properly selected time-delay starter/fusing required (many
industrial fans take up to 20 - 25 seconds to reach operating
speed)
4. Temperature at inlet is excessively low (high density)
5. Low voltage at motor terminals
6. Inadequate system resistance

Vibrations 1. Loose bolts in bearings and pedestals, or improper mounting
2. Defective bearings
3. Improper alignment of bearings or couplings
4. Out-of-balance fan wheel
5. Loose set-screws holding wheel to shaft
6. Weld cracking
7. Improper fan wheel clearance to inlet piece(s)
8. Material build-up and/or wear on wheel
9. Ensure expansion joints in ductwork are not fully compressed
10. Misalignment or loose V-belts
11. Improper wheel rotation
12. Operation near system critical speed
13. Shaft bent or distorted during high-temperature shutdown
14. Defective motor
15. Resonant frequencies of structural steel mounting
16. Beat frequency with other fans on common base
17. Loose hub-to-shaft fit
Duct pulsation 1. Control volume with a radial inlet damper
2. Install speed variation
3. Change to a special "surgeless" blower design
High motor 1. Improper ventilation of cooling air to motor (may be blocked by dirt)
temperature 2. Input power problems (especially low voltage)
3. High amperage
4. High ambient temperature

Component Problem Probable cause, remedies
Bearings Noise 1. Imperfection in bearing elements
2. Improper clearance
3. Internal wear of bearing parts
Freezing water jacket 1. When stopping water flow in freezing
weather, blow out lower portion of bearing
housing water cavity
Fan Wheel Erosion 1. Reduce dust loading
2. Reduce rpm
3. Redesign inlet ductwork
4. Damper setting
5. Damper design
6. Better liner material
7. Alternate blade design
Buildup 1. Reduce dust loading
2. May be affected by system temperature
change
3. Blade form
4. Alternate wheel material, apply "slippery"
material
Vibration 1. Rectify build-up
2. Rectify erosion
3. Tighten foundation bolts
4. Correct misalignments
5. Improve supporting structure
6. Check effects of ductwork thermal
expansion
7. Shaft bow due to "thermal set"
8. Special considerations, refer to factory
Shaft Cracks at section - Get new shaft that is machined to eliminate
change stress raisers
Natural frequency too - Redesign shaft
close to running value
Out of round at - Replace shaft
bearing
Bowing and torsion - Refer to factory
problems
Shaft dropped or - Get new shaft
damaged during transit
or installation
Hubs Loose fit on shaft - Tighten interference fit

Insufficient stiffness - Redesign
Cracks in casting - Change to weldment
Erosion - Build up or replace

7. START-UP OF FANS
Before starting the fan for the first time, complete the following list:
1) Uncouple motor from fan and check motor (fan) for proper rotation.
2) Shut off power by disconnecting motor main breaker.
3) Check and tighten hold-down bolts.
4) Check and tighten rotor set-screws.
5) Check couplings and bearing for proper alignment.
6) Move rotor to see if it is rotating freely and maintains proper inlet piece/rotor clearance.
7) Check that fan wheel is balanced.
8) Check fan and ducts for any foreign material or dirt build-up.
9) Check that physical position of damper corresponds to indication at actuator and control
panel.
10) Secure all access doors.
11) Check lubrication of bearings, couplings, drive unit etc.
12) Couple the motor again to fan and secure and check safety guards for clearance.
13) Close dampers for adequate system resistance to prevent drive unit from overloading.
14) Supply water to water-cooled bearings and start lubrication pump.
15) Make sure that all persons are away from fan and out of any other equipment of the
system to which the fan is connected.
16) Connect electric motor by closing main circuit breaker of the motor. Start equipment
according to recommendations of drive unit and starting equipment supplier.
17) Allow fan to reach full speed, then shut down. Make immediate corrections if any
vibrations or unusual sounds have been detected.
18) During a run-in period make observations of bearings at least once an hour. Higher
bearing temperatures may occur if bearings are over-lubricated.
19) Refer to trouble-shooting guide for any unusual occurrences encountered during the run-
in period. Only after any vibrations, misalignments, etc. have been corrected, may the
fan be restarted.

8. FAN IMPELLER ARRANGEMENT AND CONNECTIONS
8.1 Assembly
8.1.1 Overhung assembly (Fig. 8.1)

This is possible if the fan operates in an environment with little risk of clogging and
associated imbalance and the impeller diameter is not too large (less than approx. 2.7 m).
Advantages:
lower purchase and installation costs than centre hung assembly
connection of upstream duct directly to fan inlet is possible (no inlet box)
Fig. 8.1: Fan arrangement - overhung arrangement

8.1.2 Center hung assembly (Fig. 8.2)
This assembly is more robust and absorbs the dynamic effects of rotor imbalance better than
the above solution.
Its use is recommended if
the dust load is high with a risk of clogging or wear of the impeller, causing rotor
imbalance
when the size and weight of the impeller makes overhung assembly delicate
Fig. 8.2: Fan arrangement - arrangement between bearings
Figure 8.3 indicates the application range for the two arrangements.

Fig. 8.3
8.2 Foundations
Fig. 8.4

1) Directly onto concrete with separate motor-bearing base-plate (this is the most
economical solution and most widely used).
2) Onto concrete with common base plate with centerline axis support maintaining the
rotor-bearing-motor alignment with the stator (more expensive solution, but easy to
install onto the foundations).
Used for hot gas fans where thermal expansion must be taken into account.
3) Common base-plate with centerline axis support resting on anti-vibration mounts (this
chassis must be perfectly rigid and is thus very expensive).
Used only in very special cases (fans mounted on steel structures).
4) With spring-supported concrete block resting on anti-vibration mounts and supporting the
fan (a spring supported block is less expensive than a common base-plate with
centerline support and allows for remarkable vibration absorption. Its weight, between 10
- 60 tons must be included in design calculations).
Used only in very special cases.

8.3 Connections
8.3.1 Inlet connections

8.3.2 Outlet connections

9. INFORMATION SOURCES
Robinson Industries, Inc., Zelienople, PA, USA

TLT-Babcock, Inc., Akron, Ohio, USA
Venti Oelde, Oelde, Germany
Solyvent-Ventec, Chalon-Sur-Sane, Cedex, France
Balance Dynamics Corporation, Ann Arbor, Michigan, USA, Fax # 313 994 3690

V4 Process Technology 2

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

V4 Process Technology 2

Încărcat de

Drepturi de autor:

Formate disponibile

Volume 4

Cement Manufacturing Course Version 2005 Volume 4 - Page 1

Cement Manufacturing Course Version 2005 Volume 4 - Page 2

1. PROCESS REQUIREMENTS FOR KILN SYSTEMS ......................................................... 6

Cement Manufacturing Course Version 2005 Volume 4 - Page 3

< 1% water dry-process

Cement Manufacturing Course Version 2005 Volume 4 - Page 4

Cement Manufacturing Course Version 2005 Volume 4 - Page 5

Table 1 Sequence of Reactions occurring in a Rotary Kiln

Temperature Type of reaction

Cement Manufacturing Course Version 2005 Volume 4 - Page 6

Process Type Feed Material Cons. Feed Moisture Feed System

Cement Manufacturing Course Version 2005 Volume 4 - Page 8

> 6000 kJ/kg cli long wet or dry kilns,

OVERVIEW OF KILN AND PROCESS TYPES

Cement Manufacturing Course Version 2005 Volume 4 - Page 9

Cement Manufacturing Course Version 2005 Volume 4 - Page 10

Cement Manufacturing Course Version 2005 Volume 4 - Page 11

40 1000 Total Group

avg kiln cap

Cement Manufacturing Course Version 2005 Volume 4 - Page 12

Cement Manufacturing Course Version 2005 Volume 4 - Page 13

3.3 Wet Process Kilns with Slurry Preheaters

Cement Manufacturing Course Version 2005 Volume 4 - Page 14

dust : 0.15 kg/kg

0.9 Nm3/kg, 180C

20mbar 50mbar 200 C

Capacity : 100 - 3600 t/d Kiln feed : slurry, 28 - 43 % H2O

CC-99072.dsf / Kma 17.2.99

Cement Manufacturing Course Version 2005 Volume 4 - Page 15

4.2 Semi Wet Process Long Kilns

4.3 Semi Wet Grate Preheater Kilns

Cement Manufacturing Course Version 2005 Volume 4 - Page 16

Chalk Quarry Clay Quarry

CC-99056.dsf Kma 17.2.99

Cement Manufacturing Course Version 2005 Volume 4 - Page 17

Cement Manufacturing Course Version 2005 Volume 4 - Page 18

clinker : 1kg cli, 1400C dust

CC-99071.dsf / Kma 17.2.99

5. SEMI DRY PROCESS

Cement Manufacturing Course Version 2005 Volume 4 - Page 19

5.2 Semi-Dry Process Grate Pre-heater Kilns

Coal Raw Meal

Clinker Cement Cement

CC-99055.dsf Kma 17.2.99

Cement Manufacturing Course Version 2005 Volume 4 - Page 20

Cement Manufacturing Course Version 2005 Volume 4 - Page 21

1.3 Nm3/kg Raw Meal

Capacity : 300 - 2000 t/d Kiln feed : meal nodules, 11 - 14 % H2O

Cement Manufacturing Course Version 2005 Volume 4 - Page 22

6.1 Long Dry Kilns

6.2 Raw Meal Suspension Preheater Kilns

Cement Manufacturing Course Version 2005 Volume 4 - Page 23

Production of Cement by the Dry Process

CC-99054.dsf Kma 17.2.99

Cement Manufacturing Course Version 2005 Volume 4 - Page 24

small units 3350 ... 3550 kJ/kg cli

kiln gas temperature approx. 1100C

Cement Manufacturing Course Version 2005 Volume 4 - Page 25

CC-99073.dsf / Kma 17.2.99

Cement Manufacturing Course Version 2005 Volume 4 - Page 26

small units, 4 stage SP 3350 ... 3550 kJ/kg cli

Cement Manufacturing Course Version 2005 Volume 4 - Page 27

1800 kJ/kg (60%) 770C