Fluidized Bed Systems

Energy R&D and Demonstration Programmes
of the
Commission of the European Communities
Fluidized Bed Systems

Proceedings of the Contractors' Meetings held in
Brussels on 12-13 October 1982
D. Reidel Publishing Company

Dordrecht / Boston / Lancaster
Proceedings of Contractors' Meetings jointly organized by
the Directorate-General for Science, Research and Development
under its Energy R&D Programme and the Directorate-General for Energy
under its programme of demonstration projects in energy saving.

Proceedings of the Contractors' Meetings held in Brussels on
12-13 October 1982
PARL EUROP. Biblioth.
N. C.
~WVK
8S6Z

A MEMBER OF THE KLUWER ACADEMIC PUBLISHERS GROUP

OP
Library of Congress Cataloging in Publication Data
Main entry under title:
Fluidized bed systems.
At head of title: Commission of the European Communities.
"Proceedings of Contractors' Meetings jointly organized by the
Directorate-General for Science, Research, and Development under its
Energy R&D Programme and the Directorate-General for Energy under
its programme of demonstration projects in energy saving"
1. Fluidization-Congresses. I. Commission of the European
Communities.
TP156.F65F593 1983 660.2'842 83-8652
ISBN 90-277-1616-1
Organization of the Contractors' Meetings by

Commission of the European Communities,
Directorate-General Science, Research and Development, Brussels '
and Directorate-General Energy, Brussels
Publication arrangements by
Directorate-General Information Market and Innovation, Luxembourg
EUR 8562
Copyright 1983, ECSC, EEC, EAEC, Brussels and Luxembourg
LEGAL NOTICE
Neither the Commission of the European Communities nor any person acting on behalf of the
Commission is responsible for the use which might be made of the following information.
Published by D. Reidel Publishing Company

P.O. Box 17, 3300 AA Dordrecht, Holland
Sold and distributed in the U.S.A. and Canada

by Kluwer Boston Inc.,
190 Old Derby Street, Hingham, MA 02043, U.S.A.
In all other countries, sold and distributed

by Kluwer Academic Publishers Group,
P.O. Box 322, 3300 AH Dordrecht, Holland
All Rights Reserved

No part of the material protected by this copyright notice may be reproduced or utilized
in any form or by any means, electronic or mechanical, including photocopying,
recording or by any informational storage and retrieval system,
without written permission from the copyright owner.
Printed in The Netherlands
PREFACE
The main objectives of the Community's energy policy are to secure an

energy supply sufficient for the present and future demand of its Member
States and to reduce the Community's dependence on imported energy through
more rational use of energy and a broader diversification of supply. This
requires a full set of common efforts at all levels, including energy
research, development and demonstration.
In the framework of its strategy for scientific and technological re-

search, the Commission of the European Communities has launched several
R&D programmes in the field of energy. These programmes are implemented
either directly in its Joint Research Centre or by concluding contracts
with research institutions in the EC member countries.
One of the most important short and medium term objectives of the present
four-year Energy R&D Programme (1979 - 1983) approved by the Council of
Ministers on 13 September 1979 aims at fostering energy conservation
technologies in the three main energy consuming sectors : domestic,
industry and transport. In addition, under the Council Regulation EEC/
1303/78 of 12 June 1978, the Commission grants financial support for
Community demonstration projects in the field of energy saving.
In the European Community, 35% of the final energy is consumed by indus-

try. About 23% of this energy is oil based. More than two-thirds of the
energy is used for heating purposes. Therefore, the development of im-
proved combustion processes and waste heat recovery have high priority in
the Community's energy saving programmes. In view of the fact that the
world coal resources are an order of magnitude larger than oil and gas
resources, improved fluidized bed combustion and gasification techniques
are being developed and applied by industry. Their main advantages are in
particular high heat transfer, low SO, levels in fumes when limestone is
added, reduced fouling and corrosion of heat transfer tubes, the possi-
bility of burning low-grade fuel and reduced dimensions when compared with
conventional coal-fired boilers.
A number of classical fluidized bed combustors working at atmospheric

pressure are already in the demonstration stage, mainly in the USA, but
general experience is not yet available. "Circulating fluidized bed
combustion" eliminates a number of drawbacks of the classical fluidized
bed combustors. Pressurized fluidized bed units offer a higher efficiency
but their application depends on the development of a gas cleaning system
and of an adequate high temperature gas turbine.
Although fluidized bed combustion is not expected to decrease the overall

energy consumption, it allows the user to reduce his dependence on im-
ported oil and gas, to provide for a better diversification of fuels and
to comply with environmental requirements.
Fluidized bed techniques may also be used in heat exchangers for waste
heat recovery, in particular for application in dirty and corrosive
industrial fumes.
In order to enhance the rational use of energy and energy conservation in

the Community, it is necessary, among other things, to promote the devel-
opment and demonstration of fluidized bed combustors and fluidized bed
heat exchangers. Therefore, the Commission has concluded with industries,
public organizations and universities of the EC Member States a number of
R&D contracts dealing with the development and demonstration of fluidized
bed combustors and heat exchangers and their application. The progress of
work achieved in these contracts was reported at this contractors' meet-
ing, and the papers presented are collected in the proceedings volume in
order to inform the Member States, research bodies and other interested
Darties about existing development work on fluidized bed combustors and
heat exchanqers.
Before ending these short introductory remarks, I would like to express my

sincere gratitude to all the specialists who accepted invitations to take
an active part in the work of this meeting by preparing a paper on speci-
fic topics.
Finally, I would like to express the hope that the work reported on during
this contractors' meeting will contribute to rapid progress in the devel-
opment and demonstration of fluidized bed combustors and heat exchangers,
thereby helping to save energy and to reduce the Community's dependence on
imported oil.
Dr. Albert STRUB

Director
Head of Energy R&D Programme
ATTENDANCE LIST
1. CONTRACTORS
R+D PROJECTS
SESSION I - FLUIDIZED BED COMBUSTION
J. BANNARD The National Institute

M. POMMEROY for Higher Education
R.C. PAYNE The National Coal Board
Dr. H. DURRFELD L. & C. STEINMULLER
R. BURROWS Stone International Ltd.
J. SCUTTER Stone International Ltd.
Dr. N. SYRED The University College
Dr. T.C. CLAYPOLE of Cardiff
SESSION II FLUIDIZED BED HEAT EXCHANGER
G. CHRYSOSTOME Creusot-Loire
M. GUIGON Creusot-Loire
R. BURROWS Stone International Ltd.
J. SCUTTER Stone International Ltd.
F. FESTA FIAT
G. VIDOSSICH FIAT
DEMONSTRATION PROJECTS
SESSION III : FLUIDIZED BED HEAT EXCHANGERS
R. DENOISEUX Metallurgie Hoboken-Overpelt

R. PUFF Cerchar
Dr. A. FOLLIOT Creusot-Loire
GRAF Lurgi
B.G.H. WEBERS NEOM - Kalkzandsteenfabriek
"DE HAZELAAR"
P. O'BRIEN Kerry Cooperative Creameries
E. ROBINSON Deborah Fluidized
Combustion Ltd.
G. BERTONI ANSALDO
D.M. WILLISS The National Coal Board
U.R. LENEL Fulmer Research Laboratories
M.A. MOORE Laboratories
W. GRILLI Italtractor
2. COMMISSION
Dr. M. DAVIS, Director DG XVII

Dr. A. STRUB, Director DG XII
Dr. H. EHRINGER DG XII
Dr. U. ZITO DG XVII
Dr. P.A. PILAVACHI DG XII
P. ZEGERS DG XII
G. HOYAUX DG XII
J. SIRCHIS DG XVII
3. CHAIRMEN
Dr. J. CARRASSE SYNTEC, France

Ir. J.A. KNOBBOUT TNO, The Netherlands
C O N T E N T S
Preface V
Attendance List VII
Summary of the proceedinqs 1
PART I - R & D PROJECTS
Contractors' meeting held on 12 October 1982

Chairman: J.A. KNOBBOUT
The desiqn and development of a small-scale fluidised bed

boiler with automatic control
National Institute for Higher Education
J.E. BANNARD, A. KAY, M. POMEROY 16
Production of hot qas for drying processes using fluidised

bed combustion of coal
National Coal Board - Coal Research Establishment
R.C. PAYNE, J. HIGHLEY, A.R. WINDYBANK 29
Further exnerimental development of circulating fluidised

bed combustion for application in steam boiler furnaces
usinq low-qrade fuels and coal
L. & C. Steinmiiller GmbH
H. DUERRFELD 50
The burninq of RDF and industrial wastes in fluidized bed

industrial boilers
Stone Fluidfire
R. BURROWS 68
The development of a demonstration furnace for the

combustion of low c.v. gases
University College, Cardiff
T.C. CIAYPOLE, C.J. BATES, N. SYRED 75
SESSION II - FLUIDIZED BED HEAT EXCHANGERS

Chairman: J. CARRASSE
Amelioration des echanges thermiques dans les lits

circulants
Creusot-Loire - Division Energie
G. CHRYSOSTOME, M. MARCELLIN ' 88
Etude de la circulation autoentretenue

Creusot-Loire and Universite de Compieqne
J.-F. LARGE, P. GUIGON, G. CHRYSOSTOME 96
Fluid bed high temperature gas/gas heat exchangers

Stone Fluidfire
J.N. SCUTTER 105
Fluidized bed heat exchanger filter for waste heat recovery

from dirty corrosive gases
Centro Ricerche Fiat
C. DALLA ROLLE, D. PETRUCCIOLI, R. MARGARY, R. FESTA,
S. VACCARO 120
PART II - DEMONSTRATION PROJECTS
SESSION III - FLUIDIZED BED HEAT EXCHANGERS

Chairman: M. DAVIS
Opening address
M. DAVIS, Commission of the European Communities 139
The M.H.O. process for the combustion of shales in fluidized

bed
Metallurgie Hoboken-Overpelt
J. DE KEYSER, R. DENOJSEUX 140
Combustion of carbonaceous washery shales in fluidized bed

CERCHAR
R. PUFF , 149
Total precalcination by circulating fluid bed

LURGI Umwelt und Chemotechnik, Creusot-Loire and
Lafarge Coppee
A. FOLLIOT 162
Fluidized bed boiler used in the production of bricks

De Hazelaar
B.G.H. ngEBERS 173
Multifuel fluidized bed exploiting indigenous solid fuels
Kerrv Cooperative Creameries Ltd.
P. O'BRIEN 188
Fluidized bed combustion of waste acid tars

Deborah Fluidized Combustion Ltd.
E. ROBINSON 202
Atmospheric fluidized bed combustion boiler for electric

power generation by coal
ANSALDO - Divisione Breda Generazione Vapore
G. BERTCNI, L. GUIDI 214
The application of fluidized combustion of coal to a low

cost, large industrial water tube boiler integrated in a
combined heat and power scheme
National Coal Board
B.J. SMITH, D.M. WILLIS, J.M. TOPPER 231
Use of fluidised bed furnaces in steel heat treatment

Fulmer Research Laboratories Ltd.
U.R. LENEL, M.A. MOORE 248
Metallurgical treatment (quenching and tempering) in fluidized

bed of track chain elements recovering the residual forging
heat
ITALTRACTOR
W. GRILLI 261
Closing address
Dr. M. DAVIS, Commission of the European Ccmnunities 278
SUMMARY OF THE PROCEEDINGS
P. Zegers , J. Carrasse and J.A. Knobbout
INTRODUCTION
In the European Community, R and D on fluid bed combustion is carried out

1n the second Energy R+D programme in Directorate General XII for Science,
Research and Development. Fluidized bed demonstration projects are
carried out in the frame of the Community demonstration programme for
energy savings (DG XVII). On 12 and 13 October the contractors
working on fluid beds in the Energy R+D programme and the Demonstration
scheme were invited to present a paper on the progress of their work.
Nine papers on R+D projects and ten papers on demonstration projects were
presented. The proceedings of this two day contractors meeting are given
in this book and below a summary will be given of the papers presented and
the discussion.
STATE OF THE ART
Before discussing the progress of these projects a short description of

the principles and state of the art of fluid beds will be given.
A fluid bed is basically a container with fine inert solid particles. In

the bottom of the container holes are made through which air can be blown
in upward direction. If the air velocity is high enough the particles
can be lifted and float in the air in a turbulent motion which resembles
a fluid. If this bed is brought at 700-900 C and preheated solid fuel is
introduced, combustion takes place. Normally only 2-3 X of the fluid bed
mass consists of fuel; other materials are inert particles and ash.
The heat released in the fluid bed can be removed in three ways :
. Heat can be extracted with heat transfer tubes in the bed (e.g. in order
to heat air for gas turbines or to raise steam). The heat transfer in
the fluid bed is very efficient as it happens not only by radiation and
convection as in conventional boilers, but also by the continuous impact
of the hot particles in the bed. The rate of heat transfer in atmosphe-
ric fluid beds (AFBC) could thus increased by a factor five as compared
with conventional heat exchangers, this allows a more compact installatior
. For the case of high excess air levels heat can be removed as sensible
heat from the exhaust gases. This system is used in pressurized fluid
beds (PFBC) where the pressurized hot flue gas may be used to fire a gas
turbine for power generation. Heat extraction from flue gases is also
used in external circulating AFBC which will be discussed below.
. Often a consistent part of the heat is extracted from ash and entrained
particles; in particular for fluid beds with high excess air levels.
* Commission of the European Communities

** SYNTEC-Ingenierie, France
*** TNO, Netherlands
l-
The advantages of fluid beds as compared to conventional combustion are :
. Combustion takes place at relatively low temperatures (700-900 C ) .

This results in low N0 X levels and less material problems (e.g.
corrosion);
. More efficient heat transfer, which results in a more compact instal-
lation;
. Very efficient combustion as the contact between fuel and air is optimal;
. No or limited need for coal treatment such as pulverization;
. Low calorific fuel (e.g. biomass, wood) and waste can be burned.
Combustion of waste leads to cost savings for fuel and waste disposal;
. The SOg content in exhaust gases can be reduced by as much as 90 X by
adding Limestone (CaC03) in the bed. This allows the use of high
sulphur coal. The costs involved are small compared to stack gas
cleaning;
. Retrofitting of conventional oil boilers with a coal fired fluid bed
is possible. The fluid bed installation has lower fuel costs and
typical payback times are 3-5 years.
disadvantages are :
. The starting up of a fluid bed takes long time (1-4 hrs). It can be
realized by preheating air which is blown through the bed or by heating
the bed by oil or gas burners above the bed;
. The heat output in. a conventional fluid bed can be regulated only by
25 % (turndown value 4 : 3 ) . Much progress however is made and in the
E.C. programmes several projects are carried out which try to solve this
problem : by using several beds which can be slumped separately, by
introducing external circulating fluid beds, by using fluid beds with a
varying heat exchange surface;
. Dynamic response of the heat output for a conventional bed is slow due
to the large thermal capacity of the bed. Use of external circulating
fluid beds, which will be discussed below, may solve that problem;
. Large quantities of ash in particular with desulphurization;
. High energy cost of fans which blow air through the bed;
.. Special feed preparation may be required for fluid bed combustion of
waste;
. Investment costs are high'but lower fuel costs make fluid beds already
now often attractive;
. Erosion of parts in the fluid bed such as heat exchanger tubes by
abrasion of inert particles; in particular at high gas velocities.
Much of the R, D and D described in this book deals with the above
mentioned disadvantages : start up, turndown, dynamic response, feeding
and erosion. Another important topic is circulating beds.
The study of iyiatiD9_liyl^_^^5 is a fairly recent but important
development; 6 of the 9 R+D projects and 3 of the 10 demonstration pro-
jects deal with this subject. The main features of circulating beds will
therefore be shortly discussed. One may distinguish two cases :
For the case of low excess air and a fluid bed with a well defined bed
surface an internal_ciculation in the bed is possible, which may be
influenced by the shape of the bottom, local introduction of a stronger
air flow and the shape of in bed heat exchanger. Such a circulation
brings about a better fuel mixing, gives a better burnout and the tempera-
ture in the bed is homogeneous.
For the case of high excess air small particles are entrained from the
bed with the flue gas. These hot particles are separated from the flue
gas by a cyclone and reintroduced in the bed possibly after heat extrac-
tion with a heat exchanger. There is not a clear surface in this external
llyi2iiD9 fluid bed. We have now a lower part which contains the
heavier particles and where combustion takes place and the upper part with
entrained particles. This fast circulating bed has the following advantages
- By regulating the particle flow the heat output can be regulated down
to 15 X of the maximum heat output.
- By placing the heat exchangers in the upper part of the fluid bed the
heat transfer and combustion part can be separated.
- Fuel treatment is not required because both small and large pieces are
completely burned.
- High gas speeds give a higher heat transfer (but more erosion), which
leads to a more compact installation.
- The temperature in the bed is homogeneous.
Large installations are mainly used for the production of power. For
cogeneration or power production by fluid beds in future, pressurized
fluid bed combustors seem to have the best chances. Atmospheric fluid
bed combustors are mostly used when lower heat outputs are required. Here
combustion of waste material is an application. Retrofitting of fluid
beds in conventionally fired boilers is possible.
The excellent heat transfer properties of fluid beds make them very
efficient heat exchangers. The development of fluid bed heat exchangers
is carried out in several E.C. R+D projects.
Finally it should be clear that fluid bed combustion will not lead to
energy saving as compared to conventional coal combustion (with a possible
exception for sophisticated PFB units). Consumption of electricity by the
fan may in fact increase the energy requirements somewhat. Fluidized bed
combustion however does allow users to switch from scarse premium fuels
such as oil and gas to more abundant fuels such as coal. It can also
burn waste materials quite efficiently.
-3-
R + D ON FLUIDIZED BED COMBUSTORS AND HEAT EXCHANGERS
CARRIED OUT IN THE FRAMEWORK OF THE COMMUNITY'S ENERGY R + D PROGRAMME
Research on fluid beds in the E.C. energy R+D Programme is carried out on
five fluid bed combustors and four fluid bed heat exchangers (See table 1 ) .
Fluid bed combustor research is exclusively done on atmospheric fluid beds.
Interesting work is carried out on a 25 kW internal circulating AFBC for
small commercial and domestic applications. Design concepts are being
tested for a double gasifier/combustor fluid bed producing hot gas at
1250 C and a conventional fluid bed producing dust free hot gas of 600 C.
Furtheron an external and an internal circulating fluid bed are being
developed.
For the fluid bed heat exchangers two projects deal with heat exchangers
in a fluid bed and two projects investigate fluid bed gas-gas and gas-
water heat exchangers.
Fluid bed combustors
In view of the large potential for small scale solid fuel boilers, the
NATIONAL INSTITUTE FOR HIGHER EDUCATION, Ireland is developing a 25 kW
atmospheric fluid bed combustor (AFBC) which is meant for domestic and
small scale industrial and commercial applications. The AFBC will be
cheap, will have a high turndown ratio (50 : 1 ) , a short start up time
(20 minutes) and will be fully automatic. Within the fluid bed (bed
area 10 x 20 cm) an internal circulation is brought about by a sloped
bottom and a locally stronger air flow in the upper sloping part of the
bottom. Such a circulation brings about an efficient mixing and combus-
tion of fuel with a varying diameter. The heat is extracted by fluid bed
walls which are water cooled. The high turndown value can be obtained by
varying the height of the bed thus varying the heat transfer surface and
by partiaLly slumping the bed; which is enabled by the sloped bottom.
The start up heating is done electrically with a heater of 1 kW. The
AFBC will be operational in May 1983.
Production of hot gas for drying purposes represents a significant use of

energy in the E.C. countries. Presently coal fired fluid beds are
already being used commercially for hot gas applications up to 950 C
where contamination by ash is acceptable. The NATIONAL COAL BOARD, U.K.
is experimentally evaluating design concepts for fluid beds which are
able to produce a) hot air'containing ash at 1250 C and b) dust free
gas at 600 C.
a) In a test installation of 400 kW the concept design for a fluid bed

producing flue gas at 1250 C is experimentally tested. It consists
of two adjacent fluid beds, the first being operated as a partial
gasifier producing low c.v. gas which is burned in the oxygen containing
off-gas from the second bed. The second bed is fueled by the char
from the incomplete gasification of the first bed and is operated with
a high excess air level as a conventional fluid bed. Preliminary
tests obtained 1200 C, further experiments should demonstrate the
stable operation of this system under a range of test conditions.
b) The concept for the production of hot, dust free gas at 600 C was
tested in a 200 kW unit. Air is heated by passing it through heat
exchanger tubes in and above the fluid bed. The mullite tubes used
in the in-bed heat exchanger broke down several times. A study
suggested to use silicon carbide tubes which during 1000 hrs of
testing have proved to be satisfactory. Also the second phase of
the project, the design for a full size 1,8 MW installation for a
malting factory, has been completed.
A further development of a fast external circulating fluid bed for

production of steam is carried out by STEINMULLER, Germany. The advanta-
ges of such a circulating bed have been described previously. The main
emphasis in this project is low fuel treatment (possible as in this type
of fluid bed both small and large pieces are burned efficiently) and a
good burn out which is ensured by the longer and effective mixing in the
bed. The pilot unit has a capacity of 1 MW, will contain desulphurization
facilities, will produce steam of 450 C and 20 bar and will be fired
with coal and low grade fuels. The plant was to become operational in
December 1982.
The fluidized bed combustion of refuse derived fuel (RDF) and industrial
waste is being tested by STONE FLUIDFIRE, U.K. in a 300 kU fluid bed
installation with an internal circulation of the bed induced by the sloped
profile of the bottom. This circulation ensures a good mixing of the fuel
and a good burn out. It is important to introduce the fuel in the central
down region so that the fuel is entrained and stays a maximum time within
the bed before it reaches the surface. Still a high proportion of the
volatile hydrocarbons is released so fast, that it is burned above the
bed resulting in a high CO content (0,5 X ) . In order to avoid this, air
is also introduced above the bed. For the raw RDF a special pneumatic
feed system had to be developed. Waste generally has a low calorific
value and may be wet. Addition of coal or propane is therefore necessary.
It was possible to sustain combustion with 60 X of the heat input provided
by waste residue. The installation has a efficiency of 80 X if pelletized
fuel (RDF) is used and started testing in January 1982.
Fluid bed heat exchangers
In the R+D programme on fluid bed heat exchangers different types of heat
exchangers are being investigated. Two projects study heat exchangers in
fluid bed combustors. Two other projects investigate a gas-gas heat
exchanger and a gas-water heat exchanger based on the fluid bed concept
with the inert particles as a heat exchange medium.
-5-
CREUSOT LOIRE and THE UNIVERSITY OF COMPIEGNE, France are currently
optimizing the performance of a circulating fluid bed heat exchanger.
The fluid bed is of the external circulating type. The high gas speeds
in such a fluid bed result in better heat transfer coefficients. Another
major advantage fs the possibility to install the heat exchanger in the
upper part of the combustor, thus separating heat transfer from combustion.
Subject of the investigations are :
. the influence of the overall geometrical configurations of the heat
transfer surface on the pattern of the solid flow and on the heat
transfer coefficient;
. the influence of the tube shape and of the geometry of finned tube
surfaces (longitudinal or spiraLed fins).
A first series of tests showed that instabilities of the diphasic gas/
solid flow can strongly affect the overall performances of the heat
exchanger : suspension flow stability and gas flow friction are the two
main criteria used to compare various surface geometries.
Another project by CREUSOT LOIRE and THE UNIVERSITY OF COMPIEGNE, France

deals with the influence of an in-bed heat exchanger on the internal
circulation in the bed. In large fluid beds a natural circulation occurs
with an upward movement in the middle of the bed and downward on the
sides. In this study the self sustained circulation induced by a compact
in-bed heat exchanger and the influence of the shape of the heat exchanger
is investigated. Such a circulation allows the horizontal heat exchanger
tubes to be very close. The heat exchanger is thus very compact and only
a small part of the bed is needed for heat exchange. Heat transfer and
combustion in the bed are more or less decoupled. By varying the
circulation flow also the heat output can be regulated. First experiments
gave satisfactory results.
A fluid bed gas-gas heat exchanger is being developed by STONE FLUIDFIRE,

U.K. The heat exchanger consists of two adjacent fluid beds. The
separating wall between the two beds has openings at the height of the
fluid bed. This allows the fluid bed mass of the two beds to make a
horizontal circular movement which is induced by blowing hot gas and
cold air through the two respective beds with a horizontal component.
The partition plates allow a free passage of particles from one section
to the other but minimize gas leakage. The heat exchanger operates with
a parallel flow system and has an efficiency of 50 X. Different bed
geometries have been tried- in small scale prototypes with gas temperatures
up to 250 C. The gas leakage can be controlled within acceptable limits.
Testing will be continued up to 800 C. The most apparent advantage is
its simplicity. Potential markets are combustion preheating and space
heating with hot flue gases.
Heat recovery from diesel exhaust gases or, generally, from factory hot
flue gases has a serious drawback because of the particulate contents.
In conventional heat exchangers, tube fouling leads to a rapid decrease
in heat exchange efficiency. FLuid bed heat exchangers overcome this
inconvenience. FIAT, Italy designed a semi industrial prototype of 62 kW
-6-
where the bed is fluidized by soot containing exhaust gas of a diesel
engine at 500-600 C and the heat is extracted from the bed with cooling
water circulating in an in-bed heat exchanger. After 30 hours of
operation, the efficiency of the fluidized bed heat exchanger only falls
about 3 X. Top values of heat transfer coefficient can be reestablished
inducing tube defouling by stepping up and down the tube wall temperature.
Experiments showed that soot particles are agglomerated in the fluidized
bed at diameters 10-100 times larger than those at the engine outlet.
The efficiency of soot removal is strongly improved by this agglomeration
of the particles. Bed temperature has a small influence on soot removal
efficiency. The engine air/fuel ratio of the diesel engine however
strongly affects the agglomeration and consequantly the removal efficiency
A higher air/fuel ratio results in a better adhesivity of the soot to the
heat exchanger tubes.
COMMUNITY'S DEMONSTRATION PROGRAMME
Introduction
In the Community's Demonstration Programme carried out by Directorate

General XVII for Energy, ten projects are carried out on fluid bed
combustors (see Table 2 ) . Three projects deal with the combustion of
coal waste with a high ash content (up to 82 X ) . These fluid beds pro-
duce heat for cogeneration, steam production and manufacturing of
ingredients for cement production. Three fluid beds use normal coal as
a fuel; one produces steam for the brick industry and two are used for
cogeneration. Furtheron work on an external and an internal circulating
bed is carried out; both burn waste or low calorific fuel. Finally two
beds are used for heat treatment of steel.
Projects
An atmospheric fluid bed furnace was constructed in Overpelt by

METALLURGY HOBOKEN-OVERPELT, Belgium. It is fired with coal waste which
is brought from coal washings at Waterschei. The heat is used for the
production of steam (42 bar, 410 C) which is used in a 3,3 MWe cogene-
ration plant consisting of a backpressure steam turbine. The combustion
of the wet (10 X H 2 0) coal waste with a low calorific value (3,6 MJ/kg)
and a high ash content (up to 82 X) was very satisfactory. Heat was
extracted from the ash and the flue gases. The ash after combustion
could be used for concrete and as an ingredient in cement. Desulphuri-
zation was done with CaO + MgO which is naturally present in coal waste.
The SOj content was decreased with 80 X. This project is economically
attractive also in case the user has to bear the transport charges of the
coal waste.
-7-
A similar installation is being built by CERCHAR, France. This atmospheric
fluid bed combustor is to produce steam of 30 bar and 300 C which present-
ly is produced with a coke gas fired boiler. The waste from coal washings
has a calorific value of 2,5 - 5 MJ/kg and an ash content of 72 to 88 X.
The expected production capacity is 17 tonne steam per hour and the
required coal waste lies between 12 and 28 tonne/hour. Like in the
previous project crushing or milling of the material is not required.
The distance between the fluid bed combustor and coal washing is 500 m.
The overall efficiency of the plant is calculated to be 57 X, with losses
in the fluid bed combuster and boiler of 12,7 X and 19 X respectively.
Small scale experiments demonstrated the technical feasibility of this
installation which will be operational in October 1983.
Fluid beds can bring about considerable energy savings in the production
of cement; in particular if they are fired with coal waste. Cement is
milled clinker, which is produced by mixing CaO and clay (which contains
SiOj, AI2O3 and Fe203> and heating this mixture at 1200 C. CaO is
produced by heating CaC03 (limestone) at 890 C. This precalcination
which is energy intensive,is normally done in a conventional oil or gas
fired furnace. It is this part of the process which will now be carried
out in a fluid bed combustor, fired with coal waste (of which the ash has
approximately the same composition as the raw material for clinker pro-
duction). In this process it is important to realize a high degree of
decarbonization. In preliminary experiments it was shown that 98 % of
CaC03 is transformed into CaO. A prototype external recirculating AFBC
producing 1250 tonne of raw material of cement per day, is now being
built. About 70 % of the heat required for precalcination will be
delivered by coal shales. Energy savings for the overall clinker pro-
duction is estimated to be 32 X or 50 kg coal per tonne of clinker. Also
the use of vegetal waste as fuel will be tested. Finally it will be
investigated whether fluid bed combustors can also be used for the clinker
production. This project is being carried out by LURGI, CREUS0T-L0IRE,
LAFARGE-COPPEE, France.
A 8 MW coal fired AFBC was built for the brick works "DE HAZELAAR" in the
Netherlands. It produces steam of 210 C at 18 bar. The heat output of
this plant can be reduced to 25 X of its maximum output. The reduction of
the SO2 emission is 80 X and the ash of the fluid bed can be used for
brick production. The calculated internal rate of return for the instal-
lation is 11 % with a boiler utilization of 33 %. (Not taking into
account funding by the C E O . The AFBC was commissioned in August 1982
and a three year measuring programme is presently going on.
A fast external recirculating fluidized bed combustor of about 40 MW was

chosen by KERRY COOP, IreLand to replace an oil based conventional heating
system which produces 47 tonnes steam per hour at 224 C and 24 bar for a
factory of milk products. This fluid bed is designed to use a wide
variety of low calorific fuels such as peat, waste, wood and coal. The
industrial process requires a very fast dynamic load response (10 X of the
maximum heat output per minute) and the fluid bed should be able to operate
at 15 X of the maximum load during the winter. These requirements can be
fulfilled with the external circulating fluid bed, which is able to reach
turndown values of 7 : 1 and to rapidly fluctuate the heat output. The
heat is for 50 X recovered from the flue gases and for 50 X from the
solids extracted from the flue gases by a cyclone before they are again
fed into the fluid bed. The installation will be ready for operation in
March 1984.
A 3,5 MW internal circulation bed operating at 900 C was constructed by

DEBORAH FLUIDIZED COMBUSTION LTD, U.K. The internal circulation in the
bed, which is brought about by a sloping bottom, gives a homogeneous
distribution of the fuel even with very few feeding points. Also the
temperature in the bed is homogeneous. The bed can burn any fuel with a
calorific value higher than 11,65 MJ/kg. Heat is extracted from the bed
by water tubes in the container walls and there is no need for in-bed
heat exchanger tubes. The heat output can be regulated by expanding the
bed, thus in increasing the heat exchanger surface. The turndown ratio
is 2 : 1. The start up of the bed takes 4 hours. This plant will burn
waste acid tar, which has the advantage that fuel is free of charge and
that waste disposal costs are saved. This results in a very short payback
time of one year. Several plants have already been sold.
ANSALDO and TOSI, Italy are constructing an AFBC which produces steam at
475 C and 59 bar for a cogeneration plant of 15 MWe consisting of a steam
turbine-generator. The fluid bed consists of four compartments each of
which can be stopped operating. In this way the heat load can decreased
to about 25 X of the maximum heat output. The start up of the bed takes
one hour and heat is extracted from in-bed tubes and flue gases.
Limestone is used for desulphurization. The plant will be operational
in 1984.
The NATIONAL COAL BOARD, U.K. built the first AFBC fired with coal as a
fuel, with a "once through" coil boiler. It is a 30 MW installation which
produces 36 tonne/h steam of 440 C and 45,5 bar for a cogeneration plant.
The system should be able to respond to rapid load changes and have a high
turndown (6 : 1 ) . This high turndown can be obtained by slumping one or
two pairs of the six beds which are arranged in pairs. Half the heat is
extracted from the bed by pan cake heat exchanger coils and the other half
by heat extraction with a superheater and a economizer from the hot flue
gases. The plant became operational in 1981. Tests showed that the
efficiency at maximum load was lower than expected. This was due to a
greater carbon loss than expected. This problem could be partly solved by
using shallower beds and coarser sand. In 1982 several tube ruptures
occured in the in-bed heat exchanger. It was decided to start a major
redesign of bed coil arrangements.
Heat transfer in fluid beds is very efficient. This makes fluid beds an
interesting option for heat treatment of steel products at 800 - 1000 C.
FULMER RESEARCH LAB, U.K. compared the efficiency of heat treatment in a
fluid bed and a salt bath furnace which were both gas fired and an electric
seal quench furnace. The measured efficiencies for the three ovens were
5 X, 5 X and 15 X respectively. The reason for the low efficiency of the
fluid bed was the fact that heat from flue gases could not be recovered.
A change from batch to continuous operation would increase the efficiency
of the fluid bed considerably.
Considerable energy savings can be achieved when a part of the heat

treatment process (heating and tempting at 850 C) of tractor chain
elements is carried out with a fluid bed which replaces a conventional
furnace. The fluid bed has the advantage of a 510 times better heat
transfer and a homogeneous bed temperature within 5 C. Heat treatment
of 17000 tonne/year of tractor chain elements treated in this way will
bring about energy savings of 1000 TEP/year. The new process also gives
an improved product quality. The project is carried out by ITALC0NTRACT0R,
Italy.
GENERAL DISCUSSION
Following the presentation of the papers on R, D and D on fluid bed

combustion/ a general discussion took place on future R+D and the economic
feasibility of fluid bed demonstration projects. The discussion on future
R+D was of particular interest for the Commission as it is presently
preparing its third Energy R+D Programme.
Mr. Knobbout introduced the subject and gave a list of areas where future
R+D is still needed : starting up of the fluid bed, high turndown values,
fluid bed combustion of waste, fuel feeding system, erosion, lifetime.
To his opinion fluid beds are now on the treshold of economic feasibility.
Erosion was one of the most discussed topics. The problem often occurs
with inbed heat exchangers and mostly pragmatic ad hoc solutions are
found. The more fundamental reasons for erosion however are often not
understood. B asic research on erosion and in particular on its long term
effects is very important. Presently it is not possible to guarantee
lifetime of fluid beds for longer than 35 years. If lifetimes could be
increased to 1020 years the economic feasibility would improve a lot.
The lifetime could be considerably improved by being careful (e.g. by
using low gas velocities) but often high gas velocities are indispensible.
An external circulating fluid bed was mentioned which operated 11 years
without problems. This demonstrates that it is advantageous as far as
erosion is concerned to have combustion and heat transfer separated.
Another topic in the discussion was combustion of waste. There is a large

interest for this application of fluid beds. In France there is a parti
cular interest for fluid bed combustion of coal with a high ash (up to
80 '/.) and sulphur content.
The feeding systems for fuel and limestone are expensive and often not
reliable. Cheaper systems should therefore be developed which give a
regular and reliable supply and which give the correct distribution in the
bed. This is very important both for optimal combustion and desulphuri
zation.
10
Other suggested future research areas are : development of lower cost
systems of gas particulate cleaning systems and the problem of the slow
start up of combustors.
From the discussion on the economic feasibility of fluid bed systems one
may conclude that these systems are on the verge of economic feasibility
under normal conditions (e.g. steam production with coal as a fuel).
It will be considerably improved if longer lifetimes can be obtained.
Erosion is here one of the main problems. In special cases fluid bed
combustion may be very attractive. An example of such a case is a fluid
bed which burns waste tar, thus saving the cost of the disposal of this
waste product and the cost of fuel. This leads to a payback time of one
year.
The final part of the discussion was focussed on the question : which
fluid bed applications are presently in a demonstration phase. The
Commission is supporting demonstration projects in order to speed up
commercialization of the FBC. It welcomes proposals, in particular the
ones where several partners from different E.C. countries have a vested
interest. One proposal concerned a 7 MW fluid bed producing clean hot
air for drying processes in malting works. Another important area is
the production of electricity.
11-
TABLE 1 : OVERVIEW OF FLUID BED R+D PROJECTS
Fuel or heat Special features Ready for

Contractor Type of FBC Size R+D source or problems operation
FLUID BED COMBUSTORS

National Institute Development of a Turndown ratio
for Higher Education, Internal 25 kW coal May-1983
ci rculating very small AFBC 50 : 1
Ireland
a) Gasifier/ Production of flue Test i n s t a l -
400 kW coal
National Coal Board, combustor gas at 1250 C lations
U.K. Use of SiC tubes O p e r a t i o n a l
Production of flue for in-bed heat
b) Conventional 200 kW coal
gas at 600 C exchanger :
External Testing of exter-
Steinmuller, D coal December
circulating 1 MW nal circulating 1982
fluid bed
Testing of com- Refuse Problems with
Stone Fluidfire Ltd, Internal January
bustion of refuse derived high CO content
U.K. ci rculating 300 kW 1982
derived fuel fuel have been solved
FLUID BED HEAT EXCHANGERS

Testing of the
Creusot-Loire External cir- above bed heat coal
Un. Compiegne, F culating AFBC exchanger
Influence of heat
Creusot-Loi re Internal
exchanger on coal
Un. Compiegne, F ci rculating
internal circul.
Stone Fluidfire Ltd, Fluid bed gas-gas Testing of fluid
U.K. heat exchanger bed heat exchang.
Fluid bed gas- Testing of the Diesel
Fiat, It. water heat heat exchanger exhaust
exchanger (e.g. fouling) gases
TABLE 2 : OVERVIEW OF DEMONSTRATION PROJECTS
Special features Ready for

Contractor Type of FBC Size Application Fuel
or problems operation
Steam for cogene- Waste coal

Metallurgie
Conventional + 10 HW ration coal waste treatment not Operational
Hoboken-Overpelt, B th
410 C, 42 bar required
Steam Waste coal
Cerchar, F Conventional + 15 MW coal waste October 1983
th 300 C, 30 bar treatment not
required
Lurgi
External Cement
Creusot-Loire coal waste
circulating ingredients
Lafarge-Coppee, F
Steam for brick Turndown 4 : 1
De Hazelaar, N Conventional 8 MW manufacturing coal Fast dynamic August 1982
210 C, 18 bar load response
Steam for mi Ik
External wood, peat
Kerry Coop, Irl. 36 MW products Turndown 7 : 1 March 1984
circulating coal waste
224 C, 24 bar
Deborah Fluidized Internal Turndown 2 : 1

3,5 MW Steam acid tar Start up 4 hrs Operational
Combustion Ltd, U.K. circulating
paybacktime 1 yr
Four bed compart' Steam for cogene
Ansaldo-Tosi, It. ments which can be + 45 MW ration coal Turndown 4 : 1 1984
slumped separately 475 C, 59 bar
Six bed compart- Steam for cogene- Tube ruptures
National Coal
ments which can be 30 MW ration coal in bed heat Operational
Board, U.K.
slumped in pairs 440 C, 45 bar exchanger
Fulmer Research Heat treatment of
Batch type furnace 5 X efficiency Operational
Lab, U.K. steel products
Heat treatment of
Italcontractor, I Operational
steel producs
PART I - R & D PROJECTS
Chairman: J.A. KNOBBOUT
The desiqn and development of a small-scale fluidised bed

boiler with automatic control
Production of hot gas for drying processes using fluidised

bed combustion of coal
Further experimental development of circulating fluidised

bed combustion for application in steam boiler furnaces
usinq low-qrade fuels and coal
The burninq of RDF and industrial wastes in fluidized bed

industrial boilers
The development of a demonstration furnace for the

combustion of low c.v. gases
15-
THE DESIGN AND DEVELOPMENT OF A SMALL-SCALE FLUIDISED BED BOILER WITH
AUTOMATIC CONTROL
Authors: J E BANNARD, A KAY, M POMEROY.

Contract Number: EE-D-1-302-IR
Duration: 27 months, 1 October 1981 - 31 December 1983.
Head of Project: Dr J E Bannard
Contractor: National Institute for Higher Education
Limerick
Ireland
Summary
Tne conceptual design for a small-scale FBC boiler is well advanced and
has been aided, by tests in cold models on fluid dynamics, bed cycling
rates, elutriation, segregation and bed start-up. Many of tne sensors
and actuators have been sourced but decisions on others are still to be
made. The building of a prototype boiler is about to commence.
1. Introduction
The depletion of reserves and increases in prices of oil have encouraged
research into the comDustion of coal, because of the high cost of
ancillary equipment the application of fluidised-Ded combustion has been
restricted to larger boilers. Nevertheless, in tne context of the needs
of the smaller industrial and commercial enterprise (whicn represent the
majority of enterprises), it is believed that there exists a large
potential for small scale solid fuel boilers. However, improved Doiler
efficiencies are necessary and significant reductions in the time needed
for maintenance and attention must be made. It is proposed that a F6
Doiler be built of size appropriate to the need of the small industrial,
commercial or even domestic unit, where a number of interesting proDlems
are anticipated. For example, the boiler will need a very hign degree of
reliability and the minimum of attention in operation; it will need to be
fully automatic; it will need a rapid on/off facility and/or a high
turn-down ratio; and finally it will need to be simple in design and
cheap to produce. The objective of this project is to design, build and
assess a FB boiler of approximately 25kw (thermal) output with high
reliability and fully automatic control. Tne program is divided into
three steps as follows:
Stage 1 (Commence Octooer 1981)
Conceptual designs will be considered in respect of the following
elements: boiler size and shape; method of removing heat; type of fuel to
be handled; type of distributor plate; size of coal to be handled;
solution to elutriation problems; solution to ash handling problems.
Sensors and actuators will be developed and tested: electric start-up
heater; vibratory feeders; Og (and CO) detectors; air distribution
actuator.
Stage 2 (Commence May 1982)

Cold models will oe constructed followed by construction of tne prototype
boiler to test the following parameters: bea depth; fluidising velocity;
excess oxygen; carbon losses; neat transfer rates; start-up procedure;
turn-down ratio; influence of type/quality of fuel; asn removal
parameters.
Stage 3 (Commence May 1983)
The prototype boiler will be completed and evaluation runs will be
carried out. These will include: influence of type of coal; reliability
of control system; reliability of mecnanicals; extent of turn-oown and
response of start-up and shut-down; neat transfer rates; and general
performance. Additionally the performance will be evaluated with
reference to the fuel requirements and availability (e.g. grade and size
of coal); the amount of attention required; and the cost of production of
the boiler.
Small reserves of good-quality anthracite exist in Ireland in addition to
the larger reserves of poor-quality coals and coal shales, and a research
program is underway to explore the possibility of burning these fuels
using F8C. However, it is not anticipated that tnese fuels will Decome
available to a wide market or that they will make mucn impact on tne
marketing structure tor imported Dituminous coals. Accoraingly, the
small-scale boiler is to De built to ourn principally low-sulpnur
<^
- 17-
bituminous coals.
Coal continues to make inroads into the oil market in Ireland as in other
EEC Countries. Between 1976 and 1978 the numDer of domestic oil-fired
central heating systems dropped from 74% to 59* in Ireland whilst those
burning coal rose from 23% to 33%. Nevertheless only 25% of houses in
Ireland boasted a central-heating boiler at the end of 1978. Clearly
large potential exists for solid fuel boilers in the domestic market, and
it is with this market in view that the lower end of tne size range was
chosen as the subject of this research program.
2. The research and development program

The technical work has been divided into a number of projects, and these
are shown schematically in Figure 1. Project numDer one is seen as the
core project inasmucn as the conceptual design is carried through to cold
(perspex) models and finally to a working boiler. The otner projects are
running simultaneously in oroer to obtain essential back-up information.
Some of the work is subcontracted to the Institute for Industrial
Research and Standards, Dublin.
2.1 The conceptual design

It can be shown that with a normal fluidising velocity and with flue gas
concentrations of about 20% excess air, the volume of fluidised bed
required to product 25kW (thermal) is approximately 2.5 litres
(2.5xl0"3m3). it is obvious that on this small scale with a
parallel-walled combustor, coalescence of bubbles would lead to excessive
elutriation losses. Thus it was decided to consider a combustor with
conical-vertical section so that slugging at high velocities would be
avoided. The cross-sectional area would increase up the combustor and
the fluidising bubbles would therefore be unable to bridge the combustor.
A further requirement is for the handling of large range of size of

coal. If processing costs are to be kept to a minimum, then coal of -lOinm
would seem to be an appropriate grade to attempt to burn. Tnus a means
of fluidising large particles in the presence of fine particles was
sought, and a circulating bed was seen as a way of doing this.
If a large turn-down ratio is to De achieved, the provision must be made
for slumping the bed. This is because control of other parameters would
provide a combined turn-down ratio of only 3 to 1, and a peak-to-minimum
output ratio of nearer to 50 to 1 is required for a boiler of this
nature. It is hoped to achieve this ratio by a combination of partial
bed slump and rapid on/off capability. Metnods of bringing apout partial
slump are usually based on partitioned bed and plenum chamber, but for
reasons of simplicity, and hence cost, other methods were considered
here. One such method was tne use of a oiapnrayfn annulus to control the
passage of air to the distributor plate. This is depicted in Figure 2.
However it was felt that this design would have a number of draw-backs:
a) The diaphragm would be a precision component and hence expensive, bj
the operating mechanism would need to be mechanical or electromechanical
and hence expensive, c) the design would require an expensive stand-pipe
distributor plate, and d) the diaphragm would be susceptible to clogging
by dust dropping from the bed.
18-
Tne reactor design which seems to show most promise in terms of
simplicity is one of asymmetric vertical section with a sloping
base-plate. In this way all of tne requirements discussed above may oe
met. The change from expanded to semi-slump modes of the bed may be
acnieved simply by variation in the output of the fluidising-air fan.
This is depicted in Figure 3. Because of the sloping base-plate, it now
becomes cheaper to have a combustor of rectangular cross-section rather
than a circular one. Cycling of the bed is ensured oy having a greater
pressure drop for air passing through the up-slope part of the base plate
compared with the oottom of the slope. This is achieved Dy a gradation
of the size of the holes in the plate. As the fluidising velocity is
reduced, the oed starts to slump as shown in Figure 3 D .
The adopted design features are discussed in more detail below.
2.1.1 Comoustor size

A thermal output of 25kw is planned. Assuming a combustion efficiency of
95%, then
25 x - ^ - = 26.3 kw
of thermal heat is required from the coal.
The calorific value of imported Dituminous coal is 24 HJ/Kg, tnerefore
26.3 x I03x 3600

Coal feed rate required
24 x 10
= 3.9 kg/hr.
Thus the bed must have a tnroughput of 4 kg/hour at maximum output.
The combustor size will depend on the amount of air flow required to
maintain combustion and to fluidise the bed. For a typical bituminous
coal containing 82% carbon and 6% hydrogen (ignoring the oxygen content
of the coal and the oxygen requirement for the coinoustion of sulpnur),
calculations show that 10kg of air is required for the combustion of 1kg
of coal. Therefore 40kg of air are required/hour for coinoustion at
maximum output. Allowing for up to 25* excess air, then this figure
becomes 50 ng/hr (38.7 M^/hr). The quantity of excess air depends upon
a numDer of factors and is never likely to exceed 25*; nowever, the
maximum possiole fan capacity must be allowed for.
As the air passes through the bed, its volume will increase due to
temperature rise according to the gas laws. At 800C this volume
becomes
38700 x (273 * 800) . ^ ^ ^
The selection of fluidising velocity will be dependent on the size of the

particles in the bed. In order to fluidise 10mm particles a velocity of
approximately 3 M/s would be anticipated. However, as discussed earlier,
only a fraction of the bed will be fully fluidised at any time. A mean
fluidising velocity of 2 M/s would seem appropriate. The relationship
between fluidising velocity Vf and bed area is
-19-
f *-
where V is the volume of air passing, 152 M^/nr = 4.22x10"^ to^/s,

and A is the bed area (M^). Hence
4
A= -222x 10
"2 = 2.11 xlO"2M2
A rectangular crosssection of combustor was chosen of dimensions 200mm x

100mm.
2.1.2 Combustor shape
An asymmetric combustor shape, as depicted in Figure 3 nas been adopted.
With the appropriate baseplate characteristics, tnis shape ensures
circulation of the bed and the capability of partiallyslumping the oed.
2.1.3 B aseplate characteristics
Tne first requirement is for a sloping baseplate positioned at an angle
such that the slump continues to slide aowh towaras the oottom where the
spouting occurs. In this way it is hoped that the stagnant time for
burning particles in tne slump will be too short tor clin<ering to
occur. Different designs of distributor place have been studied in
perspex models of comparable dimensions to tne comoustor. For oase plate
inclinations of less than 45 to the horizontal proDlems were
encountered of build up of slump even under conditions of high mean
fluidising velocity. Tests showed that an angle of nearer to 60 would
be most successful in keeping the bed continuously circulating. However
the best results were obtained with the distributor plate shaped as shown
in Figure 4 with a flat portion at the bottom end and no perforations at
all at the top end. A nearspout is generated Dy having 4mm holes,
protected only by gauze, at the bottom. The sloping section had a small
number of perforations of diminishing size up the slope (equivalent to
1mm diameter holes diminishing to 0.25mm). The nature of the
perforations, when eventually placed in a stainless steel base plate,
will ensure minimum fallback into the plenum chamber. The purpose of
the perforations is to provide a horizontal moment and nence aid in the
circulating of the bea. Tests are currently underway to observe the
effect of a baffle plate above the spout to aid the circulation. This
may make the perforations on the sloping part of tne Daseplate redundant.
One aim of the circulating bed is to prevent segregation of the large

from the small particles. Tests were carried out in a perspex model of
the effect of baseplate inclination on the segregation of large from
small particles and of light from dense particles. It was found that
with the preferred angle of the base plate of 60 to the horizontal, no
measurea segregation occurred of 8mm coal pieces from a bed of sand.
Similarly good mixing was observed between the sand Ded and low density
particles like wood chips and polystyrene beos. Unpublished work from
the National Coal B oard shows that in a conical bed, very little
segregation would be expected provided the included angle of tne cone
does not exceed certain limits.
It was therefore concluded that if good circulation could be achieved,
then segregation in tne bed would not be a problem.
20
2.1.4 Heat exchange
In order to Keep costs down, the Doiler will be equipped with cold walls
rather than tubular heat exchangers. Calculations have shown that witn
less than 25* excess air, then 50* of the heat produced must De removed
directly from the bea. For a heat transfer coefficient of 2 5 0 W / K / M 2
and a temperature drop from tne Ded to tne cooling water of 800K, tnen
200 kW/M2 will be transferred. Tnerefore, to remove 12.5kw,
'200 = 0.06 M^ of cold wall must contact tne Ded when the Ded
is fully expanded. Using the dimensions given aDove, 0.2 x 0.1 M, then
the height of the expanded bed to be in contact with the cold wall should
be
0.06 M 2 . , ...
I.e. 100mm approximately. As the fluidising velocity is decreased the

bed is expected to become progressively slumpeo and the boiler will have
a proportionately lower output. At a mean bed height of half the
maximum, the thermal output is expected to be only 10* of the maximum,
i.e. 2.5kW of which 1.5kw will be removed by the fluidising air. The
remaining lkw will not De transferred directly into the cooling water
because under these conditions the bed will be below the cold wall, see
Figure 5. Instead, losses by conduction and radiation are expected to De
sufficient together witn the tact tnat the control circuitry will De
designed to periodically activate tne bed. This will De to reduce tne
possibility of dinner formation.
Heat excnange in the freeboard will De aided by a consideraDly larger

contact area, and also Dy the tact that the flue gases will De channel leu
in such a way that maximum wall contact will De ensured. Calculations
are underway to determine the extent of tne Chanel ling required in
conjunction with studies of methods of avoiding elutriation (see bection
2.1.5).
2.1.5 Elutriation of fines
Tests have shown that elutriation of fines is reduced Dy circulation of
the bed. However the high fluidising velocity above the spout was found
to produce a high rate of elutriation. A baffle placed in the cold model
just above the expanded bed was found to improve the bed circulation rate
and to reduce elutriation - see Figure 6. This Figure also shows that a
grit arrestor is to De sited internally adjacent to a water wall such
that the arrested fines feed back into the downward arm of the cycle.
2.1.6 Start-up procedure
A great advantage of having a large turn-down ratio is tnat start-up can
be initiated whilst the bed is in a low output condition. It is
therefore anticipated that an electric start-up heater of little more
than one kWatt should be sufficient. Once the reduced bed is at
appropriate temperature, the air would be increased in order to
progressively bring more bed into tluidisation. Confirmation of this
procedure must await the construction of the prototype boiler.
2.2 Sensors and actuators
2.2.1 Air distribution
Consideration of an annulus means of controlling the air throughput (Fig.
2) gave way to the simpler and cheaper notion of a distributor plate
with varying sizes of holes. Test results are promising for a range of
base plate types. The quantity of air to be distributed will be
controlled by a variable speed fan. Tne use of an induced draught fan
rather than a forced draught fan was considered (and is still Deing
considered). The advantages of an I.D. fan are
a) it could readily be sited in a position divorced from the Doiler and
hence noise could be more readily controlled.
b) Any leaks or blockages in the boiler or hoppers coula present little
nuisance or danger, and
c) arising from b) it would be an advantage in allowing more readily for
a cheap vibratory means of controlling tne coal feed.
The I.D. fan, however, would nave a number of disadvantages
a) Exhaust gases would be at a higher temperature than amDient air. The
fan would need to have a 25* higher capacity assuming the exhaust did
not rise aoove 200C.
DJ Higher temperature operation would require a somewhat more expensive
fan in terms of materials and construction.
In view of the above arguments it was decided to opt for a F.D. fan for
the first prototype boiler and to reconsider tne problems associated with
opening up the hopper, the coal feed mechanism, etc. in due course.
2.2.2 Start-up heater
Electricity as a service to a F.B. boiler is a necessity, so it was
decided to use an electric start-up heater. The options open are a
fluidising-air heater sited at the outlet of the blower, or alternatively
to have an in-bed heater. Option 1 would tend to heat up the sloping
base-plate and the surrounding equipment, which is not desirable. Option
2 would require that the 'hot-finger' be quite small Dut still be aDle to
dissipate over lkW. Calculations have shown that for a fluidising
velocity of less than 20% of the maximum a heater like that depicted in
Figure 7 should meet the requirements.
2.2.3 Coal and ash handling

Vibratory coal feeders are considered to be the simplest, cheapest and
most reliable method available. Experience has been gained in their use
in this laboratory, and their reliability has been proven. However
vibrators must normally be isolated from other pieces of equipment
through flexible connections (see Fig. 8 ) . Furtnermore, unlike a rotary
valve or a screw feeder, a vibrator will not normally isolate the coal
feed hopper from the Doiler. Subsequently opening of the hopper could
only oe accomplished after the F.L). fan has been stopped. Pneumatic feed
was considered but calculations showed that for the feed rates required
here, this mode of coal feed was inappropriate. It is therefore proposed
that a screw feeder be fitted to the prototype boiler whilst an
appropriate mode of operation for a vibratory feeder is developed. An
additional draw back with the simple arrangement shown in Fig. 8 is that
background vibrations, manipulation of hopper, etc., could easily lead to
a small amount of coal feed when not required. This might have
undesirable consequences.
-22-
An asn removal mechanism from the base of the bed will be necessary in
oroer a) to maintain required bed depths, and b) to cope with tne proolem
of pieces of stone being fed into tne boiler and falling througn the
bed. (Most of the coal available on the Irish maricet is unwashed.) It
is considered that a simple vibrator, set on a rubber shock absorber,
will be adequate for this task.
2.2.4 Flue gas analysis

Considerable advantages would be gained from having an oxygen sensor in
the flue gas of the boiler; instant and precise information on tne state
of combustion would be available to the control circuitry at any time.
An 0^ sensor is not strictly essential: a fixed stoichiometric ratio
of air/fuel should give an approximately constant excess oxygen contttAt
and hence something approaching ideal combustion. However coal available
on the Irish market is very variable in quality and hence in combustion
rate. Thus one aim of this project is to look at the possibility of
developing a cheap reliable sensor for measuring flue gas oxygen
content. One of the commonest oxygen sensors developed for the kind of
application discussed here is based on a doped zirconia solid electrolyte
cell. This is depicted in Figure 9. The potential produced by the cell
is dependent oo a source of standard oxygen (air) and is also extremely
temperature dependent. Hence available sensors are hignly sophisticated
and therefore expensive. It would seem to oe a difficult task, except a)
a source of pressurised air is available from the fluidising fan, and b)
the temperature characteristics of the sensor output could be built into
the control circuitry, hence obviating the need for a heater in the
sensor. Tnese iueas are under study.
2.3 Control circuitry

A Westinghouse 1500 control computer has been adapted to control tne
performance of a 1MW Pilot scale FbC boiler. Tne appropriate control
loops have been determined, and the circuit design has commenced. Final
control parameters for the small-scale boiler are still to be decided.
More detail will be provided in the next report.
3. Future work
The program of research and development is up to the schedule as
described in Section 1. The next step is to construct the prototype
boiler. Simultaneously, work will commence on the other projects as
described in Fig. 1.
-23-
FIGURE I : Schematic representation of the projects involved in the production of the small-scale FBC boiler
z ^v YEAR
Carry out
Studies on
s h o r t and
production
long-term
cri teri a;
tests
material s
/ selection;
/ costi ngs;
/ B u i l d MK. market criteria
boi1er
V
Development of Bui 1d prototype Tests on cold Dynamics o f
control boiler. Tests models of semi-siumped
circuitry for on maintenance possible c i r c u l a t i n g bed
FBC boilers. of combustion designs of critical V
reactors. e f f e c t o f F'
Development of segregation,
Feas ibi 1 ity of Development of
mi croci rcui try
oxygen sensor
for flue-gas
actuators and
(onceptual desig
4*X e h r t r i-at-i-onrtfc
sensors using
envi ronment, eactor s i z e , S t u d i e s on
existing plant.
rapid in responsje liase-plate type, f 1uidisation
Studies on :
and cheap in a l 1-through and dynamics: effejcts
mechanisms of
manufacture ) l o c k i ng o f f 1 u i d i s ing
start-up and
:haracteristi cs, v e l o c i t y , bed
turn-down.
>tc. Development c i r c u l a t i o n rai e ,
j f mathematical r e a c t o r shape,
and c o l d models fcf base-plate
J slump/semi slump
nodel s w. r. t .
naintenance o f
characteristic!
etc.
combustion.
PROJECT 5
FIG- 2- FIG. U. The Distributor Plate
Annulor Control of Distributor Air
FIG 5 Heat Transfer through a Cold Wall
(a) Semi-slumped (b) Expanded
FIG. 3. The Asvmmptric Combustor

lo) Expanded Bed (b) Semi-slumped Bed
FIG, 6. An ArranqemQnt for the Freeb oard FIG. 8. Conventional Vib ratory Feeder
t Hopper
Vibrator
FIG- 9- Solid Electrolyte Oxygen Sensor
FIG. 7. "Hot-Finger" Start-up Heater Filter Ft Gauze ZrO?

^Control Thermocouple ~~ I
"!'
Kanthal Wire Ceramic Sheath - - 111
Standardising O2 1 ' .
_ I ' I I
TnT rr I I , I
wx \x v W W v o Heater .1 11
Ceramic Former^ f>_ m Controller
DISCUSSION
Question - Dr Syred, University College : The presenter has described design

features of a proposed small scale fluidised bed boiler. The over-
all concept appear to be sound, but relies on the integration of
a number of complex systems. Does the presenter feel that the
system can be developed to an economic solution, i.e. is the
cost of a zerconia oxygen probe justified and is the reliability
of such probes sufficient ?
Answer Dr Syred has identified the real problems associated with Oj

sensors. Unless such a sensor could be made cheaply and to
perform reliably we do not anticipate its inclusion in the
early prototype boilers. However, we are pursuing this de-
velopment project in the belief that the rewards will be well
worth the investment in research. The only other complexity
that we recognise in the design we have presented is that of
"integration", i.e. the control system. The actuators we
believe will emerge as being both cheap and reliable and the
control circuitry itself, being solid state (or even micro-
circuitry) should also meet that description.
Question - Dr Ehringer, CEC : Do you have experience in part load operation ?

Why did you choose a 25 kU fluidised bed combustor ?
Answer No we do not have experience in the operation of our boiler under

part load conditions; it is not yet built. However cold model
tests, as depicted in Figure 3 tell us that at low air through-
put a large proportion of the bed can be essentially slumped.
Under these conditions the heat will be proportional to the
stoichiometric air/fuel ratio at the reduced rates.
In many ways we chose the 25 kU scale because it posed a number
of intriguing problems. We believe that there is a considerable
market potential for small boilers between 25 and 200 kW.
Success with this project would clear the way for the whole
range of sizes. Furthermore, we suggest that if we can so
simplify boiler design in terms of the ancillaries, costs and
reliabilities of larger scale plant will be much improved.
When one considers a fluidised bed system the combustor is
simple and cheap to construct. The ancillary equipment is the
cost raising feature. We believe that if we can solve the
problems of reliability and cost for coal feeding, flue gas
cleaning and the control systems for our 25 kW model significant
economics in larger FBC boiler systems will follow.
-27-
Question - Mr Payne, National Coal Board : With regard to the use of
electrical ignition for starting up the 25 kW unit
1. Will the ignitor achieve a sufficient bed temperature to
allow acceptable combustion of the coal ?
2. Is an ignitor immersed within the bed likely to be
sufficiently robust ?
3. Is a 2 kW ignitor of sufficient size for application to a
25 kW unit ?
CRE have experienced these type of problems in applying
electrical ignition to fluidised beds.
Answer We welcome the benefit of Mr Payne's experience, and must admit

that as this stage we have very little experience of electrical
start-up. If the bed were to be started up on full output, it
would require over 6 kW of start-up energy.
However we anticipate being able to start-up in a semi-slumped
mode where heat losses from the bed will be low and the mass
of both air and bed will be much reduced. We do not anticipate
any problems with the robustness Of the start-up heater.
Question - Mr Robinson, Deborah Fluidized Combustion : In fluidised bed

combustion it is usual to work on a fixed air/fuel ratio during
design which is then optimised during commissioning.
Having established this ratio to say 1.2 to 1.25, 02 trim
results in only slight gains in efficiency. There is also a
danger that the reactivity of fuel during the combustion process
may change due to variation in the Fixed Carbon to Volatile
Matter content with consequential 02 variations in flue gas.
Why therefore do you think using O2 is a good parameter for
control ?
Answer For precisely the reason of fuel variability, we believe an

oxygen trim would be advisable in order to maintain maximum
combustion efficiency. The average boiler operator will not be
expected to appreciate the differences between one coal and
another, so the control circuitry must be able to cope with
these variations. It must be appreciated that to set fine
limits on the quality of the coal costs money, and in some
parts of Europe it is not possible to specify such limits.
For example, most of the coal available on the Irish market
is unwashed and comes from varying sources. Accordingly, this
coal will have varying combustion characteristics.
-28-
PRODUCTION OF HOT GAS FOR DRYING PROCESSES
USING FLUIDISED BED COMBUSTION OF COAL
Authors R.C. Payne, J. Highley, A.R. Windybank.
Contract number EED-2-322-CB
Duration 27 months 1 April 1981 - 30 June 1983
Head of project R.C. Payne, Head of Appliance Development Branch
Contractor National Coal Board, Coal Research Establishment
Address National Coal Board, Coal Research Establishment,

Stoke Orchard, Cheltenham, Gloucestershire, England.
Summary
1. Production of hot gas at temperatures up to 1250 C
An existing refractory lined, furnace test unit has been modified for
an experimental programme to investigate the production of gas at
temperatures up to 1250 C from fluidised bed firing of coal. The
furnace consists of two adjacent beds with a common wall and free-
board, one bed operating as a char combustor and the other a gasifier
with particle interchange between beds through apertures in the
partition wall. All modifications have been completed, additional
instrumentation installed and the unit operated for several short
term tests to establish a suitable wall configuration for control of
particle transfer.
Although steady state conditions have not yet been achieved, the
preliminary tests have indicated that the gasifier bed can be oper-
ated at sub stoichiometric conditions to produce a low calorific
value off-gas. This has been burned in the oxygen rich exhaust from
the adjacent char combustor bed to give gas temperatures approaching
1200C.
2. Production of clean hot air at temperatures of up to 600 C
A small scale test unit incorporating a ceramic tube heat exchanger

immersed in a coal fired fluidised bed with additional recuperators
above the bed to recover heat from the off gas, had been designed
and constructed in collaboration with Encomech Engineering Services
Ltd. This unit has been operated over a range of conditions to
gather preliminary heat transfer data and investigate the integrity
of ceramic heat exchange tubes in a fluidised bed environment. Initial
tests confirmed the design predictions on heat transfer performance
-29-
but the programme was interrupted by the repeated failure of the
mullite in-bed tubes. An investigation into tube temperature
gradients and possible alternative material was carried out and
several tube compositions were tested. Tubes manufactured from
silicon carbide have proved to be the most robust and have so far
completed over 1000 hours of operation including around forty stop/
start thermal cycles. The next stage has been to design and operate
a full size unit producing representative quantities of hot clean gas
for a process industry in order to demonstrate the thermal and
mechanical integrity of the concept. To assist the design a small
gas fired rig has been manufactured for testing of larger tubes.
This test rig has been installed and a programme of work has recently
commenced. The full scale design of 1.8 MW output rating has been
completed and installation of this demonstration unit at a maltings
is underway.
1. Introduction
Production of hot gas for drying processes represents a significant
use of energy in E.E.C. countries. There is a potential for substitution
of coal for oil or gas in many process industries if suitable technology
can be developed.
Fluidised bed combustion of coal is already being used commercially,
to produce hot gas at temperatures up to 950C for drying in agricultural
and industrial processes where some product contamination by ash is
acceptable. However, although the use of dust-laden gases at 950C from
these furnaces is commercially competitive with oil and gas because of the
lower fuel cost, the limitations preclude application to processes where
dust free hot gas is required or those designed to maximise thermal
efficiency by operation at inlet gas temperatures above 950C.
Therefore, the objective of this research programme is to develop
coal-fired fluidised bed technology in order to demonstrate the production
of:
(a) Gas at temperatures of up to 1250C, by enhancing above-bed combustion;
(b) Dust free hot gas at temperatures up to 600C, by (i) passing through
heat exchange tubes immersed in a combustion bed and (ii) operating
adjacent beds with heat exchange through the division wall.
(c) Gas with low dust content at temperatures up to 700C, by direct
contact with sand heated in a combustion bed.
The work programme will comprise three projects to evaluate the three
design concepts for variants of fluidised bed combustion to achieve the
objectives set out above. Two projects will make use of an existing NCB
test rig to be modified in stages to the design requirements of each
concept. The third proj_ect will initially utilise a purpose built ceramic
heat exchanger rig constructed in collaboration with Encomech Engineering
Services Ltd.
At present, experimental programmes have only been carried out on
phases (a) and (b(i)) above.
A2 Production of hot gas at temperatures up to 1250C

2.1 Pilot plant description
An existing pilot scale hot gas furnace has been modified to incorpor-
ate two adjacent fluidised beds, the first operating as a partial gasifier
producing a low C.V. gas which is burned in the oxygen-containing off-gas
from the second combustion bed. This second bed is fuelled by the char
-30-
from incomplete gasification in the first bed, and is operated with a high
excess air level as in conventional fluidised bed hot gas producing
furnaces.
2.2 Combustor/gasifier furnace

The layout of the pilot plant is shown schematically in Figure 1.
The unit containing the two fluidised beds is refractory lined and has an
external cross section of 1.1m x 1.4m and height of 2.4m. The air dist-
ribution system comprises multiple horizontal sparge pipes carrying
vertical standpipes. The air supply fan is rated at 2,500 m3/h at 750 ran
W.G. allowing a maximum heat release of 400 kW. The hot gases produced by
the furnace are cooled and cleaned to reduce dust emission to atmosphere.
The range of operating conditions are:
Fluidised velocity 1 to 4 m/s at 900C
Coal feed rate Up to 200 kg/h
Bed material Silica sand (range of size
distribution )
Bed depth (static) 150-300 mm
Bed temperature (nominal) 800-975C
The fuel is commercially available washed bituminous coal with about
35Z volatile matter and 6Z ash. Both singles grade (12 mm to 25 mm) and
smalls grade (12 mm to zero) coal will be used, supplied to the unit with-
out further crushing or drying.
Coal is fed to the gasifier bed by a variable speed screw feeder from
an intermediate hopper of capacity 250 kg.
2.3 Control of fluidised bed

In the conventional fluidised bed furnace virtually all of the heat
generated by combustion is utilised in heating the fluidising air and coal
to produce the hot gas. The furnace output can only be modulated by
changes in the bed temperature or fluidising velocity (air flowrate).
The fluidised bed temperature is maintained constant by on-off
operation of the coal feeder. The coal feed screw is controlled by a
P.I.D. controller which, by monitoring the fluidised bed temperature,
dictates the on/off time of the feeder. The on-off action of the coal
feed system leads to transient changes in combustion chamber pressure.
To maintain a constant pressure a pneumatically operated butterfly valve
has been installed in the hot gas ducting, up-stream of the induced draught
fan. A P.I.D. controller dictates the open/closed position of the butter-
fly valve, to maintain a preset pressure in the combustion chamber.
The fluidising velocity is controlled manually by regulating the air
flowrate from the forced draught fan to the distributor.
To facilitate the operation of the furnace as a two stage combustor/
gasifier it is necessary to incorporate into the air distributor a means of
regulating the air flowrate into any one of the six horizontal sparge pipes.
This has been effected using two manifold assemblies, manufactured from
310 stainless steel each carrying six tapered plungers as shown in Figure
II. The plungers give a leak proof seal when preloaded into the tapered
seats of the horizontal sparge tubes. The plungers can be retracted into
their manifold housings during the hot gas start-up period to prevent
possible thermal distortion of the exposed plunger shafts.
The rate of particle interchange between the two beds is controlled
by adjusting the geometry and arrangement of holes in the lower section of
the division wall. The upper section of the refractory wall consists of
ceramic panels withoverall dimensions 630 mm x 500 mm x 50 mm. The lower
section of the wall is constructed of soft refractory brick of external
dimensions 225 mm x 112 mm x 50 mm which can be easily worked to effect
geometry modifications Co the interchange port.
-31 -
Electronic analysis equipment has been installed to determine the CO,
C02,CH4,H2 and 02 content of the gases leaving the gasifier and combustor
beds and the furnace exit.
A.3 Test work'

3.1 Test conditions and procedure
Variation in the division wall geometry is one means of controlling
the transfer of char and exchange of bed material between the adjacent
fluidised beds, and is therefore a major control mechanism for achieving
stable substoichiometric conditions in the gasifier bed. For the first
series of tests three configurations of transfer port were evaluated of
four, five and seven vertical slots each of 12.5 mm width and the same
height as the static bed above the standpipe tops.
For each test the bed material used was silica sand sized between 1
to 2 mm and the fuel Gedling singles coal an 801 rank bituminous coal
sized 12 to 25 mm. Size and chemical analysis of bed material and coal
is given in Table 1.
The start-up procedure was the same for.all three test series and was
carried out by heating the fluidising air to the gasifier bed to about
750 C. When the bed temperature reached 600 C coal was added to further
increase bed temperature to 700 C, when the bed was operated on coal alone.
In each case the two beds are supplied with equal quantities of fluidising
air and gasifer bed maintained at 870-920 C until steady state conditions
were also achieved in the char combustor bed.
When both beds had attained steady state conditions, the action of
the P.I.D. controller was reversed such that coal would be fed in response
to a rise in temperature from the set point. At the maximum temperature
of the char combustor bed which was not allowed to rise above 1000 C sets
of test data were recorded as given in Table II.
A.4 Analysis of results and comments

Preliminary tests were completed at a nominal fluidising flow using
three variations of division wall to control char and sand transfer
between the adjacent beds. Under steady state conditions the final temp-
erature and carbon dioxide content of the char combustor bed is a measure
of the amount of coal/char combustion taking place. With four slots the
char bed temperature is low implying insufficient heat release would be.
available in this bed to satisfactorily control the gasifier at substoich-
iometric conditions. Seven slots produced higher over bed temperatures
and carbon dioxide levels together with visible flames as an indication of
a high rate of fuel transfer. With this configuration it would be
difficult to attain substoichiometric conditions in the gasifer bed with-
out causing excessive temperature rise in the char bed.
The five slot wall produced a residence time of coal particles in the
gasifier bed of sufficient duration to allow the complete devolatilisation
of the coal. The char transfer rate maintained the char bed at 530 C.
However for substoichiometric operation of the gasifier bed, in all cases
the char bed temperature approached 1000C. This is considered as being
too close to the coal ash fusion temperature, a safe working limit being
of the order of 950C.
A. 5 Conclusions
An existing test unit has been successfully converted into a twin bed
gasifier/char combustion rig and the furnace commissioned by carrying out
preliminary tests to determine datum test conditions and geometric config-
urations to allow transfer of char and bed material through the division
wall.
-32-
In these tests hot gas temperatures of the order of 1200 C were
recorded although steady state conditions were not achieved. The
gasifier bed off gas exhibited a gross CV of 4.4 MJ/m3.
The next stage of the experimental programme is to confirm
satisfactory stable operation of the gasifier and char beds over a range
of test conditions. This work requires changes in relative bed areas to
achieve satisfactory heat dissipation and possibly creation of velocity
gradients across the beds to modify the rates of material interchange.
Preliminary tests are currently underway and initial results suggest
that a bed area ratio of 1:2 (gasifier:char) may produce stable conditions.
B.2 Production of clean heated air at temperatures of up to 600 C

2.1 Pilot plant design requirements and general description
A 200 lcW pilot scale unit had been installed in order to carry out a
preliminary investigation of the mechanical integrity of horizontal tubes
installed in a fluidised bed combustor for air heating and also to provide
initial performance data. It was designed in collaboration with Encomech
Engineering Services Limited, a manufacturer of ceramic heat exchangers,
for use in dirty gas environments, and incorporates a patented sealing
arrangement to allow thermal expansion of the tubes.
To maintain high combustion efficiency from a coal fired fluidised bed
combustor the minimum excess air requirement has been found to be about
30Z. For a combustor to operate at this excess air level it is necessary
to remove 30-40Z of the heat input by the immersed surface within the bed.
For this 200 kW test unit, estimation of overall bed-air heat transfer
coefficient indicated an immersed surface requirement of 2.58 m^.
Consideration of available standard ceramic tube length to diameter ratios,
minimum tube spacing requirements to allow circulation of coal particles
and limiting pressure difference across tube end seals defines the inbed
heat exchanger configuration, fluidised bed depth and cross-sectional area.
The in-bed bank comprises twenty tubes of 54 mm external diameter with an
unsupported length of 800 mm. The bed area is 0.44 m^ and static depth
to immerse the tube bank is 400 mm.
To remove heat from the fluidised bed off gas, two further Encomech
recuperators of standard design are mounted above the bed. The cooled
exhaust gas is passed to a single cyclone and then to atmosphere.
A schematic arrangement of the unit is shown in Figure III.
The pilot scale unit has overall dimensions 1 m x 2 m x 3.8 m high
and consists of a fluidised bed module, a transition section, and two above
bed recuperator modules.
Fluidising air supplied from a single F.D. fan, rated for 550 m-Vh at
1100 mm W.C. enters the bed via a standpipe distributor. After combustion
in the bed, the off gases are cooled by the two recuperator modules.
Start-up is achieved by burning premixed natural gas in the bed.
Cooling air, supplied from a separate forced draught fan, rated for
2000 m^/h at 650 mm W.G., enters the building through two ducts. One
carries air to the single pass in-bed tube bank and thence to atmosphere
through a hot gas exhaust duct. The other duct carries air to the upper
recuperator tube bank. Air passes once through this, round a connecting
duct to the lower recuperator,once through this, and thence to atmosphere
via the same hot gas duct.
The fuel system consists of a coal hopper and variable speed screw
which enters the rig through the refractory wall above the fluidised bed.
Fuel for all tests was washed Snibston singles an 801 rank bituminous coal
similar in characterisation to the Cedling coal used for the gasifier tests
33-
Access to the fluidised bed is provided by a refractory lined port in
one side of the base module and the transition section includes a 100 mm
diameter viewing port.
Bed material is silica sand sized between 0.5 mm to 1 mm.
Flue gas samples are taken between the above bed recuperators and
analysed for O2, CO and CO2. Cyclone fines have also been collected for
analysis after each test run.
A second test unit has been constructed to evaluate ceramic tubes of a
length (approx 1.6m span) suitable for large scale production applications,
FigurelV. The rig can accommodate up to 6 x 54 mm external diameter tubes
or 4 x 77 mm external diameter tubes. The bed is heated solely by natural
gas burnt in the bed. The air for fluidising and combustion is fed into the
bed by a distributor comprising vertical standpipes arranged on a
100 mm square pitch. This long tube test rig has also been designed to
evaluate various configurations of tube/wall seals to cope with the
increased expansion which is experienced with the longer tubes.
B.3 Test results

The initial construction of the unit incorporated mullite tubes in
all heat exchangers.
Data obtained from seven steady state performance tests are summarised
in Table III. The objectives of the test programme were to determine the
effect of fluidising velocity, bed temperature and process air flow rate
on the output, combustion efficiency and heat transfer to the in-bed and
recuperator tubes.
It should be noted that only three different values were selected for
each variable, and no repeat tests were carried out. However, it is
considered that the data provides a good indication of the performance of
the unit. During these tests repeated failure of the mullite in-bed
tubes occurred which necessitated shut down and replacement. In addition
undetected tube failure during a test produced erroneous results, particu-
larly the excess air values. It was believed that the failures were
caused by high thermal stresses in the tube walls, particularly at their
interface with the refractory wall blocks on the relatively cool inlet air
end. As a result an investigation was undertaken into the tube wall
temperatures during operation using thermocoupled steel tubes and temper-
ature sensitive paints. In order to reduce the temperature gradient
producing these high internal stresses, inlet shrouds were developed
which although effective in lowering temperature differentials, did not
prevent tube failures. Temperature measurements made with and without in-
let shrouds are given in Table IV(a).
In parallel to this work several alternative tube materials were
evaluated, all of which except for silicon carbide exhibited similar
failures. The specifications of these materials are given in Table IV(b).
As a result a full set of silicon carbide tubes were fitted to the in-bed
heat exchanger and similar performance tests repeated. These results are
given in Tables Va + V b. The rig has been operated for over 1000 hrs. on
an endurance test which included over 40 stop/start thermal cycles without
evidence of failure. The tubes are due to be removed in the near future
for a full examination. During the tests no failure of the above-bed
mullite tubes has occurred in normal operation.
Initial tests on the long tube test unit have confirmed heat transfer
and pressure drop data for the design of the industrial prototype
demonstrator.
34-
B.4 Analysis of results and comments
4.1 Heat output from inbed tubes
The heat outputs for silicon carbide tubes were significantly greater
than for mullite tubes. This was thought to be a combination of the
higher air velocities in the silicon carbide tubes (because of the greater
wall thickness) and the higher thermal conductivity of silicon carbide. A
comparison is shown in Figures V(a) and V(b).
It can be seen from the results that increasing the bed temperature
and air flowrate through the tubes significantly increases the heat output
from the inbed tubes.
Increasing the bed temperature from 850 to 950 C increased the heat
extracted by the inbed tubes by approximately 14Z.
Although the results show that it is desirable from a heat transfer
point of view to maximise the air flow rate through the inbed tubes, the
resultant high pressure drop places a restriction on this. The pressure
drop through the inbed tubes on the pilot scale rig increases from 165
mm w.G. for a flow of 1000 nr/hr to 312 mm W.G. for a flow of 1500 m 3 /hr
(bed temperature 950C). This problem will be compounded on a full scale
unit with inbed tubes of possibly twice the length of those on the pilot
rig. High pressure drops along the tubes mean high fan power requirements
and could lead to excessive pressure drops across the tube seals.
4.2 Heat output from above bed recuperators

The above bed recuperators are linked in series, airflow entering the
downstream unit first. Heat output increased with rises in bed temper
ature for all tests and also with higher inlet air flowrates, except for
an early test on mullite inbed tubes as shown in Table III. This effect
has not been observed in later tests (Tables Va + Vb).
4.3 Excess air level

The tests have established that the excess air level can be reduced
to 30Z or less with a bed temperature of 950 C and a fluidising velocity
of 1.0 m/s. In order to achieve a higher thermal efficiency, this
excess air level will need to be maintained for higher outputs.
4.4 Combustion efficiency

Combustion efficiencies ranged from 96 to 99%. None of the variables
investigated had any significant effect on combustion efficiency, within
the ranges selected.
4.5 Thermal efficiency

Test results (Tables III, Va + Vb) show flue gas temperatures ranging
from 355C to 627C representing sensible heat losses of between 22Z and
37Z. However, these could easily be reduced by incorporating an addit
ional recuperator section into the unit.
Because of the units size, and the method of construction, the heat
losses from the shell are high, possibly up to 20Z of the heat input.
Shell surface temperatures of 150250C have been recorded on areas of the
unit lined only with high strength, high thermal conductivity refractory.
The results confirmed that thermal efficiencies of the order of 80Z
should be attainable on a production size unit.
B.5 Industrial prototype demonstration unit

The 1.8 MW prototype unit comprises a fluidised bed section incorpor
ating a ceramic heat exchanger above which is mounted the first recuperator
to recover heat from the offgas. A second, two pass recuperator is
35
mounted alongside and the off-gas enters the top, passes down through the
tube bank and then up the other side before being exhausted to the cyclones
and stack. An arrangement drawing is shown in Figure VI and basic
specification shown in Table VI. The unit will be sited at a commercial
maltings in Yorkshire, where it will replace the existing direct oil-fired
system.
5.1 Air supply
The existing malt kiln already employs a heat recovery system by
exhausting the saturated gas from the malt bed through glass tube
recuperators which provide about 20C of preheat to the incoming air
upstream of the oil burners. To ensure that the efficiency of this system
is maintained the four fans for the Encomech unit (one for each heat
exchanger and the fluidising air) will take their inlet from the plenum
chamber below the recuperators. Inlet cone (B.S.726) flow measuring
devices will be fitted inside the plenum on each fan inlet duct. The
fans will be mounted below in an acoustically lined enclosure.
5.2 Coal feed

For a maltings site it was essential that the fuel system be
totally enclosed to ensure cleanliness in operation. Coal storage will
be in a plastic silo from which the fuel, Allerton Bywater singles will be
conveyed by an elevated screw to a lock hopper and thence to the bed via
a modulating rotary valve. Action of the screw and lock (slide) valve
will be controlled by high and low level capacitance probes in the lock
hopper.
5.3 Ash removal

A bed regrading/replenishing system will be fitted and ash discharged
from the twin cyclones collected in a sealed skip.
5.4 Start up system

Fluidising air will be heated by an in-duct oil burner upstream of
the plenum chamber. The burner will be controlled by a high temperature
limit sensor in addition to the usual flame failure interlocks.
5.5 Control system

Basic design is for manual control by remote adjustment of motorised
dampers on each of the fan flows, with a P.I.D. controller maintaining bed
set point temperature by modulating the coal feed rate.
The malting process control system is also currently manual operation,
involving about 12 hours on full load followed by a staged turndown (4:1)
of the oil burners as an increasing proportion of recycle gas is used.
The existing oil burners are automatically modulated to maintain present
kiln inlet temperatures but recycle volume flow is controlled by a manual
damper. Fully automatic control of the plant using a micro-processor based
system is planned for the future.
B.6 Conclusions
A small scale test rig to examine and evaluate in-bed heat exchanger
tube materials and provide heat transfer data has been operated over a
range of conditions in an initial test programme.
Originally, in-bed tube integrity was disappointing but it has been
established that the most likely cause of failure was the steep temperature
gradients at the interface of the tubes and their supporting refractory
wall blocks. These gradients are particulary severe at the cold air
inlet end, and modifications to the tube retaining collars which reduce
-36-
the thermal effect, did not prevent failure of the tubes. Tubes made of
stainless steel have performed well but may give rise to sealing problems
on a larger scale unit owing to the relatively high coefficient of
expansion in comparison to ceramics. The steel tubes have proved useful
in obtaining information on transient tube temperatures from thermocoupled
positions in conjunction with temperature sensitive paint results.
Several alternative ceramics have been assessed and silicon carbide
identified as a most promising tube material, with over 1000 hours and
forty stop/start cycles having been completed without failure.
The original performance predictions have been validated and this
confirms the viability of a full size furnace design.
A representative initial rating for an industrial process prototype
is of the order of 1.8 MW and a suitable location is available for test
of such a unit. Design of this unit is complete and installation is at
an advanced stage.
37-
TABLE 1
Size analysis
% retained Samp Le
on'screen (mm) Gedling W.S. coal 1-2 mm Si sand
+ 26.5 2.9 0
+19 - 26.5 44.4 0
+13.2 - 19 46.4 0
+6.3 - 13.2 5.6 0
+2.36 - 6.3 0
+2 - 2.36 1.2
+1.7 - 2.0 21.8
- 0.7% < 6.3 mm
+1.4 - 1.7 37.4
+ 1.18 - 1.4 33.3
+1 - 1.18 5.3
1.0%<1 mm
Bed material analysis
Chemical
Silica sand
analysis
Na20 0.26
K2O 0.1
Fe 2 0 3 0.12
A1 2 0 3 0.5
SiO? 98.5
Sintering
1500C
temp
Coal analysis
Analysis % (as fired)
Total moisture 12.0

Ash 4.5
Mineral matter 5.4
Carbon 67.5
Hydrogen 4.46
Oxygen 8.01
' Nitrogen 1.45
Sulphur (total) 1.35
Chlorine 0.35
Carbon dioxide 0.15
Volatile matter 40.0

F. carbon 51.8
Calorific value 27772 kJ/kg
Ash fusion
Def. temp. 1170
Hemi temp. 1190
Flow temp. 1290
38-
TABLE II
OPERATING PARAMETERS
GASIFIER/CHAR COMBUSTOR
Test condition Steady state Substoichiometrie

Test number 1 2 3 1 2 3
Coal type
. ._i - _
Ccdl
Bed material 2 mm i
Tile configuration
(no. of slots)
4 5 7 4 5 7
Static bed depth 150 150 150 150 150 150
Fluidised bed temp.C
(i) Gasifier 920 920 896 1080 1060 1060
(ii) Char combustor 530 580 620 900 980 920
Cas temperatures C
(i) Freeboard of
850 990 920 980 1080 980
gasifier
(ii)Freeboard of
480 530 560 780 880 700
char combustor
(iii) Mixing region 580 790 680 1080 1170 980
Air mass flow rate
kg/s
(i) To gasifier 0.14 0.15 0.16 0.14 0.15 0.16
(ii) To char
combustor 0.14 0.15 0.16 0.14 0.15 0.16
(iii) Total mass
flowrate 0.28 0.30 0.32 0.28 0.30 0.32
Gas composition
(% volume dry)
Gasifier
X co 2 8.8 11-15 7.0 16.6 16 15
3! C O Trace 0.1-2 0 6.0 5.8 4.6
X CH4 0 Trace-1 0 8.5 3.0 3.6
X H2 0 Trace 0 4.0 4.8 2.6
x o2 11.5 2.5-7.0 13.0 Trace Trace 2.0
Char combustor
X C02 1 4.0 6.8 15 16.5 14
X CO 0 0 0 Trace Trace 0
X CH4 0 0 0 0 0 0
X H2 0 0 0 0 0 0
x o2 19 16.0 12.5 5.0 4.0 6.2
Calorific value
mJ m~ s (dry basis) - - - 4.41 2.4 2.2
39-
TABLE III CLEAN HOT GAS PRODUCER
SUMMARY OF PERFORIIANCE DATA WITH MULLITE TUBES

1 2 3 4 5 6 7
BED
Bed temperature C 800 850 950 950 950 950 950
Fluidising velocity ms~l 0.8 0.8 0.8 0.65 0.8 0.8 0.95
at bed temperature
Fluidising air flow m 3 hr -1 342 333 307 246 297 300 365
at 20C
COAL FEED
Coal feed rate kghr-1 29.6 29.9 31.6 34.4 25.4 33.9 34.3
Heat input kW 226 225 237 258 191 267 270
IN-BED TUBES
Air flow at 20C m 3 hr _1 1000 1000 1000 1000 473 1295 1000
Inlet temperature oc 16 20 18 24 24 27 29
Outlet temperature C 175 187 215 224 262 199 219
Heat extracted kW 55.7 58.4 69.5 71.4 40.3 78.9 67.1
RECUPERATOR TUBES
Air flow at 20C m 3 hr _1 988 1000 1000 992 1306 480 980
Inlet temperature C 14 18 16 22 21 26 27
Outlet temperature C 191 213 237 243 173 349 262
Heat extracted kW 61.1 68.8 78.5 76.6 69.6 57.1 84.0
FLUE GAS
Quantity kghr-1 475 472 443 372 425 436 515
Temperature C 355 371 411 436 363 561 472
O2 content % 9.3 7.9 4.9 5.2 7.9 4.4 7.6
CO2 content % 10.0 11.3 13.0 13.6 11.7 14.7 11.8
CO content % 0.0 0.2 0.1 0.4 0.0 - 0.1
Heat loss kW 60.7 62.5 66.5 60.8 54.1 88.8 85.9
Combustion efficiency % 96 96 97 97 98 98 98
Excess air level % 58 56 36 20 64 18 42
Thermal efficiency 52 57 62 57 58 51 56
1
1
-40-
TABLE IV(a)
CLEAN HOT GAS PRODUCER
TEMPERATURE DATA FROM STAINLESS STEEL TUBE
Distance With 15 mm Top Hat Shroud With 100 mm Top Hat Shroud
from end
Temperature Temperature gradient Temperature Temperature gradient
t tube
J J
C/cm C/cm C/cm C/cm
Inlet
end
20 145 360
52.5
60 A 30 51 570 36
29
160 860 860
Outlet
end
20
60
550
705 } Ikl9.5
39
25
590
730
7 22
160 900 900 T
TABLE IV(b)
PROPERTIES OF TUBE MATERIALS
Modulus Coefficient Therm,

of conduct. Analysis
of
rupture thermal at
(MN/m2) expansion 1100C AL 2 0 3 Si 0 2 SiC
(X10-6 C) (W/mC)
I.A.P. 72.5 5.0 1.99 55.3 40.0 -
Mullite 86.0 5.5 2.56 58.6 36.8 -
Purox 165.0 8.4 5.00 99.7 0.05 -
Silicon
26.0 4.5 13.00 5.0 9.0 84.0
Carbide
Cr Ni
S/Steel 19.0 2 26.0 3
24.0/26.0 20.0/22.0 -
1 Balance consists of oxides of F e , T i , Ca, M g , Na & K

2 20 - 900C
3 At 800C
-41 -
TABLE Va CLEAN HOT GAS PRODUCER
SUMMARY OF PERFORMANCE DATA WITH SILICON CARBIDE TUBES
8 9 10 11 12
BED
Bed temperature C 950 850 900 950 950

Fluidising velocity
ms~l 1.0 1.0 1.0 1.0 1.0
at bed temperature
Fluidising air flow
at 20C m3hr"l 374 410 392 380 379
COAL FEED
Coal feed rate kghr" 1 38.0 34.0 35.3 35.4 32.2
Heat input kW 279.9 250.5 264.9 262.1 238.3
IN-BED TUBES
Air flow at 20C m3hr_1 990 993 995 745 517
Inlet temperature C 34 31 27 29 30
Outlet temperature C 269 239 248 282 309
Heat extracted kW 83.6 73.7 79.1 68.2 52.5
RECUPERATOR TUBES
Air flow at 20C m3hr_1 985 985 995 1230 1255
Inlet temperature C 33 30 25 27 27
Outlet temperature C 274 234 245 226 223
Heat extracted kW 86.1 71.7 78.7 87.4 87.9
FLUE GAS
Quantity kghr" 1 529 569 549 535 529
Temperature C 476 422 438 424 420
O2 content % 6.5 8.4 . 6.8 7.0 9.6
CO2 content % 13.3 11.2 12.7 12.7 10.3
Heat loss kW 89.2 83.1 84.8 88.1 77.5
Combustion efficiency % 98.6 98.1 98.5 99.0 98.8
Excess air level /a 41.0 73.0 59.2 54.0 68.6
Thermal efficiency % 60.6 58.0 59.6 59.4 58.9
-42-
TABLE Vb CLEAN HOT GAS PRODUCER
SUMMARY OF PERFORMANCE DATA l/ITH SILICON CARBIDE TUBES
13 13b 14 15
BED
Bed temperature C 950 950 950 950

Fluidising velocity
ins *-
1.0 1.0 1.0 1.0
at bed temperature
Fluidising air flow
m3hr-1 374 371 377 385
at 20C
COAL FEED
Coal feed rate kghr -1 41.7 42.1 45.1 46.3
Heat input kW 308.5 311.5 333.7 339.0
IN-BED TUBES
Air flow at 20C m3 hnr -1
m 1250 1255 1495 1560
Inlet temperature C 29 28 32 30
Outlet temperature C 247 248 234 235
Heat extracted kW 98.1 99.8 107.8 114.1
RECUPERATOR TUBES
Air flow at 20C m3hr"l 750 745 515 567
Inlet temperature C 28 27 32 29
Outlet temperature C 318 319 371 365
Heat extracted kW 79.8 79.8 64.8 70.7
I FLUE GAS
Quantity kghr" 533 530 540 552
Temperature C 545 548 625 627
O2 content X 4.4 - 3.2 3.9
CO2 content X 14.8 - 15.7 15.2
Heat loss kW 105.7 106.3 123.4 126.8
Combustion efficiency X 98.1 97.5 98.1 98.5
Excess air level X 28.5 26.4 19.8 19.3
Thermal efficiency X 57.7 57.7 51.7 54.5
-43-
TABLE VI
INDUSTRIAL PROTOTYPE UNIT - DESIGN SPECIFICATION
Heat input 2.2 MW (75 therm /hr)
Heat output 1.8 MW (61 therm/hr)
Coal feed rate 260 kg/hr
Bed dimensions 1.6 x 1.3 m
Bed depth 0.6 m (fluidised)
Fluidising velocity (max) 1.6 m/s at 950C
In bed tube bank:
45 tubes 1.83 m long 0.077 m O.D.
Process air flow: 11000 m 3 /hr
Anticipated heat output: 840 kW
Outlet air temp: 250C
Above bed tube bank:
48 tubes 1.83 m long 0.054 m O.D.
Process air flow: 5000 m-Vhr
After bed tube bank:
100 tubes 1.83 m long 0.054 m O.D.
Process air flow: 20000 m^/hr
Total process air flow: 36000 m /hr
Mean outlet air temp: 170 C

"Hot Gas
2nd Stage "1250C
Combustion
Low
LOW C.V.|Oxygen
L.V.il RLC!
Coal
lOff Gas
"lOff loff Gas
Gas l(
Sand ;- fjar..
Bed
Air Y|:M:^;|-.;ri-riiri:l_Air
FIG. I: Production of hot gas at temperatures up to 1250C
1 Secondary Air Supply
80mm Dia Air Tubes

/ ! <
ST ft
t*
fe
iz= 1
Common
Manifold
Screw Connected Standpipes Manifold

/Assembly
Primary A i r
\ Supply
1^>*I
S 7Z^\
Taper Seat
Sealing Gland
Plungers
FIG. II: Modified air distributor - Gasifier char combustor
-45-
FIG. Ill: Production of clean heated air at temperatures up to 600C
(Encomech Engineering Services Ltd - Pilot Rig.)
FIG. IV: Long tube test unit

lOoo .>> O2oC
toe as.
l r i "** A*^RI^3M/ ^ k* >3 *ttc
FIG. Va: Effect of bed temperature FIG. Vb: Effect of airflow on

on heat extracted (bed tubes) heat extracted (bed
tubes)
3 SURGE HOPPER, 4.FEED VALVE. 5 * FLUID BED, 7 . ABOVE BED HEAT

EXCHANGER, 9 AFTER BED HEAT EXC HANGER, K> CYCLONES, HC HIMNEY
17. DUMP STAC K
FIG. VI: Layout of 1.8 MW heater

DISCUSSION
Question - Mr Bannard, NIHE : I wonder if you have analysed the failure

problems of your mullite or AI2O3 tubes ? Are these failures
thermomechanical ? In the context of the start-up heater
described in our report, would you recommend us to look
directly at the more expensive silicon carbide ? What are
the relative costs of the tube materials ?
Answer We believe the failures of mullite and AI2O3 to result from

their inability to withstand the high thermal stress existing
at the interface of the fluidised bed and the tube supporting
wall block. Details of the thermal gradients'encountered are
given in the paper. We could not advise on the suitability
or otherwise of either mullite or AI2O3 for your application
but would be wary of SiC in view of its electrical conductivity.
The cost of the materials tested were all of the same order
with the exception of AI2O3 which was three times more expensive.
Question - Dr Ehringer, CEC : About the SiC heat exchanger, did you study
the possibility to use Zr02-
Answer We did not consider use of ZrC>2 as a heat exchanger tube

material on the grounds of cost and availability.
Question - Mr Pomeroy, NIHE : 1. Would it be possible or perhaps more

benefical to fuel the combustor of the hot gas generator using
the elutriation of char particles from the gasifier into the
combustor ?
2. What tube temperatures would occur in the ceramic economiser ?
Would it be possible to use metallic materials for the lower
temperature tubes ?
3. Have you considered the degradation of SiC over long periods
e.g. 10,000 - 20,000 hrs ? It may be of benefit to consider
other Si based materials for the tubes or coatings e.g. Si3N^
or sialons.
Answer 1. Under the test conditions reported, using singles grade

(25-13 mm) fuel the char size is relatively large after
devolatilisation (typically 12-8 mm) and would therefore not
be elutriated. If a smaller size of fuel were used, e.g.
smalls (13-0 mm) then the elutriated solids would devolatilise
in the freeboard region and burn out in the hot gas ducting,
which affords a relatively long residence time at elevated
temperature.
2. No direct temperature measurements have been taken in the

above-bed recuperators, but a wide range would be expected
dependent on tube position, and process heat load. Use of
metallic tubes is feasible in these circumstances but conditions
leading to their corrosion/thermal distortion would become an
additional consideration.
-48-
3. One of the purposes of the prototype industrial unit is to
evaluate long term operation of the in bed tubes. If necessary
use of such protective coatings will be considered.
Question Mr Viddosich, Fiat : In ceramic tube heat exchangers what are

the limits, due to the end sealings, to the possibility to
transfer heat to air under pressure (e.g. 36 ATM and 4090 P S D ?
Answer Our existing experience is only up to differential pressures

around 2 PSI (1,6 x 10^Pa). However, in the future we would
intend to develop tube/wall seals to cope with substantially
higher pressures.
Question Mr Festa, Fiat : I would like to have more details of the

distributor of the "combustor and gasifier" and to know
whether you had any backflow problems.
Answer The air distributor comprises six horizontal sparge pipes

with vertical standpipes arranged on a SO mm triangular pitch.
Each standpipe has five 5 mm diameter drilled holes equispaced
around the circumference at levels of 9 mm and 26 mm below
the blanked top of the pipe. Material for all components is
type 310 stainless steel.
Air is fed into the sparge pipes from two cylindrical header
boxes, one at each end of the pipes. The flow of air into
each sparge pipe can be regulated by a taper seated valve
acting on its inlet. At an air flow rate of 1210 m3/h, the
pressure drop across the distributor is 400 mm wg.
There have been no problems with either back flow of air or
bed material into the distributor.
Question Mr B urrows, Stone Fluidfire : I would like to draw your attention

to a caution about the suggested use of silicon nitride under
combustion conditions. The use of this material as superheater
spacers showed very rapid attack by trace elements in ash.
Answer Although we do not plan to use silicon nitride in the immediate

future, advice on potential problems encountered with this
material in similar environments is useful information.
Question Mr Carrasse, Expert : Because of the low heat transfer coeffi

cients on the air side, the SiC tube is at high temperatures.
Why do you not simultaneously benefit from the possibility of
heat exchange by radiation from the tube interiors by using
analogous methods which are applied at high temperature in the
steel industry ?
Answer We have investigated improvement of air side heat transfer

by fitment of inserts within the tubes, which by absorbing
radiation from the tube walls enable substantial gains in heat
output to be achieved without uncurring an excessive pressure
drop penalty.
49
FURTHER EXPERIMENTAL DEVELOPMENT OF CIRCULATING FLUIDISED
BED COMBUSTION FOR APPLICATION IN STEAM BOILER FURNACES
USING LOW-GRADE FUELS AND COAL
Author H. Durrfeld
EC Contract No. EE-D-2-342-D
Duration 30 months 01.01.1981 - 30.06.1983
Project Leader Dr. H. Durrfeld
Contractor L. & C. Steinmiiller GmbH
Address Postfach 100855/100865
D-5270 Gummersbach 1
Summary:
The object of the research is the construction of a fluidised bed
pilot plant and experiments to obtain improved design data for steam
generators with circulating fluidised bed firing. The chief areas of
emphasis were a low degree of fuel treatment, good burnout, low
anission values for noxious pollutants and simple construction.
The fluidised bed boiler consists of a natural circulation evaporator
of tube-fin-tube construction with a superheater installed in the
second gas pass and an economiser in the third gas pass. The
combustion chamber is wider at the bottom and in this area the
fluidising velocities are low. Above this is a transitional area,
which is followed by a narrow upper sector in which the fluidising
velocities are high. The coal is fed in by worm conveyor and the
limestone added pneumatically. Ash drops out of the cyclone under the
influence of gravity and is removed pneumatically from the second
pass and fabric filter. Ash removal from the fluidised bed can be
accomplished via an ash cooler and thence from the fabric filter.
Erection of the plant is largely completed and commissioning has been
provisionally set for December 1982.
The experimental programme will involve the determination of heat
transfer in the boiler heating surfaces, the plant auxiliary power
requirement, the efficiency, load alteration response rate, behaviour
in respect of erosion and corrosion and the emissions produced. After
variation of the fuel and additive types and operational tests with
the most suitable parameters, the results are evaluated and the
research programme is completed by a comparative analysis of the
technical and economic factors thus revealed.
-50-
1. Introduction
This research programme is a further development of the circulating
"fast" fluidised bed process utilised in steam boilers. The main emphasis
is laid upon a low degree of fuel treatment, good burnout, low production
of noxious emissions (NO , SO ) and a reduction of the investment costs of
large scale plants by small scale and simplified construction. Good burn-
out is ensured by the longer and more effective mixing in the bed; the
reduction in SO emissions and - in this connection - the lower Ca/S ratio
by the longer dwell period, finer granulation and higher turbulence; and
the low NO emissionfigure by the improved firing kinetics and the burning
pattern (no temperature peaks, low excess air, no local air excess,
secondary air).
The research programme includes the construction of a fluidised bed
pilot plant and experiments to optimise the design data for steam
generators with circulating fluidised bed firing.
2. Description of the Fluidised Bed Boiler System
The design of the plant is presented schematically in the flow
diagram (Fig. 1 ) .
2.1 Design Parameters
Thermal capacity of steam generator 1 MW
Cross-sectional area of fast fluidised bed 450 x 450 mm
Cross-sectional area of slow fluidised bed 450 x 1200 mm
Total height of fluidised bed area 5500 mm
Steaming capacity 1230 kg/h
Steam pressure
Max. allowable operating pressure 32 bar
Pressure at exit from superheater 20 bar
Temperature at exit from superheater 450 C
Feedwater temperature 105 C
Fuel flow requirement 147.5 kg/h
(Design calorific value Hu = 26,300 kJ/kg)
Combustion air flow 1692 kg/h
Additive quantity 30 kg/h
2.2 Plant Infrastructure
In view of the fact that the boiler is installed inside the boiler
house on the premises of Messrs. Steinmiiller, it is possible to take the
feedwater from the existing feedwater tank. The saturated steam, after
temperature and pressure reduction by addition of feedwater, is fed into
a steam saturator in the works steam supply. The flue gases can be directed
into the existing stack. When the other boilers in the boiler house are
off line, the flue gases must be mixed with excess air to preserve the
minimum stack exit gas velocity. This air is first heated by saturated
steam in order to hold the stack temperature above the dewpoint.
There is also a connection for the supply of natural gas for the
lighting-up burner.
Softened water is used for the cooling of test heating surfaces, the
-51-
ash cooler and the cooling jacket round the coal feed, for test coolers
and special coolers. It is then fed into the deaerator of the feedwater
tank.
2.3 Water/Steam Portion
The steam generator is designed as a natural circulation boiler with
gastight welded tube walls. The furnace, or boiler 1st pass, is operated
under pressure and the 2nd and 3rd passes by contrast maintained at
negative pressure by an ID fan. The boiler stands on four legs and is
supported in axial guides about halfway up.
Feedwater is taken from the feedwater tank and pressurised via a
14-stage centrifugal pump. After flowing through the control valve it
enters the economiser in the 3rd pass from above and flows through the
economiser coils (4 x 40 tube lengths) downwards. During its passage, the
water heats up from 105 C to 210 C and the flue gases are cooled from
360 C to 200 C. The feedwater then leaves the boiler pass and is led
to the drum.
The downcomer lies next to the left-hand side wall and supplies the
front wall and division wall via two branches. The water/steam mixture
from these strongly heated walls, which border on the fluidised bed, flows
into a separate header for each wall and finally through two nozzles back
into the drum. Since the steam/water mixture in the evaporator tubes is
less dense than the water column in the downcomers, natural circulation is
maintained.
The side wall tubes and the rear wall, which is constituted by a
proportion of the side wall tubes, are not supplied by the downcomer. The
side wall distributors are connected with the drum directly by the
separate, only slightly heated or alternatively insulated evaporator
tubes. (The insulation is achieved by the application of ramming mass to
these tubes.) Temperature compensation in these three walls is achieved
by the circumstance that the water in the insulated tubes flows downwards
and the water in the tubes which are partly swept by the flue gases tends
to flow upwards.
In the drum water and steam are separated. The drum internals are
designed so as to prevent entrainment of water droplets into the super-
heater.
The saturated steam leaves the drum on its upper side by two
connecting pipes, which are employed in the gas-swept portion of their
run as superheater sling tubes, and passes into the distributor situated
in the 2nd pass. From here, the steam flows through four parallel tube
coils, each containing 39 tube lengths, and is superheated from 217 C to
450 C, thereby cooling the flue gases from 850 C to 360 C.
The superheated steam flows into the header located on the downcomer
and initially vents via the start-up valve and a silencer to the
atmosphere, until the desired steam condition has been attained (20 bar,
450C). After this, the saturated steam constant pressure valve controls
blowdown into the steam saturator. In the latter, the addition of feed-
water maintains a constant level. The saturated steam leaving from the
top of the saturator is fed into the 11 bar works steam supply.
For start-up, saturated steam from the works steam supply can be fed
into the downcomer and into 4 tubes of each side wall.
Saturated steam is also used to heat the filter hopper, in order to
prevent temperature excursions below the dewpoint during start-up and
shutdown, and - as already mentioned - to heat up the air which is mixed
with the flue gas.
-52-
The nine test heating surfaces which are mounted flush with the
refractory on the right hand side wall of the furnace are supplied
serially in 3 rows of three with softened water, as are the tube-bundle
type ash cooler, the cooling jacket round the coal inlet, the probe
coolers and the sample cooler.
Should the feedwater pump cease working for any reason the fuel supply
is immediately interrupted and an emergency feedwater pump with an
independent auxiliary power supply comes into operation. The evaporating
boiler water is thus replaced.
2.4 Firing
The fuel is supplied in 2 cu. m. containers, which are lifted onto
the storage bunker by a travelling crane and whose contents are then let
into the bunker via a hand operated flap. An intermittently operated worm
conveyor then fills an intermediate bunker, from whence a continuous worm
feed meters the fuel into the furnace some 350mm above the nozzle plate.
The additive is also unloaded from 2 cu. m. containers into a storage
bunker. A rotary vane feeder continuously removes the additive from this
bunker to a discharge nozzle and a branch of the combustion air supply
transports the additive falling out of the discharge nozzle into the
combustion chamber on the same level as the entry point of the coal.
A rotary compressor supplies the total combustion air. This is divided
into different branches: one to the nozzle plate, one for the supply of
additive, another to transport the ash from the cyclone into the furnace,
to loosen this ash should it fail to move, to supply the lighting-up burner
and to transport the ash from the fabric filter and second gas pass back
into the furnace.
The windbox under the nozzle plate is connected via an expansion joint
with the distributor frame of the boiler. The nozzle plate contains 24
nozzles, each with a maximum of 16 jets 5 mm in diameter. None of these
jets is directed against the furnace walls. The ash removal pipe projects
100 mm above the surface of the nozzle plate, so that the latter is
always covered by a stationary layer.
The rotary vane wheel under the ash cooler ensures controlled removal
of the ash into 2 cu. m. containers at a rate dependent on the pressure
difference in the lower part of the fluidised bed. The combustion chamber
is designed with a widened lower section in which gas velocities are low,
a transitional region and a narrow upper section in which high gas veloci-
ties occur. This purpose of this configuration is to allow a longer dwell
time for the fuel particles on the one hand, and on the other, to maintain
a circulation with relatively good heat transfer characteristics over the
total height of the furnace.
The flue gas leaving the furnace proper carries a heavy burden of
entrained particles. It passes through a hot gas duct connected by an
expansion joint with the furnace and into a high-efficiency cyclone. The
particles extracted by the cyclone fall under gravitational influence, with
some assistance from transport air, back into the furnace and enter it
above the level of the coal inlet.
The flue gases then pass along a further hot gas duct with expansion
joint into the second pass of the boiler. At the point of deflection up-
wards into the third pass, ash drops out again, is removed by a rotary
vane-type transporter and then conveyed pneumatically, in company with ash
from the filter, through the secondary air inlet into the furnace once
again.
-53-
From the third pass the flue gases, now at a temperature of 200 C,
pass through a sheetmetal duct equipped with expansion joints into the
baghouse, which is fitted with felted glassfibre tubular filters. These
are cleaned by bursts of compressed air. The return of the ash from the
filter hopper.is accomplished by a rotary vane wheel working in conjunction
with the current of secondary air, and the ash is reintroduced into the
furnace through the secondary air inlet. Ash return is controlled as a
function of the pressure loss in the upper part of the fluidised bed. When
the filter hopper is completely full, a rotary vane wheel removes the ash
down to a minimum level in the hopper, and it is then conveyed pneumati
cally into 2 cu. m containers for removal.
The cleaned flue gases are extracted from the filter by the ID fan
and exhausted into the stack. An inlet guide vane control before the
ID fan allows the maintenance of atmospheric pressure upstream of the
cyclone. As already mentioned, the flue gases are mixed with preheated
air before entry to the stack, should the exit velocity be too low.
For startup of the fluidised bed firing, the fluidised bed is raised
to the ignition temperature of the fuel with a gasfuelled lightingup
burner before the fuel is admitted and the gas burner turned off.
2.5 Control Technology and Measuring System
A central process computer with builtin dialogue capacity was
selected for plant control. This equipment is capable of displaying real
time measured values and also presenting them graphically as curves for
the total operating period, displaying and printing out information and
alarms, and at the same time controlling all plant functions.
The main closedloop control circuits are:
the drum water level control
the fluidised bed temperature control
the superheated steam pressure control
the furnace temperature control and
the air ratio control
Additionally the following are controlled:
the additive/fuel ratio
the pressure differential in the fast sector of the fluidised bed
the pressure differential in the slow sector of the fluidised bed
the water level in the steam saturator and
the ash level in the cyclone ash return ducting.
The most important control circuits are also equipped with conventio
nal control. Some operating data are also logged using pen recorders.
Necessary interlocks are also built into the lowtension switchgear.
For emergency operation and also for greater clarity, a schematic
flow diagram with pushbutton operation of the control circuits as well as
signal lamps and instrumentation for important binary and analogue data
displays is provided.
For data logging and mathematical evaluation of the experimental data,
a second process interface with data logger, computer for evaluation
purposes and printer is envisaged.
All necessary measuring equipment for carrying out the experimental
programme is provided. A large number of nozzles is provided in the
boiler wall for investigation of the firing behaviour and the heattrans
fer in the test heating surfaces already mentioned, at various positions
within the fluidised bed. For working out the heat balance, etc. of the
system,the mass flows of material entering and leaving the boiler,
together with their condition, are also recorded.
3. Test Programme
The test programme first of all concentrates on cold functional
testing of all individual components, ascertaining the characteristics
of the metering systems such as worm feeds, rotary vane feeders, etc.
and checking all components for correct operational functioning.
Using a coal of known characteristics the plant is then tested as a
whole. The behaviour of the fluidised bed during start-up and shutdown
and its interactions with the steam generator are here of particular
interest. On the exterior of the only slightly heated furnace walls,for
example, thermocouples are welded in various positions to ascertain
temperature differences in relation to the temperature of the downcomer.
At the superheater outlet,temperature differences between the four banks
under all operating conditions are measured in the same way.
These important performance data are logged in the course of the
succeeding tests:
- The heat absorption of the boiler heating surfaces with a view to
determining the heat transfer coefficients by measuring the inlet and
outlet temperatures and mass flows across the previously mentioned
water-cooled test heating surfaces in the boiler.
- The auxiliary power requirement of the fluidised bed boiler, by
measuring the electrical power required for fuel worm chargers,
combustion air fan, ID fan, feedwater pump and the rotary vane feeders.
- Boiler efficiency, by determining burnout, the positive and negative
enthalpy balances in the various systems (air/flue gas, water/steam,
fuel/ash) and the losses in waste heat.
- Noxious emissions, by measuring the S0 2 , NO and dust content in the
flue gases.
- The rate of load change with varying fuel and air proportions.
- The rates of erosion and corrosion by the analysis of components made
from different materials.
During these tests boiler output, fuel type and additive type should
be varied and the data of material inputs and residues recorded. Finally
operational test runs with optimised groupings of parameters should be
undertaken.
The evaluation by the computer already referred to should be performed
both on-line and off-line. A consideration of results under their technical
and economic aspects and the concluding report will complete the research
programme.
4. Progress of work
The research project got under way at the beginning of 1981. The basic
engineering, which was based on the process design and the basic design,
was taken in hand first and the detail engineering largely completed by
the middle of 1982. Since July 1982 the plant has been in the process of
construction. The fluidised bed boiler, the steam saturator, the ash
cooler, the storage bunker and practically all the ducting and pipework
were manufactured by Steinmiiller. Motors, valves, the cyclone, the filter,
the electrical equipment and the instrumentation and controls were bought
in from suppliers. The complete erection, apart from insulating, painting
-55-
and instrumentation & controls, is being carried out by SteinmiJller.
The setting-up of the so-called "hot laboratory" in Steinmiiller's
boilerhouse was not financially supported by the European Community
commission. The steelwork, machine foundations, the infrastructure and
the arrangement of test measuring centres and control room were adapted
to the requirements of the fluidised bed boiler plant.
The hydrostatic test on the boiler took place on 28.09.82, the
lining of the boiler internal surfaces with refractory has just been
completed and the unit is at present being insulated externally. At the
same time the electrical and the instrumentation and control systems are
being installed. Finally the computerised control centre, which has
already been tested at the supplier's, will be connected.up. Hot
commissioning will probably take place in December 1982.
Preparations have already been made for commissioning and the test
programme. The containers, the fuel bunker and the additive bunker have
been checked to ascertain that material runs freely out of them, and the
discharge characteristics of the fuel worm feed and the rotary vane feed
for the additive have been ascertained.
A start has been made on compiling the evaluation programme for
calculating burnout, the degree of desulphurisation, boiler efficiency,
total system efficiency and the local heat transfer coefficients in the
test heating surfaces during on-line evaluation. At the moment, the
operating manual is being completed, the operating personnel trained and
preparations are being made to obtain official sanction for the use of
the hot laboratory and the prosecution of the experimental programme.
This last requires, among other things, the notification of the responsible
personnel and of the experimental programme to the Factory Inspectorate.
5. Conclusions
Since at the moment no test results are available, it is not possible
to say anything about the success or otherwise of the research programme.
We believe however, thanks to Steinmullers know-how in the areas of boiler
construction and firing systems, plus some investigations already
completed or running parallel with the planning & design phase, that the
prospects for uninterrupted progress in and usable results from the
research programme are distinctly favourable.
Annexe: pictures of the erection status of the boiler at the end of

September 1982.
-56-
Stack
ID-Fon
Saturated steam
into factory
Steam Suturotor
Water
Steam
Fuel
Ashond other
Air
Flue gas
Feedrfolcr Pump
FIG. 1 - Plant Flow Diagram

FIG. 2 - Boiler supports with distributor steelwork, nozzles for the media
entering the fluidised bed and for measuring instruments
FIG. 3 - Boiler supports with distributor steelwork, manhole, down-comer,

auxiliary steam connections, ash removal from second pass via
rotary vane transporter
-58-
oo
u
c
J3
U 3
o cr
a. oi
c oo
ra c
0) n
c .c
CD u
>
a
>. c
ra
u -a
o a>
U J3
O T3
O 1
U 3
n 6
o
o
13 in
c a
3 u-i
O
-59-
FIG. 6 - Fuel and additive bunkers from above
FIG. 7 - Containers for fuel, additive or ash

FIG. 8 - Fabric filter and ID fan with inlet guide vane control
FIG.9 - Baghouse with heating coils and rotary vane transporter for ash
removal and recirculation
M 0)
o o
X J3
o
M
En
60
C .C
00-H
h 3
>, C
nj to
> -I
E o
r-. T3
I c
C N
r
-62-
FIG. 12 - Steam drum and cyclone with connecting ducts to first and second
pass
KIC. 1J - Steam drum with manhole and boiler nameplate
-63-
r * * ' * - ' ^ "
FIG. 14 - Steam saturator with water level control valve and superheated
steam blowdown valve
FIG. 15 - Fabric filter from above, with steam drum and travelling crane
-64-
FIC. 16
FIG. 17 - Boilur drain valves

FIG. 18 - Encapsulated rotary compressor with overflow valve
FIG. 19 - 14-stage boiler feedwater pump
-66-
DISCUSSION
Question - Chairman : Can you give your view about the future application
of the circulation bed/ as well as the positive and the
negative point ?
Antwort Die zirkulierende Uirbelschichtfeuerung soil Anwendung finden

bei Industrien und Heizkraftuerken. Vorteile sind :
a) der hohe Ausbrand : uber 99 X;
b) die niedrige NO -Emission : kleiner als 200 ppm;
c) die verbesserte S02 - Feinbindung bei geringer Kalksteinzugabe :
etwa 90 X Feinbindung bei Ca/S von 1,5 bis 2 und
d) bis zu vierfachem Brennstoffdurchsatz je m' Brennraumquer-
schnitt gegenuber langsamen Wirbelschichten.
Der bei zirkulierenden Wirbelschichten erforderliche hohe
Mehraufwand fur Brenn- und Zuschlagstoffe wird durch die
(Combination von langsamer und schneller Uirberschicht bei der
Steinmiiller-Entwicklung uberf lussig.
Question - Mr Robinson, Deborah Fluidised Combustion : Your flow diagram
indicates that a great deal of attention is being given to
recycling elutriated particles. Presumably this is because
you are trying to achieve a high combustion efficiency by
carbon burn out.
Our experience shows that the majority of unburned carbon
collected down stream is fixed and of low SG in the form of
hollow hemispheres therefore reinjection usually does not
show a significant gain in carbon burn out as there is not
sufficient residence time. Would you care to comment ?
Antwort Bei der zirkulierenden Wirbelschicht sind die Ruckfuhrungen

erforderlich, da andernfalls die im Brennraum enthaltene Masse,
besonders bei ashearmen Brennstoffen, verloren ginge.
Zirkulierende Wirbelschichtgeuerungen anderer Hersteller zeigten,
dass AusbrSnde von uber 99 X erreichbar sind.
Question - Chairman to Mr Robinson : Is the blowing up of coal/carbon

particles a general phenomenon or is it related to certain
types of coal ?
Answer The majority of coal types exhibit this characteristic on

combustion in a fluidised bed at temperatures between 800 C
and 950 C. The SG varies between 0.2 and 0.4 and particle
sizes in ash are 60 to 20 Microns. As the superficial velocity
is high on reinjection the residence time is short, and
therefore not sufficient to increase burn out of the fixed
carbon. There is therefore little gain in overall efficiency.
Ash recycle also varies heat transfer coefficients.
Comment - Mr Pomeroy, NIHE: A comment to attempt to clear up the question

posed by Mr Robinson. I think Mr Robinson is referring to a
circulating fluidised bed whereas the presenter is referring
to a recirculating bed. In the latter system the particles
are entrained and returned to the combustor deliberately and
this gives very long residence times for carbon burn out.
-67-
THE BURNING OF RDF AND INDUSTRIAL WASTES IN
FLUIDIZED BED INDUSTRIAL BOILERS
Author: R BURROWS
Contract Number: EED-3-343-81-UK(H)
Duration:
Head of Project: R Burrows, Stone Fluidfire
Contractor: Stone Fluidfire
56 Second Avenue
Pensnett Trading Estate
Brierley Hill
West Midlands DY6 7PP
Summary
F o l l o w i n g p r e l i m i n a r y q u a l i t a t i v e t e s t s and t h e
r e s o 1 u t i o n of a f u e l f e e d i n g pr o b l e m w or k on t h e
c o m b u s t i o n of RDF has been s t a r t e d . I t has been shown
t h a t t h e s e l e c t i v e d i s tr i b u t i o n of s e c o n d ar y a i r over
t h e b e d may be d e s i r a b l e t o r e d u c e CO f o r m a t i o n w i t h o u t
i n c r e a s i n g e x c e s s a ir . S hro t d ur a t i o n t e s t s on two
o t h e r w a s t e m a t e r i a l s ar e r e p or t e d and i t i s concluded
t h a t a f l u i d i z e d b e d c o m b u s t i o n system i s c a p a b l e of
b u r n i n g a wid'e r a n g e of w a s t e m a t er i a l s pr o v i d e d
a t t e n t i o n i s p a i d t o t h e d e s i g n of f u e l f e e d er s , a ir
distribution etc.
1. Introduction
The o b j e c t i v e s of t h i s research are to obtain t e s t data
on t h e b u r n i n g of r e f u s e d e r i v e d f u e l (RDF) and o t h e r
i n d u s t r i a l and a g r i c u l t u r a l w a s t e s using a f l u i d i z e d bed
boiler.
T h e r e a r e many i n s t a n c e s where c o m b u s t i b l e w a s t e
m a t e r i a l s a r e c r e a t e d on s i t e s where t h e r e i s a l s o an
a p p r e c i a b l e h e a t load which might be met in p a r t , a t l e a s t ,
by s t e a m or h o t w a t e r from such a b o i l e r . Also, i t i s not
unusual to find s i g n i f i c a n t c o s t s a s s o c i a t e d with the
disposal of such wastes.
S u i t a b l y p r e p a r e d m u n i c i p a l refuse may not be used on
t h e s i t e where i t i s created but could r e a d i l y be used within
t h e m u n i c i p a l i t y i n c u r r i n g minimum t r a n s p o r t c o s t s and
replacing expensive imported f u e l s .
B e c a u s e RDF i s p o t e n t i a l l y the most widely a v a i l a b l e of
t h e w a s t e s l i s t e d in t h i s programme i t has been s e l e c t e d as
t h e f i r s t t o receive a t t e n t i o n . However, o p p o r t u n i t i e s have
a r i s e n t o c a r r y out short duration t e s t s burning other waste
m a t e r i a l s and notes on the r e s u l t s obtained are included for
general information.
Where a p p r o p r i a t e i t i s i n t e n d e d t o return to these
m a t e r i a l s t o obtain more comprehensive data a t a l a t e r stage
in the programme.
Soon a f t e r i s s u i n g our f i r s t r e p o r t on t h e t e s t
i n s t a l l a t i o n d i s c r i b i n g preliminary q u a l i t a t i v e combustion
t e s t s b u r n i n g RDF i n J a n u a r y t h i s y e a r , S t o n e P i a t t
I n d u s t r i e s w e n t i n t o r e c e i v e r s h i p and t h e work was
i n t e r r u p t e d f o r some m o n t h s p e n d i n g t h e outcome of
n e g o t i a t i o n s f o r t h e t r a n s f e r of our company to the newly
formed 'Stone I n t e r n a t i o n a l Company'. We are happy to report
t h a t ' S t o n e F l u i d f i r e ' i s now p a r t of a f i n a n c i a l l y sound
company w h o ' s p o l i c y i s t o d e v e l o p t h e 'Energy Systems'
b u s i n e s s and in p a r t i c u l a r fluidized bed combustion and waste
heat recovery systems.
Work on t h i s p r o j e c t was resumed r e c e n t l y and further
c o m b u s t i o n t e s t s have been c a r r i e d out burning RDF. We have
a l s o done some work on the combustion of a residue from the
food processing industry and waste from the t y r e reclaimation
process.
2. Installation
The t e s t b o i l e r used f o r t h i s work i s i l l u s t r a t e d in
F i g u r e 1. I t s r a t e d c a p a c i t y i s 2 5 0 , 0 0 0 Kcal/hr as low
pressure hot water.
The f l u i d i z e d bed i s of the " F l u i d f i r e " r e - c i r c u l a t i n g
d e s i g n , c i r c u l a t i o n of bed m a t e r i a l being brought about by
t h e p r o f i l e of the d i s t r i b u t o r t i l e s as i l l u s t r a t e d . To take
a d v a n t a g e of the c i r c u l a t i n g p a t t e r n s formed i t i s important
t h a t t h e f u e l be introduced in the c e n t r a l down flow region
so t h a t i n c o m i n g f u e l i s e n t r a i n e d with the bed m a t e r i a l
i m m e d i a t e l y and i s s u b j e c t e d to the maximum residence time
w i t h i n t h e bed before reaching the s u r f a c e . This feature i s
c o n s i d e r e d t o have an important bearing on the combustion of
v o l a t i l e s and low d e n s i t y fuel c o n s t i t u e n t s within the bed
which w i l l be i m p o r t a n t for the e f f i c i e n t combustion of RDF
and other waste m a t e r i a l .
Heat i n h o t w a t e r i s d i s s i p a t e d i n a forced draught
cooling tower. The w a t e r i s pumped round the system and
p r o v i s i o n i s made for measuring flow. The water temperature
r i s e can be m e a s u r e d s e p a r a t e l y a c r o s s the "in bed" tube
p a n e l s a n d t h e c o n v e c t i o n bank so t h a t t h e two h e a t
absorptions can be accounted s e p a r a t e l y .
B o i l e r f l u e gas passes f i r s t through a cyclone
s e p a r a t o r , t h e l a r g e r s o l i d p a r t i c l e s from t h i s being stored
f o r t h e duration of a t e s t period for subsequent weighing and
analysis. A f t e r t h e cyclone the gases pass to a bag f i l t e r
u n i t and h e r e again the dust i s stored for the duration of a
t e s t , t h e bags then being cleaned and the accumulated s o l i d s
w e i g h e d and a n a l y s e d f o r carbon content e t c . Multiple gas
s a m p l i n g p o i n t s a r e p r o v i d e d for flue gas a n a l y s i s during
tests.
The b o i l e r i s s t a r t e d up by burning propane gas to r a i s e
bed t e m p e r a t u r e t o i g n i t i o n temperature and t h i s gas supply
i s c a p a b l e of b r i n g i n g bed temperature up to approximately
80ffC. From t h i s p o i n t t h e t e s t f u e l may be i n t r o d u c e d
d i r e c t l y or a l t e r n a t i v e l y , coal f i r i n g can be used to r a i s e
t h e t e m p e r a t u r e f u r t h e r t o 9 0 <S C maximum if necessary when
b u r n i n g d i f f i c u l t f u e l s . Equipment i s a v a i l a b l e for weighing
t h e t e s t f u e l i n t r o d u c e d into the storage hopper during the
t e s t period.
The t e s t b o i l e r h a s been i n s t a l l e d i n our works a t
B r i e r l e y H i l l together with the propane gas s t a r t up system,
c o a l s t o r a g e c o n v e y i n g and f e e d i n g s y s t e m , w a s t e f u e l
h a n d l i n g s y s t e m , automatic c o n t r o l s and t e s t instrumentation
sensing and sampling p o i n t s .
3. Progress of Work and Results
A f t e r an i n i t i a l a t t e m p t t o f e e d raw RDF v i a t h e
e x i s t i n g c o a l f e e d e r p r o v e d unsuccessful a pneumatic feed
s y s t e m was d e v e l o p e d t o handle t h i s type of waste. I t was
u s e d i n i t i a l l y t o h a n d l e the waste nylon, rayon and rubber
crumb from t y r e r e c l a i m processes and these m a t e r i a l s were
s u c c e s s f u l l y burned in the b o i l e r . Unfortunately, these
r e c l a i m p r o c e s s e s were s h u t down because of the changing
e c o n o m i c s i t u a t i o n i n t h e r u b b e r i n d u s t r y before we had
obtained f u l l q u a n t i t a t i v e t e s t data on combustion.
We c a n , h o w e v e r , r e p o r t t h a t good smokeless combustion
was a c h i e v e d with low carbon in ash loss and the deposits on
t h e convection heating surface consisted of an e a s i l y removed
c o a t i n g of d u s t . In common with a l l 'high v o l a t i l e ' fuels
t h e r e was some combustion of v o l a t i l e s above the bed surface
as evidenced by luminous flames in the freeboard space.
When we were a b l e t o resume t h e work on RDF i t was
d e c i d e d t o u s e t h e p e l l e t i z e d form fed. via the coal feeder
f o r t h e i n i t i a l a s s e s s m e n t of combustion efficiency since
t h i s c o u l d be more rapidly obtained and s e t up for t e s t i n g .
The a n a l y s i s of t h i s m a t e r i a l i s very s i m i l a r to the raw RDF
b u t i t i s extruded into p e l l e t s which s i m p l i f i e s the storage
and handling problems.
-70-
I n i t i a l t e s t s showed t h a t w h i l s t the RDF burned f r e e l y
w i t h a smoke f r e e s t a c k , a proportion of the v o l a t i l e s burned
a b o v e the bed and even with r e l a t i v e l y high e x c e s s a i r l e v e l s
of 6 0 t o 70% we were g e t t i n g high carbon monoxide r e a d i n g s up
t o 0.5%. I t s e e m e d l i k e l y t h a t t h e v o l a t i l e s were being
r e l e a s e d u n e v e n l y o v e r t h e bed area with a higher
c o n c e n t r a t i o n l o c a l to the fuel i n j e c t i o n p o i n t . The
d i s t r i b u t i o n of oxygen in the freeboard space i s e f f e c t e d by
t h e primary a i r s e t t i n g s c o n t r o l l i n g d i s t r i b u t i o n between the
v a r i o u s d i s t r i b u t o r t i l e s and t h e s e s e t t i n g s are d i c t a t e d by
the requirements for c o r r e c t f l u i d i z a t i o n .
F r o m t h e s e r e s u l t s we concluded t h a t the i n t r o d u c t i o n of
s e c o n d a r y a i r over the bed, s e l e c t i v e l y d i s t r i b u t e d , o f f e r e d
t h e b e s t c h a n c e of r e d u c i n g c a r b o n m o n o x i d e and o v e r a l l
excess air l e v e l s . A simple secondary a i r i n j e c t i o n system
was f i t t e d t o d i s t r i b u t e a i r j u s t above the bed s u r f a c e on
t h e f r o n t t o back c e n t r e l i n e o f the b o i l e r , i . e . over the
fuel injection area.
E a r l y r e s u l t s u s i n g secondary a i r i n d i c a t e t h a t carbon
m o n o x i d e can be reduced below 100 ppm with 60% e x c e s s a i r and
f u r t h e r r e f i n e m e n t of t h e a i r d i s t r i b u t i o n should enable
t h e s e f i g u r e s t o be improved upon.
The u n b u r n e d c a r b o n i n a s h h a s not y e t been properly
d e t e r m i n e d b u t t h e a s h w e i g h t s u g g e s t s that t h i s w i l l be
s a t i s f a c t o r y and the ash i s a l i g h t grey c o l o u r .
D u r i n g t h e s e t e s t s the heat removed d i r e c t l y fromthe bed
v i a t h e ' i n bed' tubes was only 20% of the t o t a l heat t o hot
w a t e r as compared with 50% burning c o a l . This i s , of c o u r s e ,
due t o t h e combined e f f e c t s of high moisture in f u e l and the
combustion v o l a t i l e s above the bed.
The b o i l e r e f f i c i e n c y has not y e t been p r o p e r l y
d e t e r m i n e d s i n c e we are w a i t i n g for the a n a l y s i s of f u e l and
a s h s a m p l e s t a k e n during t e s t s . Based on a t y p i c a l f u e l CV
and d i r e c t m e a s u r e m e n t of f u e l w e i g h t and h e a t o u t p u t ,
e f f i c i e n c i e s i n t h e r e g i o n of 80% appear t o be a t t a i n a b l e
burning p e l l e t i z e d RDF.
We w e r e r e c e n t l y afforded an opportunity t o t e s t burn a
v e g e t a b l e r e s i d u e from t h e food p r o c e s s i n g i n d u s t r y . This
m a t e r i a l c o n s i s t s of f i n e p a r t i c l e s bound by 60% moisture
i n t o a p a s t e and having a GCV of approximately 2030 K c a l / k g .
T h i s m a t e r i a l w i l l n o t s u s t a i n bed temperature on i t s own
b e c a u s e t h e n e t t h e a t r e l e a s e d in the bed i s too low even
w i t h the s i d e panel f l u i d i z i n g a i r f u l l y i s o l a t e d . This heat
r e l e a s e r a t e i s , of c o u r s e , reduced by the combustion of a
p r o p o r t i o n of t h e v o l a t i l e s w h i c h o c c u r s a b o v e t h e bed
surface.
T e s t s w e r e c a r r i e d out with e i t h e r propane gas or c o a l
as s u p p o r t i n g f u e l s . Both were s u c c e s s f u l and i t was found
p o s s i b l e t o s u s t a i n bed t e m p e r a t u r e and good combustion
c o n d i t i o n s w i t h some 60% of the t o t a l heat input provided by
the waste r e s i d u e . D e s p i t e t h e s m a l l p a r t i c l e s i z e the
u n b u r n t l o s s due t o carry over was small as evidenced by the
c a r b o n c o n t e n t and q u a n t i t y of s o l i d s c o l l e c t e d in the bag
f i l t e r plant.
4. Conclusions
The l i m i t e d amount of work which we have been able to
c o m p l e t e so far has shown t h a t a r e c i r c u l a t i n g f l u i d i z e d bed
combustion system with the a b i l i t y to vary the amount of heat
e x t r a c t e d d i r e c t l y f r o m t h e b e d , as provided by s e p a r a t e l y
f l u i d i z e d side panel zones, i s p o t e n t i a l l y capable of burning
a wide v a r i e t y of waste m a t e r i a l s .
a) The p h y s i c a l p r o p e r t i e s of t h e w a s t e w i l l r e q u i r e
s p e c i a l a t t e n t i o n to the design of the feeder system to
e n s u r e a c o n t i n u o u s c o n t r o l l a b l e f e e d r a t e i n t o the
b o i l e r b u t when d i s c h a r g e d below the bed surface even
low d e n s i t y and f i n e l y divided m a t e r i a l s are adequately
e n t r a i n e d and burned to give acceptable unburnt s o l i d s
losses.
b) W a s t e s c o n t a i n i n g a h i g h p r o p o r t i o n of v o l a t i l e
h y d r o c a r b o n s tend to r e l e a s e these very r a p i d l y
r e s u l t i n g i n some c o m b u s t i o n above t h e bed and
c o n c e n t r a t e d in the region of the fuel i n j e c t i o n p o i n t s .
I t w i l l be necessary in many cases to provide secondary
a i r above t h e bed s u r f a c e s e l e c t i v e l y d i s t r i b u t e d to
ensure complete combustion of these v o l a t i l e s .
c) The n e t t h e a t r e l e a s e d within the bed and a v a i l a b l e to
s u s t a i n t h e t e m p e r a t u r e of t h e bed m a t e r i a l w i l l be
r e d u c e d by m o i s t u r e i n t h e fuel and also by the rapid
r e l e a s e of v o l a t i l e s a s the fuel e n t e r s the f l u i d i z e d
bed causing combustion in the freeboard. The a b i l i t y to
r e d u c e ' i n b e d ' h e a t a b s o r p t i o n allows v a r i a t i o n s in
h e a t r e l e a s e t o be accommodated within c e r t a i n l i m i t s .
Beyond t h e s e l i m i t s i t i s necessary to introduce a high
CV support f u e l .
d) The f a c i l i t y t o v a r y ' i n bed' heat absorption and the
p r o v i s i o n of c o n t r o l l a b l e secondary a i r together with
s u i t a b l e fuel feed systems should enable a wide range of
w a s t e m a t e r i a l s to be e f f i c i e n t l y burned in a fluidized
bed b o i l e r . Work t o d e f i n e t h e operating l i m i t s for
t h i s system i s proceeding.
-72-
Exhaust to Stack
Secondary Heat, Coal Feed Hopper

Transfer Tubes
Cyclone/
Dustbin
Screw Feed Hopper
Primary Heat
Transfer Tubes -
Water in
Water out
Distributor
Tiles
FIGURE 1 FLUIDFIRE PILOT BOILER DESIGN

DISCUSSION
Question - Chairman : When burning waste material with a high water

content, the incentive seems to be more pollution control
than energy conservation? please give comment.
Answer In fact the vegetable residue containing 60 % moisture contri-

buted some 60 X of the total heat input. 40 % from propane or
coal. This represents a significant fuel saving as well as the
benefit of pollution control by disposing of a waste which has
a strong and unpleasant odour. In many cases there will be
significant savings on disposal costs as well as the value of
the heat recovered.
Question - Mr Payne, NCB : What are the risks of bed fusion from low
melting point ash which may be introduced in the RDF feedstock
to the fluid bed ?
Answer There is clearly a risk of such materials being included in

refuse though the quantities after preparation seem to be small.
Host of the glass for instance is extracted in RDF preparation.
The circulating fluidized bed maintains very uniform temperature
conditions - tests were carried out between 800 and 850 C.
We burned approximately 1 1/2 tonnes of RDF with no indications
of fusion or sintering.
THE DEVELOPMENT OF A DEMONSTRATION FURNACE
FOR THE COMBUSTION OF LOW C.V. GASES
Authors : T.C. CLAYPOLE, C.J. BATES, N. SYRED
Contract number : EE-D-2-325-CB
Duration : 36 months 1st January 1981 - 31st December 1983
Head of Project : Dr. N. Syred, University College, Cardiff.
Contractor : University College, Cardiff
Address : Department of Mechanical Engineering & Energy Studies,

University College,
Newport Road,
GB - CARDIFF, Wales. CF2 1TA
Summary
The object of this project is to demonstrate the capability of swirl

burner/furnace systems to efficiently burn and recover heat from low
calorific value gases. The swirl burner/furnace systems have been
optimised both by industrial and laboratory trials.
The industrial trials have been on the incineration of the waste gas
produced during the discharge of the ovens at National Smokeless Fuels
Ltd.'s Phurnacite Plant. These have shown the potential of such systems
to incinerate low c.v. gases utilising their own inherent potential as
fuel. This part of the project has also led to the development of
automatic control systems to ensure efficient matching of the air and
support fuel requirements to minimise fuel consumption while retaining
efficient combustion. The control system is at present undergoing
reliability trials.
The design of the furnace has been optimised by a parallel series of

small scale laboratory trials. These have provided a valuable insight
into the combustion processes occurring within the furnace.
A commercial scale demonstration unit is currently being constructed in

the University. This paper reviews the development work which has
culminated in the design of the demonstration swirl burner/furnace
system.
-75-
1.1 Introduction
Low calorific value waste gases are often produced as a by-product of
industrial processes (e.g. the off gases from the manufacture of Carbon
Black and Smokeless fuels). These are regarded as a nuisance and are
frequently disposed of by incineration. This requires the addition of
premium fuel to ensure stable combustion. However, these gases can be dis-
posed of more cost effectively in a swirl burner/furnace system which
utilises their inherent potential as a low grade fuel. A useful heat
output can be extracted from these systems for use in the process (e.g. for
drying or steam generation), which would further reduce the consumption of
premium grade fuel. These systems also have a considerable potential for
burning difficult fuels (e.g. those which are low in volatiles).
The system comprises a swirl burner firing into a narrow cylindrical
furnace (Fig.l). A region of reverse flow is generated in the exit of the
combustor by the swirling action of the flow. The flame stabilises in the
wake of this aerodynamic blockage. The high levels of turbulence where
the forward flow stagnates causes intimate mixing of the fresh reactants
with the hot active species. This, coupled with the re-radiation of heat
from the furnace walls, gives rise to a highly stable, intense flame. A
small proportion of the flow (less than 10%) is drawn into the reverse flow
and recirculated. The majority of the flow passes from the well stirred
region into a less turbulent plug flow region, in which the reactions are
completed. Particles tend to be retained in the furnace by the centrifugal
action of the swirling flow until they are completely consumed. The
system is therefore well suited to the incineration of tar and dust-laden
gases.
Fundamental research has been undertaken into the mechanisms which
occur in swirl burners (1,2) and cyclone combustors (3,4). Spatially
resolved measurements of temperature, velocity and species concentration
have been made on laboratory scale combustion systems. They have also been
tested on industrial sites to show their ability to burn waste gases (5,6).
This project is to construct a swirl burner/furnace to demonstrate the
potential and to optimise the design of such systems. It is of the same
size as an industrial system, having 16 times the capacity of the existing
laboratory combustors.
2.1 The Design of the Demonstration Furnace

2.2 Industrial feasibility studies
The National Smokeless Fuels Ltd. Phurnacite Works at Aberaman manu-
factures smokeless fuel from coal blended with tar and pressed to form
ovoids. These are fired in ovens where the constitutents which generate
the smoke are driven off. The coke oven gas evolved during this process
is re-cycled to the ovenfe.
When the ovens are discharged it is necessary, for safety, to vent
this gas to atmosphere. Coke oven gas of high calorofic value (approxi-
mately 12 MJ/m 3 ) has to be disposed of at the start of the discharge. The
calorific value (c.v.) of the gas rapidly decays as' the ovens are emptied
(to less than 1 MJ/m 3 ).
Flarestacks are at present used to incinerate the gases. Natural gas
is mixed with the waste gases at the outlet of the flare and combustion
takes place immediately downstream. The design of tZie flare is optimised
for a gas of constant c.v. This is far from the case for this process
and a mean value of the c.v. is used for the design calculations. The
enrichment of the waste gas by the support fuel at the start of the cycle
serves only to deplete further the flame of oxygen which delays combustion
and increases the flame length. There is insufficient support fuel at the
-76-
at the end of the discharge cycle and the majority of the tars and particu-
lates escape incineration. The flame is readily disrupted by cross winds
which cause significant quantities of black smoke and particles to be
emitted.
The feasibility of achieving greater combustion efficiency using swirl
technology was demonstrated by a pilot study (5). A swirl burner/furnace
system was constructed based on the design of the research combustor (1)
and on earlier work for the Continental Carbon Company (7). Waste gas
was drawn from the flare stack and fed to this small scale system. The
system completely incinerated the waste gas while reducing the support fuel
requirement by more than 50%.
The success of these initial trials has led to a continuing programme
of development both in the laboratory and on site. Although this work has
concentrated on the problems of automatic control, the opportunity has also
been taken to improve the design of the furnace. The furnace has been
reduced in length by 30% without impairing its performance (6). The square
exit has been replaced with a converging nozzle (Fig.2). This has been
found to reduce noise by eliminating oscillations associated with the
corner eddy. It also produces a smooth flame front movement in response
to changes in the mixture ratio. The modifications both to the operation
and design of the system have produced support fuel savings of greater than
75%.
The pressure loss through the system is independent of the mixture
ratio (Fig.3) and also of the length of the furnace. The pressure loss is
greatly influenced by the level of swirl, which controls the size of the
reverse flow and hence the resistance in the throat of the combustion.
2.3 Small scale trials

A series of quarter scale trials have been initiated in order to
optimise the design of the furnace and to provide data for an investigation
into the effects of scale. Three furnace geometries have so far been
tested (8), Fig.4, at a constant swirl level (geometric swirl number,
Sg > 1.08).
The flow in the furnace can be divided into four main regions (Fig.5):
1. the swirling flow <-*-< 1.5)
re
2. the main reaction region ( 1 . 5 <c -f^
(1.5 J L < 3)
X re
3. the burn out region (re > 3)
4. the exit flow
Region 1 is similar to the free flame (1,2). Fresh reactants enter

furnace in a low temperature swirling forward flow (Fig.5). This ignites
at _x_=1.5 producing an increase in temperature. The forward jet entrains
re
air, which, since it is an enclosed flame can only be drawn from downstream.
This produces a corner eddy and increases the rate of divergence of the
forward flow compared to the free flame. The central reverse flow
probably represents <6% of the total flow (1) and contains high temperature
products of combustion (9).
The heat released by the chemical reactions in region 2 produces a
rapid rise in temperature.
The main difference between free and enclosed flows is found in region
3. This is a high temperature flow in which reactions can continue, unlike
the free flame where is considerable quenching due to the entrainment of
isothermal air (2). The heat released is equal to the heat lost via the
walls, giving rise to parallel temperature contours. However, the heat
-77-
f * ItUfato
lost via the walls only represents 5% of the total heat release In the
combustor.
The flow in the exit region 4 is influenced by the reactions upstream.
The expansion of the gas produces a considerable acceleration of the flow,
whereas the angular momentum is reduced by viscosity. Therefore the swirl
level is reduced and there is no tendency for a reverse flow zone to form
in the exit. The optimum exit geometry is the convergent nozzle. This
eliminates the corner eddy which reduces turbulence and hence the noise
output of the burner.
The heat released in the furnace is approximately 68% of the input
chemical energy, regardless of the furnace geometry or loading. The swirl
level and the initial expansion of the furnace were held constant in these
tests, giving similar flow patterns in regions 1 and 2 of the furnace.
Also high concentrations of CO and H 2 have been found in this region in
free flames (9). The majority of the chemical reactions must occur in
region 2.
The flow from the exit was influenced by the exit geometry and the
loading, but not the length of the furnace. In view of the small amount of
reaction which occurs in region 3, this could be reduced without impairing
the performance of the combustor. Therefore, the optimum geometry is
furnace 3, the short parallel section with the conical exit.
2.4 The demonstration furnace

A furnace has been designed, based on the experience gained from the
industrial and laboratory trials (Fig.6). It is four times the linear
size of the systems used for the earlier trials, which increases its flow
capacity by a factor of 16 (Table 1 ) .
A lightweight refractory brick can be used for the main chamber
because of the low local velocities. This reduces the weight of the system
and also affords better insulating properties. The reduction in heat loss
in this, the critical area for flame stabilisation, should further enhance
the stability limits of the system. However, in view of the high veloci-
ties in the throat and exit of the furnace, it will be necessary to use a
denser castable refractory in these regions.
The level of swirl can be varied by the use of inserts in the tangen-
tial inlets to the swirl chamber (Fig.7). Two access ports are provided
which can be fitted with tangential inlets to supply low c.v. gas to the
furnace.
The system is to fire vertically downwards (Fig.8). Water sprays will
reduce the exhaust gas temperature to 250C before it enters the chimney.
No provision has been made to extract the exhaust fumes in view of the
natural draught available from the chimney.
Access ports are to be provided along the length of the furnace.
These will enable probe measurements to be made and the ports will also be
capable of being fitted with windows so non-intrusive optical measuring
techniques can be used.
The mean temperatures within the laboratory combustors have been
measured using bare wire thermocouples (1,3) and gas samples collected, using
sonically quenching quartz microprobes (1,3). These techniques have been
proved (10) to be satisfactory for use in furnaces of the same size as this
demonstration unit. Therefore, the existing experimental techniques are
to be adapted for use with this system.
Laser anemometry has been used in furnaces (11). The systems which
are available in the University are being modified to enable measurements
of axial and tangential velocity to be made. A laser anemometry system is
currently being developed in conjunction with United Kingdom Atomic Energy
-78-
Authority.Harwell, for the Measurement of particle size and velocity
simultaneously (12). This is showing considerable potential and should
provide useful data especially in the combustion of particulates.
In addition to the spatially resolved measurements, the exhaust gases
are to be continuously monitored for 02, CO and CO2 levels. The rate of
temperature rise and the heat released in the furnace are to be measured
using thermocouples embedded in the refractory wall.
The flow of air into the system is to be measured using a bell-mouth
system which has been successfully applied to other high flow rate rigs(13).
3.1 Progress
An automatic control system has been developed to vary the supply of
combustion air and support fuel to match the requirements of the waste gas.
This is based on monitoring the exhaust temperature and oxygen level of
support fuel required.
The effect of furnace geometry on the system performance has already
been investigated on the quarter-scale prototype. These trials are
continuing, with the influence of swirl level and residence time to be
investigated. Also, blow off limits are to be measured.
The construction of the full scale furnace, the site preparation and
the development of measuring techniques are in hand.
4.1 Discussion
The majority of the reaction within the furnace occurs within the main
reaction region. The rate of reaction is highly dependent on the scale of
turbulence (14).
This was constant in the small scale tests, but should increase sig-
nificantly with the demonstration furnace. Therefore, one would anticipate
a greater quantity of the input chemical energy to be released as heat in
the region. The reactions have been found to be completed within a
furnace of approximately this scale (7) which supports this hypothesis.
However, it could be argued that the residence time was higher since the
furnace was slightly longer and, since the Reynolds number was of the same
order, the mean axial velocities are the same as the small scale furnace.
Therefore the time to travel the length of the furnace is increased by a
factor of 4. A longer furnace is being constructed which will give the
same residence time on the quarter scale model as is found with the demon-
stration furnace.
The furnace design shows several significant advances. These are:
the reduction in length, the conical exit and the simple construction of
the swirl burner.
Commercially available Incineration systems rely on maintaining the
gas at an elevated temperature for a long period. This necessitates a
large, well-insulated furnace to give low velocities and long residence
times. The use of swirl produces an intense reaction zone, which
eliminates the need for a long residence time. The system developed for
N.S.F. Ltd. is approximately one third of the size of systems proposed
based on long residence time combustion. It is also shorter than those
developed for the Carbon Black works (7) (approximately two-thirds the
length). The reduction in furnace size gives considerable savings in
capital cost. The associated reduction in weight enables the system to
be sited near the supply of waste gas. This can provide an even longer
capital saving, as for example at the Phurnacite Works, where the system
can be sited on top of the coke ovens, eliminating long irrigated pipe runs
which were a major portion of the capital expenditure on the full scale
project.
79-
The conical exit reduces the noise from the system. It also produces
a furnace in which it is easier to modulate the flame front.
The use of simple inlets for the swirl burner enhances its appeal to
industry. It reduces the manufacturing costs and makes the system less
susceptible toclogging (a major problem when handling particle-laden waste
gases).
5.1 Conclusions
A swirl burner/furnace system is being constructed to demonstrate the
feasibility of burning low c.v. gases. The design of the furnace has been
optimised by a series of small scale trials both on site and in the labora-
tory. The resulting furnace is considerably shorter than those previously
used for this purpose, thus considerably reducing the capital costs of an
installation.
The laboratory trials have enabled four regions in the furance to be
defined. The main reaction region is of most significance as the majority
of heat release occurs within it. The rate of heat release is dependent
on the flow in region 1, the swirl region. The flow in this region is
similar to that found in swirl burners and dictates the pressure loss
through the system.
Acknowledgements
The authors wish to acknowledge the financial support of the Commission
of the European Communities and National Smokeless Fuels Ltd. for this work.
References
1. Claypole, T.C. "Pollutant formation in swirling jets". Ph.D. Thesis,
University. College Cardiff, 1980.
2. Williams, E. "Characterisation of a swirl combustor". Internal
Report No.707, University College Cardiff, 1981.
3. Najim, S.E. "An aerodynamic study of cyclone combustor with gaseous
fuel". Ph.D. Thesis, University College Cardiff, 1979.
4. Sahati-Mehr, A. Ph.D. Thesis, University College Cardiff, 1982.
5. Claypole, T . C , Styles, A.C., Syred, N. "The incineration of discharge
gases by high intensity combustion". Internal Report No.619,
University College Cardiff, 1981.
6. Syred, N., Styles, A.C., Claypole, T.C. "Discharge gas incineration
system for the Purnacite Works, Aberaman. An interim report".
Internal Report No.776, University College Cardiff, 1981.
7. Syred, N., Dahman, K.R. "Effect of high levels of confinement upon the
aerodynamics of swirl burners". J. Energy, Vol.2, No.l, pp.8-15,
8. Hakimian, M. "Heat-transfer and aerodynamic interactions in swirl

burner furnace combinations". Internal Report No.836, University
College Cardiff, 1982.
9. Pomeroy, E. "The chemistry of swirl stabilised combustors".
Internal Report No.565, University College Cardiff, 1980.
10. Lockwood, F. Private Communication, 1982.
11. Hutchinson, P., Khalil, E.E., Whitelaw, J.H. "The measurement and
calculation of furnace flow properties". J. Energy, Vol.1, p.212,
1977.
12. White, H. "LDA study of two phase flows". M.Sc. Thesis, University
College Cardiff, 1982.
13. MacGregor, S.A. Private Communication, 1982.
14. Abdel-Grayed, R.G., Bradley, D. "A two-eddy theory of turbulent flame
propagation". Phil.Trans.Roy.Soc., Vol.301, pp.1-25, No.1457, 1981.
-80-
TABLE 1
Furnace Flow Parameters
3
/
Maximum air flow rate 50 m /nun
3
Maximum gas flow rate 4.8 m /min
Maximum input power 1.5 MW
. . *1
Maximum velocities
Combustor throat 33 m/s
Impingement point of fo rward flow 22 m/s
*2
Mean axial velocities
Furnace 19 m/s
Exhaust pipe 75 m/s
*1 derived from L.D.A. velocity measurements in free flames
*2 derived from total flow at mean temperature of 1800 k
Eilwutt
4 4 4 4 4
0//t\\0 Corrm tddy
////// Stagnation pant
C\W'3 Comtf <Jdy
III Rvfi flow
> 3 w v t cnvntMr
^
Fig. 1 Schematic diagram of flow in furnace
81
k
Sg = 1.26
10
I
1
in
61 " T -
> Sg = 0.63
i /
o
d c
6
/
, ' . a
>
X
10
o
/
/
A
/
o
1 1
' Swirl combustor
n
Ignitor
n^r-
0 1000. 2000 3000 4000 5000
Fig. 2 - Schematic diagram Total flow rale (l/min)
of swirl burner/furnace
system for phurnacite plant Fig. 3 - Pressure drop measurements for
short furnace combustion with 100% excess
air
Al dimensions In mm
152
|. l
1. Long furnace 2. Short furnace 3. Short furnace

with conical exit
Fig. 4 - Furnace geometries
-82-
To exhaust system
r*gh JUxrana refractory bncfc
datable refractory.
Ex* regwn 4
6004
Light weight
refractory brick
Bun out regwn 3
Majn reaction region 2 ! x i i
Swwl regwn 1
|, 300*
Flanga to fit swirl chamber Al dfcrwnticra in mm
Fig. 5 - Furnace flow schematic diagram and mean Fig. 6 - Demonstration furnace
temperature contours (k) (gas flow 1/min,
air flow - 1575 1/min)
Combustion air
Natural gas supply (for pre-mixed combustion)
To chimney
Access port
z=y
Natural gas supply
Plan
Axial .gas supply
hsert to alter swirl level Combustion swirl

i i " burner
r-,-!-J-1
Furnace
Al dimensions in mm
t Fig. 7 - Swirl Chamber

To chimney . r~Water sprays
. _ Water seal vent
-.^r /explosion
Fig. 8 - Demonstration Furnace Site Plan
Elevation
DISCUSSION
Question - Mr Davis, CEC : 1. How do you arrange your stability in your

system when using varying calorific value fuels ?
2. Do you see any problems when moving from the laboratory to
the actual plant ?
Answer 1. The swirl burner/furnace system is inherently stable over a

wide range of mixture ratios. The range is slightly dependent
on the flow rate reducing the flow rate increases the blow off
limits. Changes in mixture ratio produce flame front movements
within the body of the furnace. To deal efficiently with
changes in calorific value the air supplied to the system is
matched to the combustion requirements of the gas, thus
keeping the mixture ratio approximately constant. Careful
selection of the system loading insures stability through
transient variations in calorific value such as occur in
practise at the Phurnacite plant. An automatic control system
to regulate the air supply has been developed and it is intended
to present this work in the next report to the CEC.
2. The small scale system has already been successfully

demonstrated on the Phurnacite works, with only the normal
engineering problems associated with commissioning.
The full scale system being constructed in our laboratories
should prove directly comparable with an industrial unit.
-8S-
SESSION II - FLUIDIZED BED HEAT EXCHANGERS
Chairman: J. CARRASSE
Amelioration des echanges thermiques dans les lits

circulants
Etude de la circulation autoentretenue
Fluid bed high temperature qas/gas heat exchangers
Fluidized bed heat exchanger filter for waste heat

recovery from dirty corrosive gases
87-
AMELIORATION PES ECHANGES THERMIQUES
DANS LES LITS CIRCULANTS
Auteurs: G. CHRYSOSTOME, M. MARCELLIN

Contrat: n EE-B-113-F
Duree: 36 mois - ler Janvier 1981 - 31 decembre 1983
Responsable scientifique: G. CHRYSOSTOME
CREUSOT-LOIRE
Contractant: CREUSOT-LOIRE
Division Energie
Adresse: 71208 LE CREUSOT
Sous-contrat: UNIVERSITE DE TECHNOLOGIE DE COMPIEGNE
Responsable: J.F. LARGE
RESUME
Le but de la recherche est de concevoir des echangeurs de performances

optimales en s'appuyant sur une etude :
- de 1'influence de la geometrie generale de l'echangeur (disposition gene-

rale des tubes par exemple) sur l'ecoulement global de la suspension,
- de la forme de tubes d'echange et de leur geometrie de surface : tubes a

ailettes longitudinales ou spiralees par exemple.
La recherche fera intervenir une maquette existante de lit circulant ainsi

qu'un pilote de diametre 50 cm egalement existant.
-88-
PROGRAMME DE TRAVAIL
La recherche fait intervener deux partenaires : la Societe CREUSOT-LOIRE

-constructeur de materiel- (C.L.) et l'UNIVERSITE DE COMPIEGNE -Centre de
Recherche- (U.T.C.). Elle associe des etudes de caractere fondamental sur
1'hydrodynamique et les caracteristiques thermiques d'echangeurs a paroi
places dans un lit fluidise de type circulant, et des essais sur une ins-
tallation pilote existante permettant de tester leurs performances et leurs
tenues dans des conditions d'utilisation industrielle.
Pour ce faire, la recherche comportera quatre phases principales :
Phase 1 - Etude fondamentale de 1'hydrodynamique et des transferts thermi-

quea (Universite de Compiegne)
L'etude comprendra tout d'abord la definition des parametres hydrodynami-

ques susceptibles d'influencer le transfert thermique entre une suspension
en ecoulement vertical et une paroi d'echange. Ces resultats pourront etre
acquis a partir d'experiences conduites avec un tube lisse place dans des
conditions variables d'ecoulement d'une suspension gaz-solides.
Une fois obtenus ces resultats, on choisira une geometrie de tube particu-
lierement adaptee a un bon transfert thermique dans les conditions nabitu-
elles de fonctionnement des lits circulants. Compte-tenu des caracteristi-
ques propres des lits circulants, des tubes a ailettes longitudinales ou
spiralees (a large pas) seront vraisemblablement retenus. S'il s'avere que
le ruissellemcnt des solides le long des parois accrott l'efficacite de
l'echange, on examinera 1'influence des caracteristiques geometriques des
ailettes : nombre, hauteur, epaisseur, profil.
De memo, on etudiera la disposition relative des tubes et 1'influence de la

geometrie de 1'echangeur sur les performances.
Des etudes deja realisees a l'Uhiversite de Compiegne ayant montre que les
instabilities inhe rentes aux ecoulements diphasiques gaz-solides en regime
lache peuvent affecter les performances globales d'un echangeur de ce type.
On retiendra les dispositifs qui, tout en augmentant peu les pertes de
charges, ameliorent la stabilite de 1'ecoulement de la suspension et de ce
fait, peuvent promouvoir l'echange thermique.
De telles recherches seraient realisees sur une maquette reproduisant une

portion d'echangeur (echelle I) dans ses conditions d'utilisation en lit
circulant. Cette maquette, particulierement souple d'utilisation, pourra
fonctionner sur une gamme et endue de parametres hydrodynamiques mais ne
pourra cependant operer qu'a des niveaux thermiques de l'ordre de 150C
pour pouvoir beneficier des dispositifs de mesure de transfert thermique
d6j3 dcveloppes a l'U.T.C.
Phase 2 - Conception d'un echangeur industriel (C.L.)
L'etude precedente ( I) devrait conduire a plusieurs configurations d'echan-

geur particulierement adaptees aux conditions operatoires des lits circu-
lants (2 ou 3) .
La partie 2 de l'etude sera effectuee par une equipe CREUSOT-LOIRE specia-

lisee dans la fabrication des tubes a ailettes pour diverses industries.
-89-
Elle portera essentiellement sur les moyens d'obtenir industriellement les
tubes ailetes. La liaison avec l'U.T.C. sera assuree des le debut de l'etu-
de fondamentale pour eviter l1etude de tubes non industrialisables.
Les points principaux Studies seront :
. possibility d'obtenir directement les tubes ailetes par filage,

. possibilite d'obtenir les ailettes par extrusion,
. ailettes rapportees - probleme de la liaison thermique - soudage,
. possibilite d'utiliser des machines existantes modifiees,
. cout de fabrication.
Phase 3 - Definition des echangeurs d'essai (C.L.)
L'etude fondamentale des transferts thermiques sera effectuee sur un seul

tube.
II faudra done tenir compte des caracteristiques propres des lits circulants
pour definir le faisceau de tubes ailetes :
. geometrie du faisceau : pas des tubes, disposition...

. probleme des collecteurs qui ne devront pas gener le flux gazeux charge
en solides, ni etre sensibles a l'erosion.
Phase 4 - Essais dans un pilote a. lit circulant (C.L.)
Afin de limiter les problemes a l'etude du transfert thermique une combus-

tion de gaz naturel ou de fuel domestique sera effectuee dans le lit circu-
lant. Les essais seront effectues sur des tubes obtenus si possible par un
procede industriel (pour autant que les machines existent ou que leurs mo-
difications ne soient pas couteuses) ; ils comprendront :
- la verification des caracteristiques de transfert thermique definies dans

l'etude fondamentale :
. la localisation de l'echangeur dans le lit circulant, en tenant compte

des phenomenes d'erosion et d'encrassement lies a l'impact ou au depot des
particules,
. les mesures des pertes de charge supplementaires engendrees par le fais-
ceau tubulaire suivant les conditions d'utilisation,
. la tenue des materiaux.
3 - ETUDE FONDAMENTALE DE L'HYDRODYNAHIQUE ET DES TRANSFERTS THERMIQUES
3.1 - Installation experimentale

L'installation experimentale est une boucle composee d'une ligne de trans-
port vertical, d'un dispositif de mise en suspension des particules et d'un
systeme de separation et de recyclage du solide.
-90-
La ligne de transport de 20 mm de diametre interieur comporte trois por-
tions : une longueur d'acceleration de 4,5 m ; une section d'essais a
proprement parler de 4,2 m de long ; une portion aval de 2,5 m debouchant
dans le systeme de desengagement des particules. Pour 1'etude hydrodynami-
que, la section d'essai est equipee de 8 prises de pression parietales dis-
tantes de 50 cm. Elle est remplacee, pour les essais thermiques, par un tu-
be calorifuge chauffant la suspension par effet joule. Des resistances de
platine plaquees sur la paroi exterieure du tube -qui est tres mince
(0,5 m m ) - permettent de mesurer le profil longitudinal des temperatures de
paroi. La portion aval non chauffee mais calorifugee assure l'homogeneisa-
tion en temperature de la suspension en vue de verifier le bilan thermique.
Les solides sont introduits dans la ligne de transport a partir d'un lit
fluidise. Un systeme de reglage souple base sur le controle de la pression
dans le lit au point d'injection assure un debit de solide sensiblement
constant. Ce debit est mesure sur la ligne de recyclage qui comporte un
systSme de refroidissement des particules lors des essais thermiques.
Le debit de gaz est mesure a la sortie du tube de transport. Dans la jambe

de retour du solide, une garde defluidisee de 3 m est maintenue de maniere
a rendre negligeables les fuites de gaz par cette voie.
Deux vannes a fermeture rapide et simultanee (le decalage est inferieur a

0,05 s) placees de part et d'autre de la section d'essai permettent la me-
sure de la concentration moyenne dans la ligne. Lors des essais thermiques,
la densite de flux de chaleur est deduite de la puissance electrique fournie
au tube chauffant.
3.2 - Etat d'avancement des travaux
La precision des systemes de mesure et la reproductibilite des resultats

ont ete verifies.
Les essais systematiques n'ont porte jusqu'a present, que sur l'hydrodyna-
mique. Des nombres de Reynolds de l'ecoulement variant de 1600 a 15000
on ete realises.
Dans une premiere campagne d'essais,du sable de fonderie a ete utilise

(diametre moyen des solides 150 mm). Les resultats portant sur les pertes
de charge ont confirme les previsions theoriques. Une campagne d'essais
avec d'autres solides est en cours, qui permettra d'une part de confirmer
les previsions theoriques. Une campagne d'essais avec d'autres solides est
en cours, qui permettra d'une part de confirmer les interpretations avan-
cees et de mieux cerner, d'autre part, l'alteration de la structure de
l'ecoulement du gaz par la presence du solide.
Les essais thermiques sont en preparation.
A - ESSAIS D'ECHANGEURS DANS UN REACTEUR A LIT CIRCULANT
4.1 - Avantages attendus des Lits fluidises circulants

L'originalite du lit fluidise circulant reside dans le fait que l'on utili-
se une vitesse de fluidisation suffisante pour entrainer les particules.
-91-
Contrairement a ce qui se produit en fluidisation classique, il n'y a plus
de surface de separation nettement definie entre le lit fluidise dense sub-
sistant a la base de l'appareil et la zone diluee surmontant le lit flui-
dise. Le solide pulverulent qui est ainsi entraine, est separe du courant
gazeux au moyen d'un cyclone et recycl la base du racteur pour un con-
duit ayant approximativement la forme d'un siphon. Ce dernier permet d'equi-
librer grace a 1'intervention d'une colonne de particules, la difference
de pression regnant entre la base du reacteur et le cyclone.
Les avantages essentiels attendus de 1'utilisation des lits fluidises cir-

culants, comparativement aux lits fluidises classiques, sont les suivants :
- pour un meme volume, plus grande charge thermique,

- meilleure utilisation du volume pour des reactions chimiques telles que
desulfuration in situ de combustibles soufres,
- possibility d'utiliser des particules de plus faible diametre d'ou
accroissement de la surface reactive lors des reactions chimiques,
- uniformite de temperature d'ou meilleur controle des combustions et des
reactions chimiques,
- plus grande vitesse relative gaz-particules d : ou amelioration des trans-
ferts de chaleur et de masse,
- possibility de placer des surfaces d'echange thermique en partie haute du
reacteur, done possibility de "decoupler" transfert thermique et combustion,
ce qui est en general impossible en lit fluidise classique,
- plus grande souplesse du lit circulant, au moins 1 a 2, probablement 1 a
3 (peut-etre 1 a 4 pour des particules non abrasives),
- facilite de controle et de regulation,
- faible temperature de combustion ce qui evite la fusion des cendres et
reduit les emissions d'oxyde d'azote.
4.2 - Description de L'appareiLLage
L'appareillage se compose essentiellement des elements suivants :
- une chambre de combustion de diametre interieur 500 mm, garnie de refrac-

taire destin rsister l'rosion sur une epaisseur de 250 mm, de hauteur
utile 5 m et munie a sa base d'une boite a vent fermee par une grille a
tuyeres servant de distributeur, la virole exterieure ayant un diametre
de 1 m,
- un cyclone primaire assurant la separation des particules a recycler,
- un dispositif de recyclage des particules, en forme de siphon, de diame-
tre interieur 110 mm,
- un cyclone secondaire permettant la separation des fines avant le rejet
des gaz de combustion a 1'atmosphere,
- une tremie d'alimentation d'une capacite de 0,8 m3 equipee, a la base,
d'une vis sans fin a vitesse de rotation variable pour le reglage du debit
de combustible.
-92-
En debut d ' e s s a i , l a montee en temperature du r e a c t e u r s ' e f f e c t u e par com-
bustion de gaz n a t u r e l melange a l ' a i r de f l u i d i s a t i o n i n j e c t e dans un l i t
f l u i d i s e c o n s t i t u e d'un materiau i n e r t e . Cet i n e r t e e s t une q u a r t z i t e dont
l e s c a r a c t e r i s t i q u e s p r i n c i p a l e s sont :
- forme spherique des g r a i n s ,

- f o r t e teneur en s i l i c e (Si 02 > 99 Z ) ,
- t r c s f a i b l e teneur en c a l c a i r e e t elements a l c a l i n s ,
- d e n s i t e apparente non t a s s e e : 1,3 a 1,5,
- point de fusion : 1750C,
- granuloroe t r i e : 0,5 a 1,6 mm,
- charge i n t r o d u i t e : 75 kg.
Cette o p e r a t i o n se deroule au moyen de deux b r u l e u r s a b a f f l e , equipes cha-

cun d'unbruleur p i l o t e c o n t r o l e par une sonde a i o n i s a t i o n de flararae. Un
d e t e c t e u r U.V., sur chaque b r u l e u r , complete l e d i s p o s i t i f de s e c u r i t e . On
p o r t e a i n s i l e r e a c t e u r a une temperature comprise e n t r e 8U0 e t 900C. Onze
thermocouples chromel-alumel permettent de t r a c e r l a c a r t e thermique de
1 ' i n s t a l l a t i o n . Des p r i s e s de pression permettent de c o n t r o l e r l a p r e s s i o n
au-dessous e t au-dessus du l i t f l u i d i s e . La p e r t e de charge au niveau de
l a g r i l l e de f l u i d i s a t i o n e s t exprimee en mm de Hg au moyen d'un manometre
dif f e r e n t i e l . Le m a i n t i e n , au-dessus du l i t f l u i d i s e , d'une p r e s s i o n l e g e -
rement i n f e r i e u r e a l a pression atmospherique e v i t e 1'emission de gaz n o c i f s
dans 1'ambiance de l a s t a t i o n .
Une p r i s e d ' e c h a n t i l l o n s i t u e e e n t r e l e s deux cyclones permet de d e r i v e r

v e r s l e s a p p a r e i l s d ' a n a l y s e une f r a c t i o n des gaz de combustion. Leur com-
p o s i t i o n e s t mesuree au moyen d'un analyseur a absorption dans l ' i n f r a r o u g e
en continu pour CO e t C02, e t d'un analyseur en continu u t i l i s a n t l e s p r o -
p r i e t e s paramagnetiques de l'oxygene pour 02.
Les gaz de combustion sont e n s u i t e d i l u e s par de l ' a i r froid de maniere a

a b a i s s e r l e u r temperature a un niveau i n f e r i e u r a 200C puis f i l t r e s dans
une i n s t a l l a t i o n de depoussierage equipee de 50 f i l t r e s a manehe avant
d ' e t r e r e j e t e s a 1'atmosphere.
Tous l e s d e b i t s de f l u i d e s sont mesures au moyen de r o t a m c t r e s ou de d e b i t -

metres a diaphragmmes. Le d e b i t de combustible e s t determine par l e c t u r e
sur une courbe d ' e t a l o n n a g e e t par p e s e e .
La c a p a c i t e theorique maximum de I ' i n s t a l l a t i o n e s t de 1000 thermies/heure

(1160 kW) .
Compte-tenu de la masse r e f r a c t a i r e p r e s e n t e dans l a chambre de combustion,

l a montee en temperature se f a i t lentement e t demande en general 3 h pour
passer de 20C a 800C.
-93-
Les echangeurs a tester sont introduits par la partie superieure demontable
de 1'installation. Ces echangeurs sont parcourus par de l'eau dont on mesu-
re les temperatures d'entree et de sortie. Le debit d'eau est reglable entre
5 et 20 m3/h, permettant ainsi d'obtenir une hausse de temperature signifi-
cative (20 a 50^C) et done de limiter l'erreur sur la quantite de chaleur
echangee. C'est a partir de la valeur de cette quantite de chaleur qu'est
calcule le coefficient d'echange moyen. Apres passage dans l'echangeur
l'eau est refroidie au moyen d'un aerorefrigerant puis recyclee vers
l'echangeur.
4.3 - Avancenent des travaux
Les premiers essais sur tube lisse ont eu lieu en juin et ont permis de
verifier le bon fonctionnement de 1'installation. La prochaine campagne
d'essais est en preparation.
LIT FLUIDISE CIRCULANT EXPERIMENTAL
-94-
DISCUSSION
Question - Mr Pomeroy, NIHE : 1. What are particle speeds in your recircu-

lating fluidised bed ? If the velocity is in excess of 11 m/sec
erosion is a definite possibility.
2. What excess air level would be used in a recirculating
fluidised bed ?
Reponse 1. Les vitesses utiUsees sont inferieures a 11 m/s.

Pour reduire I'erosion il faut dans toute la mesure du possible
placer les surfaces d'echange paral lelement a I'ecoulement
gaz-solides.
2. L'exces d'air est lie aux caracteristiques du solide combus-

tible et a la temperature. Generalement des exces d'air de
15 X sont suffisants.
Question - Mr Robinson, Deborah Fluidised Combustion : In fluidised bed

systems having superficial velocities greater than 4 m/sec it
is necessary to separate the boiler heat transfer surfaces from
the region of high velocity to avoid erosion. Would the author
care to comment on the probable physical size of plan required
to produce 10 MW capacity using a fast circulating fluidised
bed.
Reponse L'erosion est severe pour toutes les parois disposees perpendi-
culairement a I'ecoulement gaz-solides. II faut done eviter une
telle disposition ou proteger les surfaces.
Une installation de 10 MW thermiques aurait un diametre d'envi-
ron 2 m a 2,5 m.
95-
ETUDE DE LA CIRCULATION AUTOENTRETENUE
Au thor s J.-F. LARGE, P. GUIGON, G. CHRYSOSTOME
Contract number EE - B - 1 - 114-F
Duration 21 months ; 1 August 1981 - 31 May 1983

Head of project G. CHRYSOSTOME and J.-F. LARGE,
Creusot-Loire and Universite de Compiegne
Contractor Creusot-Loire, France
Sub-contrac tor Universite de Compiegne, France
Address Creusot-Loire
71208 LE CREUSOT (France)
Preambule
La presentation qui suit constitue l'etat d'avancement de la recherche ef-
fectuee a 1'Universite de Compiegne sur la circulation autoentretenue de-
puis la signature du contrat avec la Societe Creusot-Loire, de'but Janvier
1982. La recherche n'a demarre effectivement qu'en mars 1982 apres le re-
crutement d'un chercheur.
Resume
La recherche concerne 1'etude du phenomene de circulation autoentrete-

nue induite par la presence d'un echangeur de chaleur compact forme de
tubes horizontaux dans un lit fluidise. Le phenomene est mis en evi-
dence et la force motrice de la circulation quantifiee. Les differents
systemes de mesure du debit de circulation sont passes en revue.
Summary
A selfsustaining circulation induced by the immersion of a compact

heat exchanger with horizontal tubes in a fluidized bed has been in-
vestigated. The phenomenon is demonstrated and the circulation driving
force is characterized quantitatively. The different systems used to
measure the circulation rate are reviewed.
-96-
1. Introduction
Cette recherche a pour but 1'etude de la circulation autoentretenue en
vue d'ameliorer le transfert thennique entre un lit fluidise et des tubes
echangeurs horizontaux.
Cette circulation est creee par 1'immersion au centre du lit d'une zo-
ne garnie de tubes faiblement espaces.
La circulation autoentretenue (definie au paragraphe suivant) est un
phenomene qui s'etablit naturellement.
Un des buts de la recherche est d'etudier la geometrie de la grille
de fluidisation afin de controler au mieux cet effet de circulation et d'en
faire varier l'intensite selon les applications et les contraintes opera-
toires independamment de la geometrie de 1'echangeur.
La circulation des solides a travers l'echangeur devrait conduire 3
une amelioration du coefficient de transfert thermique lit fluidise-tubes,
due au renouvellement accelere des particules en contact avec les tubes et
a la suppression du talus de particules immobiles qui apparaissent a la
partie superieure des tubes lorsque ceux-ci sont disposes horizontalement.
2. Etude preliminaire. Mise en evidence du phenomene de circulation

2.1. Etude bibliographicjue
Le phenomene de circulation dans un lit fluidise est un phenomene qui
apparatt naturellement dans les lits de grand diametre. Une macro-circula-
tion dans le lit (ou, tres souvent, la zone centrale est une zone ascen-
dante et la peripheric une zone descendante) a ete observee par de nom-
breux auteurs comme le mentionnent Grace et Harrison |'|.
Une circulation naturelle apparait aussi tres nettement quand on im-
merge dans un lit fluidise des paniers remplis de petites pieces metalH-
ques ou de billes, comme l'ont montre des experiences faites par Tamalet
| |. Dans ce cas le panier est constitue d'un cylindre vertical ouvert a
son extremite inferieure et il est muni d'une grille a larges mailles qui
supporte les billes (diametre 10 a 100 mm par exemple), de facon a laisser
passer les particules du lit fluidise. La figure 1 qui rend compte de ces
experiences montre que lorsque le panier est partiellement immerge, les fi-
nes particules remontent au travers de la couche de billes a un niveau net-
tement superieur a celui du lit fluidise. Cette difference de niveau est
d'autant plus importante que la hauteur de panier immergee est importante.
Lorsque l'on enfonce progressivement le panier dans le lit fluidise
(figure lb) il arrive un moment ou la difference de niveau devient egale
puis superieure a la hauteur de panier non encore immergee. A partir de cet
instant il se produit un debordement des particules fines qui retombent
alors dans le lit fluidise. II s'etablit ainsi une circulation des particu-
les constituant le lit fluidise au travers de la couche de billes. Le meme
phenomene de circulation autoentretenue a pu etre mis en evidence lorsque
le panier est totalement immerge dans le lit fluidise (figure lc).
L'effet de circulation autoentretenue peut egalement etre obtenu pour
des pieces de forme quelconque disposees en vrac dans un panier immerge
dans un lit fluidise. Plusieurs installations mettant a profit ce phenomene
ont ete construites pour realiser industriellement des traitements thermi-
ques de pieces de formes diverses.
Ce phenomene a aussi ete observe et utilise par Laguerie et al. | 2 |
dans un reacteur d'oxydation du butane en anhydride maleique ; la selecti-
vity a pu etre amelioree par la presence d'obstacles dans la colonne car
ceux-ci limitent la croissance des bulles et favorisent les echanges de
masse. Hydrodynamiquement, la presence d'obstacles dans une partie d'un lit
fluidise cree une zone de densite plus faible que dans le reste du lit et
amorce une circulation des solides de type thermosiphon.
-97-
Cette circulation peut etre amplifiee et maitrisee par une aeration
selective au niveau de la grille. Dans ce cas la presence d'obstacles n'est
plus indispensable. De La Nauze et Davidson | 3 | ont etudie un systeme sche-
matise sur la figure 2. lis ont mesure le debit de circulation des solides
en fonction de ia vitesse du gaz, ainsi que l1influence de la distance en-
tre le bas du tube et la grille. La vitesse de circulation des solides s'e-
tablit de telle sorte que les forces de frottements des particules equili-
brent la force due au gradient de concentration.
Un autre systeme se rapprochant de notre etude est le lit a recircula-
tion developpe par Westinghouse !**> 51 (Figure 3) pour la gazeification du
charbon. Dans cette application le jet central a une assez grande vitesse :
6 a 12 m/s. L'echange de chaleur est prevu dans les zones parietales de
circulation descendante des solides.
2.2. Premiere_caracterisation_exgerimentale_du_henomene_a_lÛniversite
de_Comgiegne
Quelques essais preliminaires ont ete effectues dans un lit rectangu-
laire de 0,02 m2 de section et de 0,3 m de haut muni de prises de pression.
Le but de ces essais etait de voir si la presence d'obstacles dans un lit
fluidise modifiait la densite du lit. Le lit a ete rempli d'anneaux de type
Pall soit totalement, soit sur une tranche horizontale. La figure 4 montre
qu'effectivement la presence d'obstacles modifie la densite du lit fluidi-
se. La difference constatee pour les faibles vitesses utilisees (5 cm/s)
n'est pas significative. D'autre part cette installation rudimentaire n'a
pas permis d'etudier les vitesses de gaz plus importantes a cause de l'en-
trainement des solides.
Pour 1'etude complete du phenomene on a ete amene a concevoir un dis-
positif permettant de mieux mettre en evidence la circulation et de mesurer
avec precision la difference de densite pour quantifier la force motrice de
circulation des solides en fonction des diverses conditions et configura-
tions operatoires. Ce dispositif est decrit dans le paragraphe suivant.
3. Conception et realisation d'un banc d'essai. Methode de mesure de la

force motrice
Un banc d'essai a ete concu et realise pour mesurer dans un premier
temps la force motrice (difference de densite) de circulation en fonction
des caracteristiques de l'echangeur (geometrie et arrangement des tubes) et
de la distance du faisceau par rapport a la grille.
3.1. Aggareillage
Un schema d'ensemble de 1'installation apparait sur la figure 5.
L'air est fourni par un surpresseur permettant de delivrer jusqu'a
200 m3/h sous 0,6 bars. Les debits d'air sont mesures a l'aide de diaphrag-
mes normalises. La colonne est formee de plusieurs sections en Plexiglass
de diametre interieur 0,29 m et de 0,54 m de haut, equipees de prises de
pression reparties regulierement tous les 0,05 m.
L'echangeur est contenu dans un parallelepiped'e en Plexiglass de
0,195 m de long, 0,125 m de large et 0,6 m de haut (figure 6 ) . La forme
rectangulaire facilite la variation de la geometrie de l'echangeur. Le dis-
positif est tel que l'on peut modifier facilement le diametre des tubes et
le pas horizontal entre les tubes. L'espacement vertical entre 2 rangees de
tubes est fixe actuellement a 0,05 m. Cet espacement peut eventuellement
etre diminue de moitie. Le diametre des tubes d'echange a ete fixe a
0,254 m, mais tout autre diametre peut facilement etre etudie. Cet element
est muni de prises de pression regulierement reparties afin de pouvoir me-
surer la densite du lit a 1'interieur de l'echangeur.
-98-
Lea solides utilises sonc dans les premiers essais des billes de verro
de diametre moyen 800 um.
3.2. Pr emiers_resu^tats_exj>erimentaux
La figure 7 montre la difference de densite entre 1'interieur de l'e-
changeur et la zone peripherique. Cette difference de densite ne semble pas
beaucoup influenced par la vitesse de fluidisation dans la gamme des vites-
ses etudiees, elle est de l'ordre de 240 kg/m3.
3.3. Plan d'exprience pour la premire phase d'tude

Pour cette premiere phase on raesure la force motrice pour differentes
geometries de l'echangeur. On etudie principalement en fonction de la vi-
tesse de fluidisation 1'influence du nombre de rangees de tubes et de l'es-
pacement horizontal entre tubes. Ensuite la distance du bas de l'echangeur
a la grille sera modifiee afin de mettre en evidence la hauteur a partir de
laquelle cette hauteur n'a plus d'influence.
Ces experiences seront faites avec plusieurs diametres de tube et plu-
sieurs types de solides afin de degager 1'influence de la masse volumique
et de la granulometrie.
Pendant que ces mesures seront faites, une autre installation sera
concue afin de mesurer le debit de circulation et de voir l'influence d'une
aeration selective a la grille.
4. Mesure du debit de circulation

Le debit de circulation est certainement un des parametres principaux
gouvcrnant l'echange de chaleur. II est done essentiel de connattre ce de-
bit en fonction de la geometrie du systeme et des conditions operatoires
(debit de gaz de fluidisation et repartition de ce gaz).
La mesure du debit de circulation n'est pas simple si l'on ne veut pas
perturber le phenomene. Toutes les metnodes consistant a interrompre la
circulation pendant un temps, meme court, sont a proscrire. II ne reste que
les methodes permettant d'obtenir la vitesse des solides, ces methodes
etant conjugees avec une methode de mesure de la densite.
4.1. Mthode_de_mesure_de_la_vitesse des garticules

La vitesse des particules peut etre obtenue a priori de quatre manie-
res differentes :
1) en suivant une particule tracee ;
2) en mesurant le transfert de chaleur entre les solides et un objet ;
3) en mesurant 1'impact des solides sur un objet ;
4) l'aide de fibres optiques.
Dans la premiere methode qui consiste a suivre une particule tracee il
s'agit souvent d'un tracage radioactif. C'est la technique employee par De
La Nauze et al. | 3 |. Elle necessite beaucoup de precautions.
Si la particule est une particule coloree, la vitesse est mesuree en
suivant cette particule a travers une paroi transparente a l'aide d'une ca-
mera. Cette technique est plus facile a mettre en oeuvre que la precedente
et nous pensons l'utiliser dans un premier temps. Le reproche que l'on peut
faire a cette technique est que la vitesse a la paroi n'est pas necessaire-
ment la meme que celle a l'interieur du lit. Cependant Thorley et al. | 6 |
ont montre que la vitesse a la paroi mesuree dans un "spouted bed" consti-
tue une assez bonne methode pour obtenir le debit de solides dans la zone
du jet central.
La deuxime methode a ete utilisee notamment par Marsheck et Gomez-
Plata |'| pour etudier le mouvement des particules dans un lit fluidise. La
-99-
sonde est formee de deux thermistances, l'une sert de source de chaleur et
l'autre de recepteur. On mesure la perte de resistance du recepteur quand
il s'echauffe et l'on en deduit la vitesse locale massique apres etalonnage.
La troisieme methode consiste a utiliser les impacts sur un cristal
piezoelectrique'| . |. Les chocs des particules sur le cristal sont trans-
formes en un signal electrique. Les particules dont le diametre est du mime
ordre de grandeur que le cristal sont reperees separement par un pic sur un
oscilloscope. La hauteur du pic est proportionnelle a la quantite de mouve-
ment de la particule. Quand les particules sont plus petites que le cristal,
les pics sur 1'oscilloscope ne sont plus separes et la hauteur des pics de-
pend a la fois de 1'impact et de la frequence. La relation entre la hauteur
du pic et la vitesse est fonction de la porosite. Dans ce cas la hauteur
moyenne des pics est une mesure de la vitesse massique. Cette sonde neces-
site un calibrage prealable.
La derniere technique pour obtenir la vitesse des particules consiste
a utiliser une sonde formee de fibres optiques | |. La fibre centrale ame-
ne le flux lumineux, les autres fibres recueillent la lumiere reflechie par
les particules.
4.2. Mesure_de_la_2orosite_du_lit
Dans un lit fluidise ferae non garni la masse volumique du lit peut
etre obtenue facilement a l'aide des prises de pression. La perte de charge
AP entre deux prises distantes d'une hauteur h est egale a :
AP = p g h
p masse volumique du lit.
Dans le cas d'un lit fluidise ouvert avec ecoulement (interne ou ex-
terne) cette relation n'est plus valable a cause de la perte de charge due
au frottement sur des obstacles ou parois. Pour mesurer la porosite en un
point donne on utilisera une sonde capacitive du type de celle decrite dans
l'article de Taganov et al. | 9 |. Cette sonde capacitive est introduite dans
le circuit d'un autogenerateur dont la frequence varie en fonction de la
capacite de la sonde. A l'aide d'un discriminateur la deviation de la fre-
quence est convertie en une difference de potentiel proportionnelle a la
porosite du lit.
5. Conclusions
Le montage realise a confirme que la circulation autoentretenue exis-
tait bien quand on immerge dans un lit fluidise uri echangeur compact n'oc-
cupant que le centre du lit. Des resultats preliminaires permettant de
quantifier la force motrice de la circulation ont ete obtenus et une serie
d'experiences est en cours pour determiner 1'influence des divers parame-
tres operatoires.
Un second montage est en cours d'etude pour permettre de faire varier
la distribution de l'air au niveau de la grille. La zone centrale sera ali-
mentee avec une vitesse superieure a celle des zones peripheriques. Cette
repartition de l'air devrait permettre d'amplifier et de mieux controler le
phenomene de circulation que l'on cherche a creer.
Bibliographie
I1! GRACE, J.R., HARRISON, D., dans Fluidization, J.F. DAVIDSON et
D. HARRISON, ed. Academic Press, 1971.
|2| LAGUERIE et al., Congres "La fluidisation et ses applications",
Toulouse, 1973.
r 31 LA NAUZE, R.D., et al., "Fluidization Technology Vol. II", p. 113,
D.L. KEAIRNS Ed., 1976.
-100-
I'M YANG, W.C., et al., I. E. C. Proc. Des. Dev. J4 (3) 259 (1975).
l5l YANG, W.C., et al., "Fluidization Technology Vol. II", p. 51,
D.L. KEAIRNS Ed. , 1976.
l6l THORLEY, B., et al.. Can. J. Chem. Eng. 37, 184 (1959).
ri MARSHECK, R.M. , et al., A. I. Ch. E. J. j_l^ (1), 167 (1965).
i8i HEERTJES, P.M., Powder Technol. 4 (1), 38 (1970-71)
n
|10|
TAGANOV, I.N., et al., Theor. Osn. Khim. Tekh. J4 (4), 558 (1980).
SITNAI, 0., I. E. C. Proc. Des. Dev. 20 (3), 533 (1981).
mi TAMALET, M., DECHEMA Vol. 65, p. 57-65 (1970).
oo _
W-
nt
r b
Figure 1
Phenomene de circulation auto-entretenue
n
Tamalet | |
-Vf*'
,o-
' l-VM
:
:i . T .T. c ft
%
H2
u. ' uj u. \
t

Figure 2 Figure 3
Draft tube apparatus Lit a recirculation
De La Nauze | 3 | de Westinghouse | 3 i **t 5 |
101-
P (cm CE)
A U sans garnissage
0 Lir avec garmssage tixe
* Lit avec garmssage entre A e t B
Figure 4
Mise en evidence de la difference de densite
Garnissage : anneaux Pall 15 mm
Sable dp = 220 ym
V = 5 cm/s
1 . Surpresseur 6 . Vanne
9
2.Soupape 7. Distribut-eur de gaz 9
9'
3-Manometre (0-15bar3B.Colonne de fluidisation 9-
9
4.Diaprragme 9.Prises de pression 9-
9'
5.Manometre a eau IClEchangeur 9"
9-
A
3<
-*
9"
9;
AIR
O w
6
Figure 5
Schema de 1 ' i n s t a l l a t i o n
102-
W ra
Ii
A mz a/////////
195 mm
\
\
y.%
235 mm
Figure 6
Echangeur Vue de dessus
" p kg/m 3
1600
/~~. ~ p zone
peri pheri que
1500
force motrice
H00
A
A
1300 centrale
1200
Ug m/s

0,05 0,10 0.15 0.20
Figure 7
Force tnotrice de circulation
Billes de verre dp 800 um
Tubes : diametre 25,4 mm
pas horizontal 78 mm
pas vertical 50 mm
Distance echangeur grille 180 mm
103
DISCUSSION
Question - Mr Knobbout, Expert : For Laboratory tests mostly small

fluidised beds are used, and then the wall effects can be
very-high. This can give a problem of scaling up the bed.
What is the situation in your bed ?
This is a general problem for calculating the effects of an
industrial bed from laboratory measurements.
Reponse Dans notre cas le fait que le lit soit petit est plutot
defavorable et on peut s'attendre a de meilleurs resultats dans
un lit de plus grande taille. L'influence de la taille pourra
etre mise en evidence par comparaison avec les resultats
obtenus au Creusot.
Question - President : Dans une circulation auto-entretenue, la vitesse des

solides augmente mais le poids specifique diminue. Les deux
phenomenes ont des effets contradictoires sur le coefficient
d'echange. Peut-on prevoir La part de chaque effet et leur
resultante ?
Reponse Si on se base sur les lits fluidises classiques, la courbe du

coefficient de transfert en fonction de la vitesse du gaz est
d'abord croissante puis atteint un maximum avant de decroTtre.
Quand on augmente la vitesse, la masse volumique diminue certes
mais la mobilite des particules augmente.
Question - Mr Pomeroy, NIHE : The curve for heat transfer coefficient vs

velocity shows a maximum at a critical velocity. The density
of the fluidised bed in the centre is less, presumably because
of greater voidage. With this in mind do you not think that
you will be on the down leg of this curve when high voidages
are involved ?
Reponse Non, je pense que nous sommes sur la partie ascendante de la

courbe. Des experiences faites sur des paniers de solides
immerges dans le lit fluidise ont montre que le transfert etait
nettement ameliore dans le cas de la circulation auto-entretenue.
Question - Mr Graf, Lurgi : Up to which size of the fluid bed can you use
your draft tube system ?
Reponse Des draft tubes de I'ordre de 1 m2 semblent realisables.
-104-
FLUID BED HIGH TEMPERATURE GAS/GAS HEAT EXCHANGERS
Author: J N SCUTTER, BSc(Hons)

Contract number; EE-B-1-142-UK
Duration:
Head of Project: R Burrows, Stone Fluidfire
Contractor: Stone Fluidfire
Address: 56 Second Avenue
Pensnett Trading E s t a t e
Brierley Hill
West Midlands
DY6 7PP
Summary
T h i s pape r s u m m a r i s e s t h e h i s t o r y of the F l u i d f i r e
f l u i d i z e d bed g a s t o g a s h e a t e x c h a n g e r from i t s
c o n c e p t i o n i n 19 75 by P r o f . Douglas E l l i o t t upto the
c o m p l e t i o n of t h e r e s e a r c h work c o n d u c t e d a t t h e
U n i v e r s i t y of Ason i n 19 81 by D C Newey. The paper
t h e n d e s c r i b e s t h e work c a r r i e d out under EEC funding
since t h e A s t o n r e s e a r c h was c o m p l e t e d and a l s o
o u t l i n e s t h e c u r r e n t t e s t programme and o b j e c t i v e s ,
f i n i s h i n g w i t h an o u t l i n e of f u t u r e work proposed
r e l a t i n g t o t h e d e s i g n of a p i l o t p l a n t s c a l e heat
exchanger.
105-
1. Introduction
1.1 Description of Method of Operation of Heat Exchanger
The p u r p o s e of the heat exchanger i s to provide a means
of exchanging heat from a hot gas stream to a cold gas stream
u t i l i s i n g a c i r c u l a t i n g f l u i d i z e d bed of i n e r t granular
p a r t i c l e s (Alumina g r i t ) as the heat t r a n s f e r medium.
F i g u r e 1 i l l u s t r a t e s the method of o p e r a t i o n . The heat
e x c h a n g e r i s d i v i d e d i n t o a hot gas and a cold gas s e c t i o n .
The h o t flue gas i s drawn i n t o the plenum chamber, from where
i t p a s s e s up through the d i s t r i b u t o r . The hot flue gas gives
up i t s h e a t t o t h e c i r c u l a t i n g f l u i d i z e d bed and then i s
drawn o u t from the freeboard chamber and through the induced
draught fan.
S i m i l a r l y , the cold a i r i s drawn i n t o the combustion a i r
f o r c e d d r a u g h t fan and i s exhausted into the plenum chamber.
The a i r passes through the d i s t r i b u t o r and picks up heat from
t h e b e d . The heated a i r i s then exhausted from the freeboard
c h a m b e r t o t h e b u r n e r s / c o m b u s t i o n c h a m b e r of t h e
furnace/boiler.
U n l i k e some of t h e c i r c u l a t i n g fludized beds t h a t have
been i n v e s t i g a t e d by other researchers ( B o t t e r i l l , e t c ) t h i s
s y s t e m d o e s n o t u s e a mechanical method of bed c i r c u l a t i o n
( e . g . paddle wheel).
The d i s t r i b u t o r s h e e t s are manufactured from s t a i n l e s s
s t e e l sheet and the s l o t s are formed by male punch and female
die. The s l o t s a r e p r o d u c e d such t h a t the gas i s given a
h o r i z o n t a l component of momentum a s w e l l a s an upward
component. The upward component ' f l u i d i s e s ' the bed of
p a r t i c l e s , and t h e h o r i z o n t a l component (when s u f f i c i e n t l y
h i g h ) d r i v e s the p a r t i c l e s across the d i s t r i b u t o r . When the
d i s t r i b u t o r s a r e l i n k e d t o g e t h e r in the form of a closed
c i r c u i t , t h e bed i s driven around the c i r c u i t by t r a n s f e r of
momentum from t h e g a s t o t h e b e d . F i g u r e 11 shows the
d i s t r i b u t o r s linked together in a closed c i r c u i t , and figures
111 and IV show a cross section through a d i s t r i b u t o r sheet
and i l l u s t r a t e the s l o t o r i e n t a t i o n .
The two g a s s t r e a m s a r e k e p t s e p a r a t e by p a r t i t i o n
plates. The p a r t i t i o n p l a t e s i n t h e plenum chamber are
p o s i t i v e l y sealed a l l the way round. The p a r t i t i o n p l a t e s in
t h e f r e e b o a r d chamber p r o j e c t into the bed such t h a t they
m i n i m i s e g a s leakage but allow the free passage of p a r t i c l e s
from one section to the other (hot to cold and vice v e r s a ) .
The h e a t e x c h a n g e r o p e r a t e s as a p a r a l l e l flow system
where t h e c o l d g a s e x i t temperature can only approach, but
not exceed, the hot gas e x i t temperature.
In o r d e r t o a s s e s t h e t e c h n i c a l f e a s a b i l i t y of the
c o n c e p t and t h e p o t e n t i a l market, a period of research work
was c a r r i e d out a t the University of Aston by D C Newey under
the I n t e r d i s c i p l i n a r y Higher Degrees Scheme.
1.2 Objectives of the Research work c a r r i e d out a t the
University of Aston
The work c a r r i e d out a t Aston had three p r i n c i p a l a i m s : -
a) To study the t e c h n i c a l f e a s i b i l i t y of the concept.
b) To develop s u f f i c i e n t data to enable the design of an
-106-
i n d u s t r i a l s c a l e p r o t o t y p e t o be undertaken,
c) To i d e n t i f y the waste heat recovery market s e c t o r s in
which the heat exchanger i s l i k e l y to be c o m p e t i t i v e .
S t u d y i n g the t e c h n i c a l f e a s i b i l i t y required the
m a n u f a c t u r e of a small s c a l e and i n e x p e n s i v e u n i t which would
b e c a p a b l e of s h o w i n g t h a t t h e bed w o u l d c i r c u l a t e a s
d e s c r i b e d , and w o u l d t r a n s f e r heat as a p a r a l l e l flow heat
exchanger. With t h e s e f a c t o r s in mind, the heat exchanger
was c o n s t r u c t e d as shown i n Figure V. The heat exchanger can
be s e e n i n t h e m i d d l e of t h e p i c t u r e below the e x t r a c t o r
hood. The a i r i s f e d from t h e f a n ( o u t of p i c t u r e ) and
t h r o u g h t h e Rotameters, part of the a i r i s blown through the
e l e c t r i c heater and then i n t o the r i g h t hand s i d e of the heat
exchanger. The c o l d a i r supply i s fed in from the Rotameters
and t h e n t h r o u g h t h e f r o n t of t h e h e a t e x c h a n g e r . The
e q u i p m e n t shown on the l e f t of the p i c t u r e was used for the
d e t e c t i o n of leakage from one gas stream t o a n o t h e r .
In o r d e r f o r f u t u r e d e s i g n s t o be c a r r i e d o u t , and
p e r f o r m a n c e p r e d i c t i o n s t o be made, i t i s e s s e n t i a l for the
d e s i g n e r t o h a v e t h e n e c e s s a r y d a t a a t h i s d i s p o s a l . To
s a t i s f y t h e s e requirements the research work was conducted in
the f o l l o w i n g manner:-
i) A m a t h e m a t i c a l model was proposed which would s i m u l a t e
the heat exchangers' performance,
ii) T h i s m o d e l r e q u i r e d e m p i r i c a l v e r i f i c a t i o n and the
o p t i m i s a t i o n of c e r t a i n v a r i a b l e s which are necessary in
o r d e r t o work t h e m o d e l ( e . g . o p t i m u m f l u i d i s i n g
velocities, b e d d e p t h and p a r t i t i o n g a p ) . The
p e r f o r m a n c e of the heat exchanger was a s s e s s e d in terms
of the a f f e c t of the above v a r i a b l e s upon e f f e c t i v e n e s s ,
i i i ) The t h i r d p a r t of t h e Aston research work i n v o l v e d an
i n v e s t i g a t i o n i n t o t h e p o t e n t i a l markets for the heat
exchanger, t h e s e were i n i t i a l l y seen a s :
a) space h e a t i n g from hot f l u e g a s e s
b) combustion a i r p r e - h e a t i n g
The m a r k e t r e s e a r c h a l s o e n t a i l e d an assessment of the
s u i t a b i l i t y of e x i s t i n g heat recovery equipment in t h o s e
sectors.
1.3 Summary of Conclusions from U n i v e r s i t y of Aston
Research Work
The t e s t i n g of t h e h e a t e x c h a n g e r v e r i f i e d t h a t the
c o n c e p t was i n d e e d f e a s i b l e , and t h a t i t o p e r a t e d as a
p a r a l l e l flow heat exchanger. The i n v e s t i g a t i o n of gas
l e a k a g e revealed t h a t the leakage r a t e could be c o n t r o l l e d t o
w i t h i n a c c e p t a b l e l i m i t s , and would enable the heat exchanger
t o operate s a t i s f a c t o r i l y in c e r t a i n market a r e a s .
The p e r f o r m a n c e of t h e u n i t was a s s e s s e d i n terms of
heat exchanger effectiveness. This i s d e f i n e d as the r a t i o
of t h e a c t u a l h e a t t r a n s f e r t o t h e maximum t h e o r e t i c a l l y
p o s s i b l e ( w h e r e t h e maximum t h e o r e t i c a l l y p o s s i b l e assumes
t h a t t h e c o l d f l u i d could be heated t o the i n l e t temperature
of t h e hot f l u i d ) . For a p e r f e c t p a r a l l e l f l o w h e a t
e x c h a n g e r , w i t h a p p r o x i m a t e l y e q u a l h o t and c o l d f l u i d
t h e r m a l c a p a c i t i e s , t h e maximum e f f e c t i v e n e s s of a s i n g l e
s t a g e heat exchanger i s approximately 50%. The g e n e r a l
107-
t r e n d s of e f f e c t i v e n e s s with varying f l u i d i s i n g v e l o c i t y , bed
d e p t h and p a r t i t i o n gap are shown in Figure VI. The general
c o n c l u s i o n t h a t can be drawn from these graphs i s t h a t any
p a r a m e t e r which a d v e r s e l y a f f e c t s bed c i r c u l a t i o n w i l l
a d v e r s e l y a f f e c t e f f e c t i v e n e s s , i . e . decreasing f l u i d i s i n g
v e l o c i t y , i n c r e a s i n g bed depth and decreasing p a r t i t i o n gap
a l l reduce e f f e c t i v e n e s s .
The i n i t i a l m a r k e t s u r v e y i n d i c a t e d t h a t there i s a
r e q u i r e m e n t for a cheap and r e l i a b l e heat exchanger t h a t w i l l
o p e r a t e a t t e m p e r a t u r e s i n e x c e s s of 400C. Within t h i s
r a n g e , c u r r e n t l y a v a i l a b l e equipment suffers from various
p r o b l e m s which r e s u l t in e i t h e r excessive cost or lack of
reliability. The p r i n c i p a l market for the heat exchanger
w i l l be i n c o m b u s t i o n a i r p r e - h e a t i n g . The o t h e r idea
p r o p o s e d , space h e a t i n g , can be discounted on the grounds of
g a s l e a k a g e and p o s s i b l e p a r t i c l e c a r r y - o v e r . Any design
b a s e d upon t h i s a p p l i c a t i o n would have to look very c l o s e l y
a t l e v e l s of p o l l u t i o n - even bearing in mind t h a t future
d e v e l o p m e n t s c o u l d r e d u c e t h e leakage r a t e , and, t h a t the
h e a t e d a i r would be h e a v i l y d i l u t e d downstream of the heat
exchanger.
With the combustion a i r pre-heating system the two fluid
mass flow r a t e s w i l l be approximately e q u a l , the u n i t w i l l
need t o o p e r a t e w i t h h i g h h o t g a s t e m p e r a t u r e s , and the
l e a k a g e r a t e w i l l n o t be s o c r i t i c a l a s f o r o t h e r
a p p l i c a t i o n s (save for unnecessary fan power consumption).
The main p o t e n t i a l a d v a n t a g e of t h e s y s t e m i s i t s '
s i m p l i c i t y , which manifests i t s e l f in terms of cheapness of
c o n s t r u c t i o n and p o t e n t i a l l y high r e l i a b i l i t y . The p r i n c i p a l
d i s a d v a n t a g e s are the p r e s s u r e drop through the heat
e x c h a n g e r , and the s i z e of the heat exchanger. (The bed area
of t h e h e a t exchanger i s determined by the gas flow r a t e and
the f l u i d i s i n g v e l o c i t y ) .
2. Description of Small Scale Prototype
2.1 Introduction
F o l l o w i n g the work c a r r i e d out a t Aston and completed in
A p r i l , 19 8 1 , the next stage of development was determined to
be t h e c o n s t r u c t i o n of another small scale heat exchanger.
I t was d e c i d e d t h a t many of t h e v a r i a b l e s needed further
o p t i m i s a t i o n since the Aston t e s t i n g did not experiment with
g a s t e m p e r a t u r e s a b o v e 250 C , o r w i t h different
distributor geometries. When more information i s a t hand,
t h e n t h e large prototype can be b u i l t in order to i n v e s t i g a t e
problems of s c a l e - u p .
2.2 Design Requirements and Test Objectives
2.2.1 F i r s t l y , a d i f f e r e n t bed geometry needed to be
tried. The Aston h e a t e x c h a n g e r s u f f e r e d from p a r t i c l e
c i r c u l a t i o n p r o b l e m s , t h i s can be seen on the d i s t r i b u t o r
p l a n view of f i g u r e V l l . The arrows show the d i r e c t i o n of
t h e bed c i r c u l a t i o n , and the shaded areas show where the bed
was n o t properly f l u i d i s e d . This phenomena was caused by too
rapid a change of d i r e c t i o n a t the end of the c e n t r a l
-108-
d i v i s i o n p l a t e ( v i r t u a l l y a s i n g l e 1 8 0 c h a n g e of
d i r e c t i o n i n s t e a d of two of 9 0 ) .
The most o b v i o u s c h a n g e of g e o m e t r y was t o make the
c i r c u i t c i r c u l a r . T his was seen t o have s e v e r a l a d v a n t a g e s : -
a) I t w o u l d be p o s s i b l e t o c h a n g e t h e d i r e c t i o n of the
p a r t i c l e s more g r a d u a l l y a t e a c h d i s t r i b u t o r p l a t e
junction.
b) A l l of t h e p l a t e s would be the same shape - t h i s could
have production advantages a t a l a t e r d a t e .
The d i f f e r e n c e between the Aston heat exchanger
d i s t r i b u t o r g e o m e t r y and t h e l a t e s t geometry i s shown in
F i g u r e V l l l . T he s m a l l s c a l e prototype diagram shows s i x
p l a t e s j o i n e d t o g e t h e r . T he number was chosen a r b i t r a r i l y
but with due c o n s i d e r a t i o n t o : -
a) t h e more p l a t e s t h e r e a r e , t h e l e s s t h e c h a n g e of
d i r e c t i o n required a t each j u n c t i o n .
b) t h e more p l a t e s there a r e , the g r e a t e r the non-
f l u i d i z i n g areas where the p l a t e s are welded t o g e t h e r .
The c e n t r a l column was provided t o avoid the c i r c u l a t i o n
problems encountered with the Aston heat exchanger.
2 . 2 . 2 T he s e c o n d r e q u i r e m e n t for the heat exchanger was
t h a t i t w o u l d h a v e t o be d e s i g n e d s u c h t h a t i t could be
t e s t e d u p t o a b o u t 8 0 0 C, t h i s w o u l d be a c h i e v e d by
supplying heat from a s u i t a b l e gas burner.
T h i s t e s t i n g would then p a r t i a l l y a s s e s s the s u i t a b i l i t y
of t h e h e a t e x c h a n g e r f o r t h e 400 C p l u s s e c t o r of the
market already mentioned.
2 . 2 . 3 T h i r d l y , t h e m a t h e m a t i c a l model proposed by David
Newey f o r making p e r f o r m a n c e p r e d i c t i o n s needed f u r t h e r
v e r i f i c a t i o n with l a r g e r gas temperature d i f f e r e n c e s .
2.2.4 F o u r t h l y , a more d e t a i l e d i n v e s t i g a t i o n was
r e q u i r e d of g a s l e a k e a g e r a t e s from one gas stream t o the
o t h e r , and of e f f e c t i v e means of c o n t r o l l i n g the l e a k a g e .
T h i s c a n e i t h e r be a c h i e v e d by s t a g g e r i n g the plenum and
f r e e b o a r d p a r t i t i o n p l a t e s such that the leakage i s forced t o
g o i n a p r e f e r r e d d i r e c t i o n , o r , a p r e s s u r e d i f f e r e n t i a l can
be s e t a c r o s s t h e two f r e e b o a r d c h a m b e r s , again f o r c i n g
l e a k a g e in a p r e f e r r e d d i r e c t i o n . With most a p p l i c a t i o n s
t h e r e w i l l be a p o s i t i v e p r e s s u r e d i f f e r e n t i a l between the
c o l d and h o t s i d e s o w i n g t o t h e f o r c e d d r a u g h t / i n d u c e d
draught fan arrangement.
2.2.5 F i n a l l y , more i n f o r m a t i o n was required about many
v a r i a b l e parameters in order t h a t the l a r g e s c a l e p i l o t p l a n t
h e a t e x c h a n g e r c o u l d be d e s i g n e d . In p a r t i c u l a r , from a
g i v e n gas flow r a t e , s u f f i c i e n t information must be a v a i l a b l e
r e g a r d i n g optimum f l u i d i z i n g v e l o c i t i e s in order t h a t the
d i s t r i b u t o r / b e d area can be a c c u r a t e l y determined. Similarly
more i n f o r m a t i o n i s r e q u i r e d about optimum bed c i r c u l a t i o n
r a t e s and p a r t i c l e r e s i d e n c e t i m e s in each gas s e c t i o n in
o r d e r t h a t t h e d i s t r i b u t o r geometry can be determined ( i n
terms of l e n g t h / w i d t h r a t i o s ) .
109-
2.3 Description of Beat Exchanger and Test Rig Layout
F i g u r e IX shows the small scale heat exchanger as i t was
m a n u f a c t u r e d in April/May 1982. The gas and a i r connections
can be s e e n q u i t e c l e a r l y as can the f l u i d i s e d bed and the
distributors. U n d e r n e a t h t h e d i s t r i b u t o r s are the plenum
chamber d i v i s i o n p l a t e s . These s l i d e such t h a t the amount of
s t a g g e r w i t h t h e f r e e b o a r d chamber p a r t i t i o n p l a t e can be
altered. The plenum p a r t i t i o n has been constructed in t h i s
way so t h a t the effect of staggering the p a r t i t i o n s upon gas
leakage r a t e can be i n v e s t i g a t e d .
The f r e e b o a r d chamber p a r t i t i o n p l a t e s are adjustable
such t h a t t h e p l a t e p e n e t r a t i o n i n t o the bed (or the gap
b e t w e e n t h e p l a t e and the d i s t r i b u t o r ) can be varied and i t s
affect investigated.
The l i g h t e n t r y and viewing p o r t s have been constructed
so t h a t t h e flow c h a r a c t e r i s t i c s of the bed can be checked
d u r i n g o p e r a t i o n a t temperature, i n i t i a l i n v e s t i g a t i o n s into
f l u i d i s i n g and c i r c u l a t i o n c h a r a c t e r i s t i c s w i l l be conducted
with the l i d removed.
The sand f i l l e r p o r t s have been added so t h a t the bed
l e v e l can be adjusted e a s i l y when i n v e s t i g a t i n g the effect of
d i f f e r e n t bed depths.
The c a s i n g i n s u l a t i o n i s T r i t o n Kaowool which i s a
m i n e r a l wool t h a t can be a p p l i e d e x t e r n a l l y to the heat
e x c h a n g e r . This method has the advantages t h a t i t s i m p l i f i e s
c o n s t r u c t i o n and m i n i m i s e s t h e r m a l expansion problems by
having the casing and d i s t r i b u t o r s a t similar temperatures.
O t h e r v a r i a b l e s t h a t w i l l be i n v e s t i g a t e d using t h i s
heat exchanger a r e : -
i) The p a r t i c l e size of the alumina g r i t bed material
i i ) Freeboard chamber height
i i i ) D i s t r i b u t o r sheet design - in terms of numbers of p l a t e s
and optimum s l o t design.
The p h o t o g r a p h s shown in Figures X, XI and Xll show the
c o m p l e t e d h e a t exchanger and the freeboard chamber p a r t i t i o n
plates.
F i g u r e X l l l shows the t e s t r i g layout. I t can be seen
t h a t t h e h o t and c o l d s i d e s a r e supplied by the same fan.
The b u r n e r i s mounted to a small insulated chamber where i t s
c o m b u s t i o n p r o d u c t s can be mixed with attemperating a i r in
o r d e r to achieve the temperature/flow rate condition
required.
3. Summary of Test Programme
3.1 F l u i d i s a t i o n and p a r t i c l e c i r c u l a t i o n c h a r a c t e r i s t i c s
w i l l be i n v e s t i g a t e d by t e s t i n g with cold a i r . The top of the
h e a t e x c h a n g e r w i l l be removed such t h a t v i s u a l examination
is possible.
3.2 A s e t of t e s t s w i l l be c a r r i e d o u t w i t h a h o t gas
t e m p e r a t u r e of 250 C i n order to compare r e s u l t s with, and
verify the work of David Newey.
no-
3.3 T e s t i n g w i l l be c o n d u c t e d with higher hot gas
temperatures (up t o about 800C).
3.4 T he e f f e c t s of g a s l e a k a g e will be observed and gas
leakage r a t e s w i l l be q u a n t i f i e d .
3.5 When a l l o t h e r p a r a m e t e r s have been o p t i m i s e d through
t h e t e s t i n g m e n t i o n e d , the e f f e c t of bed p a r t i c l e s i z e w i l l
be i n v e s t i g a t e d w i t h a l l o t h e r v a r i a b l e s a t t h e i r optimum
settings.
3.6 T he e f f e c t of d i s t r i b u t o r d e s i g n w i l l be i n v e s t i g a t e d .
4. Future Work
The r e s u l t s o b t a i n e d from t e s t i n g the small s c a l e heat
e x c h a n g e r s h o u l d y i e l d s u f f i c i e n t data to enable the next
s t a g e of r e s e a r c h t o be i n i t i a t e d . T his w i l l i n v o l v e the
d e s i g n of a l a r g e r p i l o t p l a n t heat exchanger, c o n s t r u c t e d
a l o n g the same l i n e s as the small s c a l e u n i t . T he production
of t h e l a r g e r u n i t w i l l a l l o w p r o b l e m s ot s c a l e up t o be
investigated.
5. References
Some T e c h n i c a l and Commercial Aspects of f ' l u i d i s e d Bed Gas t o
Gas Heat E x h a n g e r s - D C Newey PhD T h e s i s , U n i v e r s i t y of
Aston, Birmingham, England.
in-
rue a*, TT HO
CUT U FIG. I
CROSS SECTION THRO1 HEAT EXCHANGER
SEATED Alg O f f
Tea
RCJLTI',j
v
CD'-O * ' R i
^ ^//////^ t
i=D> =>
FIG. Ill
SECTION THRO1 DISTRIB UTOR
-NBED SOLIDS. n.ow
t= f t f f e
112
FIG. IV - GAS TO GAS HEAT EXCHANGER DISTRIBUTOR
'1
FIG. V - UNIVERSITY OF ASTON TEST RIG LAYOUT
113-
FIG. VI
TYPICAL PERFORMANCE CHARACTERISTICS
OBTAINED FROM ASTON HEAT EXCHANGER
to
to
BED DEPTH
OO
00
PARTITION GAP
00
to
Umf FLUIDISING VELOCITY
-114-
DISTRIBUTOR SLOT
rornmmBir-
SUJMPEp/NON-
ZOMSFD AREA
BED CIRCUUTION C H A ^ C T E R J S T I C S
OF ASTON HEAT EXCHANGER
*I21HEAJ_XCHANGER
_3L0T DIRECTION
FIG. V I I I
-MS-
SAND REPLENISHMENT AIR OUTLET
TUBE, FITTED WITH CONNECTION
SEALING CAP.
(AIR & GAS SIDES)
CALIBRATED
SLIDING
GUILLOTINE
PLATES
GAS
OUTLET
CONNECTION
SIGHT TUBE
AIR & GAS SIDES
LIGHT ENTRY
PORTS. AIR &
GAS SIDES
CIRCULATING
FLUIDISED
BED
DISTRIBUTOR
SHEET WITH
UNDIRECTIONAL
SLOTS
CENTRAL DIVISION
PLATE
VARIABLE SLIDES
BENEATH BED
ADJUSTABLE
EXTERNALLY
GAS INLET
CONNECTION
AIR INLET
CONNECTION
FIG. IX
SECTIONALISED VIEW OF FLUIDISED BED GAS TO GAS HEAT
EXCHANGER
116-
FIG. XII LID OF HEAT EXCHANGER SHOWING PARTITION PLATES
Pilot Air
Pilot Gas
Ultra Violet
r Heat Exchanger
Flame Failure
ki/J Detection
J<w- -S ><P&w
Stordy
Gas Burner
r, KEY:
Propane Gas
Air
Back Pressure Governor Loading
F/D FAN
I8
-&*3- Valve or Damper
ck|- Solend Valve
Governor'
.]_ -Of- Non-Return Valve
-CK3- Gas/Air Mixer (Plot)
Flow Measurement
Air/Gas System Schematic
Pressure Switch
FIG. XIII
DISCUSSION
Question - Mr Knobbout, Expert : 1. Can you give some indication of the

price difference between a heat wheel system and this system
for industrial application ?
2. What is the maximum pressure difference between the hot and
cold stream, as the leakage in a heat wheel can easily be
controlled, in principle at least ?
Answer 1. Development of the fluidised bed heat exchanger is not

sufficiently advanced at the moment to be able to offer firm
indications of price. However, the most apparent advantage
of the heat exchanger is its simplicity, in as much as the
need for complicated seal arrangements and rotating machinery
is eliminated. Ue would, therefore, expect the unit to be
considerably cheaper.
2. Without having conducted any testing on the heat exchanger,

it is difficult to give a realistic indication of what the
maximum pressure differential might be. Indeed, this is one
of the principal areas of investigation in terms of its affect
on the leakage of gas from one sector of the heat exchanger
to the other.
Question - Dr Ehringer, CEC : What is the maximum temperature.
Answer 800 C.
Question - Dr Ehringer, CEC : But you would have corrosion problems ?
Answer You might be right.
-119-
FLUIDIZED BED HEAT EXCHANGER FILTER FOR
WASTE HEAT RECOVERY FROM DIRTY CORROSIVE GASES
Authors C. DALLA ROLLE, D.PETRUCCIOLI, R.MARGARY,

R.FESTA, S.VACCARO
Contract number EEB-1-110-I
Duration 48 months 1 July 1981 - 30 June 1983
Head of project D. Petruccioli - Centro Ricerche Fiat
Contractor Centro Ricerche Fiat
Address Centro Ricerche Fiat

Strada Torino 50
I 10043 ORBASSANO (Torino)
Summary
This paper briefly summarizes uhe results achieved about the flui-
dized bed heat exchanger investigation. These results, obtained un
der the test conditions and by means of the experimental apparatus
described beyand, have been compeared with the ones of a conventio
nal heat exchanger. Furthermore a sensitivity study has been held
to quantify the influence of the main geometrical and thermodinamic
parameters on the heat exchanger performance. From this work, it
can be pointed out that the developed fluidized bed heat exchanger
has shown a very good behaviour both under the point of view of the
heat exchanger and of the soot removal from exhausts, moreover, ba-
rely affecting the engine performance.
120-
1. Introduction
Hf-at recovery from diesel exhaust or, generally, from factory hot
flu" ftases, has a serious drawback because of their particulate contents.
In conventional shell and tube heat exchangers, tube fouling leads to
such a dramatic decrease of the exchanger efficiency that they have to be
serviced after few days of operation (1),(2). Fluidized bed heat exchan-
g e s overcome this inconvenience, even if fouling problems may still arise
in the exhaust gas distributor (3).
The present research mainly aims to design and test a fluidized bed
heat exhanger/fliter as part of a diesel total-energy system.
Experiments have been carried out with a cold model of the exchanger
fed by a steady state gas s t r e a m , a laboratory scale apparatus COIJ
pled with a 5 kW engine (1), and with a semi-industrial scale exchanger/
filter coupled with a 62 kW engine. An industrial prototype of an exchan-
ger/filter for a 210 electric kW cogeneration group has been designed by
scaling up the latter one. Two peculiar features characterize the exchan-
ger/filter developped: the gas distributor (4), consisting of horizontal
manifolds with lateral nozzles, the design of which prevents plugging even
with high particulate content in the gases, and the capability of the appa
ratus to remove part of the soot from the engine exhaust gases.
In this paper the results obtained with the semi-industrial scale
apparatus are reported and the following aspects analyzed :
- the performance of the distributor, with reference to pressure drop,
blackflow of solids into the manifolds through the nozzles and uniformi-
ty of gas distribution;
- the performance of the apparatus as heat exchanger with reference to bed
temperature distribution, bed to tube heat transfer coefficients and tu-
be fouling;
- the results of soot filtering, its mechanism and removal efficiency as a
function of some process variables;
- the influence of overall pressure drop on engine performances.
2. Experimental apparatus
Exhaust gases from a naturally aspirated, four strokes, four cylin-
ders 62 KW OM Tigrotto C02D/1 diesel engine are discarged into the fluidi-
zed bed heat exchanger represented in fig.I. The column, made of AISI 304,
has a rectangular cross section 250x400 mm^ and a height of 700 mm. When
a higher free-board is required, column height can be raised to 1350 and
2000 mm with extension sections 650 mm high. Two tempered glass windows a^.
low observation of the upper part of the bed. The gas distributor (fig.IIA
and B) is made of two, 400 mm long, 40 mm ID, stainless steel manifolds
with their axes placed at 75 mm from the bed bottom and 62 mm from the cen
treline. The two manifolds are connected via an Y shaped joint to the ex-
haust pipe of the engine and have a closed end at the opposite side. Each
manifold, coated with a 10 mm thick thermal insulation, feeds 20 nozzles,
6.5 mm I.D. with chanfered edge at the inlet. Nozzles are oriented 30
downwards from the horizontal and are 12 mm long. A bank of horizontal
AISI 304 tubes, 400 mm long, 12 mm ID, 16 mm OD, is arranged in the bed
with an horizontal pitch of 45 mm and a vertical pitch of either 12 mm or
36 mm. The highest row of tubes is located at 420 mm from the bottom.Both
manifolds and tubes are parallel to the longer side of the column. The gas
bearing the bed is sent to cyclones for the separation of soot agglomera
tes. Tests are carried out with a low pressure drop cyclone having a 40x80
mm2 entrance section and with four high efficiency cyclones in parallel,
each with a 9x50 mm 2 cross section.
Cooling water circulates in 17 passes through 85 tubes, from bottom
to top to prevent formation of air bubbles. Inlet temperature is set at
the desired value by appropriate cool water make up to a loop. Most of the
runs are performed with a water inlet temperature of about 60C, correspon
ding on the average to an outlet temperature of 85C. Dashed circles in
fig.I indicate where adiabatic dummy tubes, thermocouples to measure bed
temperatures, or probe tubes to perform bed to clean tube heat transfer
rate measurements are inserted.
Two bed solid materials are tested: 220'yU.m average diameter quartz
sand and 600y0.ni glass ballotini. Static bed heights are of 240 mm and 310
mm respectively in such cases. Bed temperatures are in the range of 120
220C. At these temperatures superficial minimum fluidizing velocities of
about 4 and 24 m/s are evaluated for the two materials. Under engine ope
rating conditions, superficial fluidizing velocities range between 50 and
110 m/s. Changes of these velocities reflect those of the engine revolu
tion speed. Exhaust gas temperatures vary between 500 and 600C, depending
on the engine speed, and the air/fuel ratio.
Soot concentration upstream and downstream the apparatus are measured
by on line gas sampling. A scheme of the soot measurement system is shown
in fig.III.Particle size distribution of particulate separated at cyclones
and collected on bed solids are obtained by Coulter Counter analysis.
3. Results and discussion
3.1 Gas distribution*
The dispersion of gas in bubbles throughout the bed is satisfactory.
Gas channelling or a zone with unfluidized solids are avoided. Experiments
performed with the cold model show a little difference in bed expansion
between a porous plate distributor and the manifold distributor (fig.IV).
However a slight non uniformity in gas discharge distribution through the
nozzles at different distances from the manifold entrance, appears both
from measurements with steady state gas feeding and from the results of a
numerical model. Dimensionless static and total pressure along manifold
are shown in fig.V. Total pressure slightly decreases with distance from
the gas entrance, indicating low friction losses in the manifolds, fur
thermore the increase of static pressure is due to the reduction of the
gas velocity along the manifolds itselves. This results in a slight increa
se of the flowrates through the nozzles starting from the manifold entran
ce as shown in fig.VI. They range within the 20% around the value of 1.89;
that is the ratio of the cross section manifold area to the total nozzles
area used in the semiindustrial apparatus.
* a detailed analysis of the distributor is reported in (5)
122
The distributor design has to match the two conflicting requirements
of limited pressure drop and of no bed solids backflow. With the descri-
bed configuration and under experimental conditions, the distributor pres
sure drop ranges from 2 to 5 kPa. Solids backflow is caused by a tempora-
ry inversion of the pressure gradient across the nozzles. In fig.VII it is
shown that, at steady state engine conditions, the pressure gradient does
not reverse expecially at high engine revolution speed. Solids backflow,
seldom observed and almost exclusively during the engine shut-down, have
negligible effect. Bed solids incidentally refluxed are always subsequen-
tly ejected.
3.2 Heat transfer
Bed temperature distribution is relevant for the evaluation of the
bed to tube heat transfer coefficients and for the investigation on the
gas distributor performance. Contrarily to the temperature uniformity ty_
pical of fluidized beds, differences in local temperatures may result in
the apparatus used because of the bank of tubes and of the peculiarity
of the distribution adopted.
Temperatures in beds of 60n^jim glass ballotini at locations (4), (7)
and (9) of the t ibe array (fig.I) and at three fluidizing velocities are
reported infig.VHas a function of dimensionless the distance X/L from
the side wall through which manifolds are inserted in the bed. Temperatu
res increase along the horizontal direction with X/L. About the vertical
direction, bed temperatures are generally higher in (4) than in (9) near
the closed end of the distributor (X/L=l), whereas the reserve occurs at
the apposite side (X/L-O). These trends suggest that a gulf stream circu
latlon is superimposed to the bubbling bed, with fluidized solids moving
upward close to the column vail at X/L-l and downwards near the opposite
walL. This gulf stream circulation is according to the expectation with
regard to changes in nozzle discharge flow rate along the manifold.
Heat transfer coefficients, reported in table 1, are measured without
tube deposits at three locations in beds of both sand and glass ballotini.
Although small, differences in the measured values suggest that heat tran
sfer coefficients are higher in the loosest section of the tube array for
both solids tested. Changes in the values of these coefficients with par-
ticle size are significant whatever the location of the probe tube. In
heat recovery systems, particle size, as pointed out by Elliot (7), appe-
ars to be the most relevant variable for bed to horizontal tube bank con-
vective heat transfer.
Tube fouling due to soot laden gases is monitored. In fig. IX and X
the thermal resistance due to soot deposits is reported, as a function of
time.
This latter is evaluated from: Rf = 1 - 1
H Ho
where H is the overall heat transfer coefficient and Ho its initial value
with clean tubes. Test conditions are the same for the first 34 and 24
A more extensive analysis of heat transfer performance is reported

in (6)
-123-
hours of fig.IX and X respectively.
Data of Rf overlap within the experimental accuracy and exhibit a
linear increase with time.
The abrupt fall of Rf in fig.ix and" X is related to the reestablishment
of the initial situation with clean tubes beacause of defouling. Condi
tions determining defouling in the two cases, however, are different. Do
fouling during the run of fig.ixis induced by a rise of surface tempera
ture of the tube bank, simply obtained by interrupting temporarily cooling
water feed. Consequently deposits spontaneously and completely detach from
the surface. Severe cooling of the tube surface from 85C down to 60C
also produces defouling as shown for the run in fig. X. This figure also
indicates that further decreases of tube surface temperature cause another
lowering of the fouling thermal resistance, i.e. further defouling.
Optical measurements indicate that the average thickness, td, of

carbonaceous materials deposited on the tubes after 30 hr of operation is
about 150ju.m. A value of about n.1.3 W/mC for the deposit thermal conduc
tivity K d = t d /R f is determined.
The sensitivity of the overall heat transfer coefficient to changes
in Rf is 9H = H 2
dRf
In fluidized bed the measured values of Rf significantly affect the eve
rail heat transfer coefficient H due to its tipical high valves. The re
lat> ve reduction of H, in 30 hours of operation under the conditions of
fig.IXis AH/H = 0.19.
On the other hand, it can be shown that in fluidized bed heat exchan
gers the sensitivity of the exchanger efficiency 6 to relative changes
in H is lower tân in conventional exchangers. Under the experimental con
ditions of fig. IX a.A= 0.025 is measured after 30 hours of operation.
Fig.XI gives as a function of the ratio of cumulative fuel consumption
to the heat transfer area. In addition, this figure reports the efficiency
of a conventional shell and tube heat exchanger operating with diesel ex
hausts under conditions close to the ones of the present work (1).
3.3 Soot removal **
The mechanism of soot removal is the following: soot is collected by
bed solids and agglomerates on them. Then, because of bed solids attri
tion, soot is reentrained in the gas stream in coarse agglomerates that
can be collected by the cyclones.
Soot removal efficiency results are summarized in fig.XII e XIII.
The removal efficiency of the apparatus ^ is defined as the ratio of the
soot removed to the soot emitted from the engine and is variously evalua
ted (B arone 1982).
At about 21 engine air/fuel ratio, soot removal efficiency is less
than 50% and scarsely affected by the bed temperature which ranges from
120 to 220C(fig.XII). On the contrary, increasing air/fuel ratio,
*** A more complete discussion about this subject is reported in (8)
124
at bed temperature about 110C, soot removal efficiency grows up to 85%
(fig.XIII). As shown in fig.XII and XIII the higher the efficiency, the
higher the agglomeration and the average diameter of the carbonaceous ma
terial separated by the cyclones. A tipical size distribution of the par
ticulate collected on bed solids is presented in fig.XIV. It refers to
the highest efficiency test condition of fig.XIII.
Soot agglomerates show different hydrogen to carbon ratio H/C. In
particular this ratio results to vary from 0.2 to 0.7 for air/fuel ratio
which ranges from 20 to 30. This suggests that the higher agglomeration
observed when the air/fuel ratio increases, depends from a better adhesivit)
of the soot to the bed collector, due to the different features of the soot
and/or the different content of condensable hydrocarbons.
3.4 Engine performance with backpressure
The increase of the engine backpressure, under the experimental flui
dizing velocities, is due to the pressure drops in the distributor (25
kPa), through the bed (3.54.5 kPa depending on static bed height) and
through the cyclones (0.53.5 kPa for the low pressure drop cyclone and
416 kPa for the high efficiency cyclones). The influence of this overpres
sure (Ps Pso) on the engine performance in shown in fig.XV. The diffe
rence between the mean equivalent pressure (m.e.p.) and that obtainable
without overpressure (m.e.p0 .) decreases with (Ps P s o ) almost linearly
and substantially independently from the engine revolution speed and load.
Consequently, the encrease in percentage of the specific fuel consumption
is higher at a lower load. At a higher load it is limited to 13% for the
backpressures induced by the system.
4. Conclusions
FLuidized bed heat exchangers equipped with banks of water tubes can
be conveniently used to recover heat from exhausts of Diesel engines.
The manifold distributor tested in the present work overcomes what appea
red to be a serious inconvenience in past applications of these units,
i.e. the plugging of ordinary plate type distributors by soot deposits.
Bed temperature profiles indicate that, provided the manifold distributor
is properly designed, slight non uniformities of flow rates of gases
going out from nozzles at different distances from the entrance do not
significantly affect the transport properties of fluidized solids.
Measured bed to clean tube heat transfer coefficients are in agree
ment with those reported in the literature for heat exchangers of the sa
me type and same bed solids size. Soot fouling of the tubes results in a
relatively rapid decay of the overall bed to water heat transfer coeffi
cient because of the low thermal conductivity of the deposits. Within the
experimental conditions of this work, the deposit thickness increases
about linearly with time to build up a resistance which, after 30 hr of
operation, is about 20% higher than that related to the heat transfer from
bed to clean tube. Whereas the sensitivity of the overall heat transfer
coefficient is high, the sensitivity of the efficiency of the heat exchan
JUT unit to changes of such coefficient is low.
125
After 30 hr of operation it only falls of about 3%. Top values of bed to
clean tube heat transfer can be reestablished inducing tube defouling by
stepping up or down the tube wall temperature.
Soot particles are agglomerated in the fluidized bed at diameters
10-100 times higher than those at the engine outlet and eventually can be
separated by means of cyclones.
Bed temperature has slight influence on soot removal afficiency; on
the contrary, engine air/fuel ratio strongly affects the agglomeration
and consequently the removal efficiency of the soot. Changes in air/fuel
ratio cause differences in features of diesel emissions. It results, at
higher air/fuel ratio, a better adhesivity of the soot to bed collectors
which is likely due to the different characteristics of soot itself and/or
to a higher condensable hydrocarbon content.
The system overall pressure drop barely influences the engine perfor
mances.
TABLE 1 - Bed to clean tube heat transfer coefficients
Tube Thermocouple hDt

position
pos: position (W/m2 C)
Silica sand (4) (4) 305
Dp = 0.220 mm (7) (7) 320
U = 0 . 9 0 m/s (13) (11) 285
Glass ballotini (4) (9) 225

Dp = 0.600 mm (7) (9) 230
U = 0 . 8 0 m/s (13) (9) 220
126-
Nomenclature
Am manifold cross section area

Atn total nozzles area
Cs specific fuel consumption
Dp bed particle average diameter
dp diameter of collected soot
dpm average diameter of collected soot
H overall heat transfer coefficient, bed height
Hmf static bed height
H0 overall heat transfer coefficient with clean tubes
hbt bed to tube heat transfer coefficient
kd deposit thermal conductivity
L length of the longer side of the column and of the manifolds
m.e.p. engine mean equivalent pressure
n number of nozzles per manifold, engine revolution speed
p static pressure
P-p total p r e s s u r e
Ps engine backpressure
Rf fouling thermal resistance
t time
td average thickness of deposits on tube
T temperature
U fluidization velocity
u velocity inside the manifold
W gas massflow rate in side the manifolds
at engine air/fuel ratio
, heat exchanger efficiency
<ll soot removal efficiency
Subscripts
b bed
gi gas bed inlet
go gas bed outlet
o absence of overpressure to the engine, distributor inlet
t tube
wi water in
-127-
References
(1) Campanile A., Carlomagno G.M., De Vita A., Scognamiglio A.,

Donsi G., Massimilla L., 1980."Soot Removal and Heat Recovery
from Diesel Engine Exhaust by means of a Fluidized Bed Filter/
Agglomerator and Heat Exchanger". Proceedings International
Conference ISATA p. 164,1
(2) Thielbahr W.H., 1980. "Fouling in Diesel Fuel Exhaust Streams",

I E A Conference,Berlin
(3) Cusdin D.R. and Virr M.J., 1979. "A Marin Fluidized Bed Waste
Heat Boiler" Trans. I. Mar. E. 91.
(4) Campanile A., Carlomagno G.M., De Vita A., Donsi G., Massimilla L.,
Scognamiglio A., 1980. Italian Patent N 53499-B/10.
(5) Carlomagno G.M., Festa R., Massimilla L., 1982. "Sulla Distribu-
zione dei Gas di Scarico di un Motore Diesel in Letto Fluido".
Atti VI Congresso Nazionale AIMETA, Genova, in press.
(6) Carlomagno G.M., Festa R., Massimilla L., 1982. "Heat Recovery
from Diesel Exhausts by means of a Fluidized Bed Heat Exchanger".
A.I.Ch.E.75th Annual Meeting, Los Angeles, November, paper 80 d.
(7) Elliot D.E., Healey E.M., Roberts A.G., 1971. "Fluidized Bed Heat
Exchanger", Conference arranged by the Institute of Fuel and the
Institut Francais des Combustibles et de l'Energie, p. 303.
(8) Barone L., 1982. Thesis in Chemical Engineering, University of

Naples.
-128-
en
88!388
( 3)
o8o88388
( r) o o o o o
I s)
( 3)
(
mpcP^.
B.
<J.
Fig. I - Experimental equipment Fig. II - Distributor

A - Fluidized bed heat exchanger: 1) Fluid bed A - Picture of the distributor with Y shaped joint
column, 2) Column bottom plate, 3) Manifold B - Detail of the manifold and nozzles
distributor, 4) Cyclone, 5) One of cooling D 45 mm, P - AO mm, 1 - 12 mm, d - 6,5 mm, OC/- 30
tubes (numbers in brackets indicate the row
number in the array)
Fig.Ill Gas sampling system upstream and downstream the apparatus.
1 Engine. 2 Hydraulic brake. 3 Filter/Exchanger and cyclones. 4 Distri
butor. 5 Sampling probe location. 6 Filter. 7 Filter holder. 8 Sampling
probe location. 9 Gas exhaust pipe. 10 Gas exhaust/air cocurrent heat ex
changer. 11 Hot air generator. 12 Bypass on gas sampling line. 13Filter.
14 Filter holder. 15 Gas sampling pipe downstream the cyclones. 16Water
traps. 17heat exchangers. 18Pump. 19 Temperature display. 20Air flow
rate regulator. 21Rotameter. 22Voltage regulators. 23 Gas volume measu
rator thermocouples.
i 1 i
0.30
H -H*r
Hmf f
A
0.20-
A'
7^ *
0.40
tf
M Silica sand 15Q-350>m
/ A Porous plat
^ O Manifold
0 1 1 1 :
fO 20 30 AU.[cmlstcJ
Fig. IV Comparison between bed expansion with porous plate distributor

and manifoldsiistributor (5);
130
* . / , = 1.89
O Static pritsirt
A A Total prtssin
0.2 0.<r 0.6 0.8 X/Z -f.0
Fig. V Static and total pressure inside the manifolds: 0,A air flow rate
Wo 35 g/sj 0 , A air flow rate Wo = 52 g/s; () from numerical model (5).
'.< A
0
-0.1
*w 4
S?
+0.1
0
t 4
i
-or
o A D * A
Rm/fttn 183 1.83 1.89 178 3.78
Wo [Xl*] 35 +J 52 ia 2Z
1 1 1 1
OA 0.6 0.8 X/Z 1.0
o.z
Fig. VI Nozzle flow rate nonuniformity at different distributor inlet

flow rates and Am/A^n ratios () from numerical model (5).
131
220
/200 r.p.m.
/so
MRMIFOCD
^ ra* &D S0T70H
to J20 ms O
/S
220
SO ms 90
F i g . V I I P r e s s u r e i n s i d e t h e m a n i f o l d and a t t h e bed b o t t o m a t two e n g i

ne r e v o l u t i o n speeds v e r s u s time ( 5 ) .
i i 1 c
C D
180
A A
A A a A
A
II O
e e 8
I R
o 8
A
8
O D
* A
460- 6

A
MO- 1

A ft A

1
i
420-
n 1 1 1
0.25 0.5 075 , Oi
Xfi.
Fig. VIII Bed temperature distributions for different fluidizing veloci

ties ( 6).
132
t(h)
Fig. IX Fouling thermal resistance as a function of time. Defouling obtai^

ned by a step up of tube wall temperature (6).
1 r -|
/*

"I
1 KN
/<! o
4.0-
oc) 0
o oo
0.8 o
o
( CO
o 3 o
U6 o o oo
o
O o
0A- O o &o
o 0
o 0
Oi . <h CD O oo.
6 o
f> 1 1 H 2_
10 ZO
30
'() "
H u . X Fouling thermal resistance as a function of time. Defouling obtaj_
nod by a step down of tube wall temperature (6).
133
to 1
3
O
I
A<A
v l ' 0.8 AA A
O
A AA A AA
0
0
O 0
0.6 O
0 O 0
II
0.4-
O.Z A FLUIDIZED BED HERT EtCHMOEB. -

O SHELL RND TUBE HERT EXCHRNG ER.
1 1
100 zoo 300
Comulatat ive Fuel" Consumption s / I

Heat Transfer Surface
Surfac '
F i g . XI Exchanger e f f i c i e n c y as a function of r a t i o between cumulative

fuel consumption t o h e a t t r a n s f e r surface ( 6 ) .
0,80
6
0*0
I
\
0.40
0,20
IH,r.. O
111 .
*00
XT
Fig. XII Soot removal efficiency and mean soot particle diameter as a
function of bed temperature (8).
134
uo
OS
n-1700
08-
0.4
02
-^w ir 30
Fig. XIII - Soot removal efficiency and mean soot particle diameter as a
function of engine air/fuel ratio ( 8 ) .
1.15 tSi
Fig. XIV - Particle size distribution of soot collected on bed solids(8).
-135-
SIK.^ <
( -60
%Max. fuel
i
injection
i -8 . o 50
30 i 1 i

i
i i 1
20 4-0 6(P*-PSo)[jtPa]
Fig. XV - Influence of the engine overpressure on specific fuel consumption

and mean equivalent pressure at 1200, 1700 and 2200 r.p.m. (5).
DISCUSSION
Question - Mr Robinson, Deborah FLuidized Combustion : What is the

silencing effect of this system on the diesel-engine noise ?
Answer We did no measurements but it centainly equals a silencer.
136-
PART II - DEMONSTRATION PROJECTS
SESSION III - FLUIDIZED BED HEAT EXCHANGERS
Chairman: M. DAVIS
Openinq address
The M.H.O. process for the combustion of shales in fluidized

bed
Combustion of carbonaceous washery shales in fluidized bed
Total precalcination by circulating fluid bed
Fluidized bed boiler used in the production of bricks
Multifuel fluidized bed exploiting indigenous solid fuels
Fluidized bed combustion of waste acid tars
AtmosDheric fluidized bed combustion boiler for electric

power generation by coal
The aoplication of fluidized combustion of coal to a low

cost, large industrial water tube boiler integrated in a
combined heat and power scheme
Use of fluidised bed furnaces in steel heat treatment
Metallurgical treatment (quenching and tempering) in fluidized

bed of track chain elements recovering the residual forging
heat
Closing address
-137-
OPENING ADDRESS
by
Dr. Michael DAVIS
I'd like to welcome you for this second day of our 2-day Contractors
meeting on fluidised beds which, as you know, has been arranged jointly
between DG XII, Research and DG XVII, Energy. The first day was devoted
to research contracts and this second day to demonstration. Now, I hope
that this will give an opportunity to those who are concerned with,
respectively, Research and Demonstration to know what each other is
doing and to establish contacts. This is very important.
I would like to say just a few words about the idea of demonstration.
Often, when research has been successfully completed, it is very
difficult to get the results exploited by industry. Frequently there
are residual risks which deter enterprises from going forward on a new
process or method. The idea of the Community demonstration scheme is
that selected projects benefit from the financial risk being shared
by the Community. If a project turns out to be unsuccessful, then the
amount of loss to an enterprise is correspondingly reduced. If a pro-
ject is successful, a reimbursement of the Community funds is required.
Such a repayment, incidentally, provides a very good inde;, of success.
The contractor for a demonstration project has the obligation to make
available for publication a full final report. These reports ensure
that the results are widely known in the Community, so that others can
benefit from it. This is the demonstrative value. Of course, the
contractor of a successful project has clearly an initial advantage
to exploit the innovation himself with the commercial rights that he
enjoys.
Today we have ten communications, two of them concern fuel combustion

of where the fuel is of low calorific value; two on the electricity
production; two where there is heat recycling. A further two deal with
burning of waste material and the remaining two on metalurgical
treatment.
-139-
THE M.H.O. PROCESS FOR THE COMBUSTION OF SHALES IN FLUIDIZED BED
Authors J. DE KEYSER, R. DENOISEUX
Contract number : EE 027.80
Duration : October 1, 1980 - December 31, 1983
Contractor : Metallurgle Hoboken-Overpelt
Address : A. Grelnerstraat 14
B - 2710 Hoboken
Belgium
Summary
In the present energy context, wastes from coal washing

units present obviously an advantage. Traditional techniques
being not adequate to process such wastes with high ash contents
- often exceeding 82 per cent -, they seem worthless and are,
hence, dumped.
Metallurgie Hoboken-Overpelt (M.H.O.) was able to valorize
these wastes by an original process called "combustion in flui-
dized bed", characterized in that wastes with an average size of
0 to 12 mm do not need crushing or drying, prior to their pro-
cessing. The combustion remains attractive when the user has to
bear the transport charges to its works. The combustion of these
products, despite their very low calorific value of about
3,600 kjoules per kilo and their humidity, enables the production
of nearly one tonne of usable live steam that can be fed to the
electric power plant of the works.
After combustion the ashes contain no more than some 0.4 per
cent of carbon and present a size that is close to the crude
shales before treatment; this makes them perfectly suitable for
use in the production of concrete elements. They can also be
used as cement ingredients.
Gasses freeed during combustion contain but very low dust
quantities and their temperature is reduced to some 110C. Their
release in the atmosphere through a metal stack raises no whatso-
ever environmental problem.
Resume
L'utilisation des rejets de laveries de charbonnages pre-

sente un interet evident dans le contexte Snergetique actuel.
Leur teneur en cendres depasse souvent 82 %, leur enlevant toute
valeur, parce que les techniques conventionnelles ne sont pas
adaptees a de tels produits. lis sont done mis en terril.
Metallurgie Hoboken-Overpelt (M.H.O.) est parvenue a valori-
ser ces rejets par un procede de combustion en lit fluidise, dont
l'originalite repose notamment sur le fait que des rejets de ca-
-140-
llbre 0 & 12 mm sont traites sans broyage ni sechage. Cette ope-
ration reste intressante, merae en prenant en compte le cout du
transport depuis les charbonnages.
La combustion de ces produits permet de produlre prSs d'une
tonne de vapeur vlve utilisable a la centrale electrique autonome
de l'uslne, et ce malgre leur tres faible pouvoir calorifique (de
l'ordre de 3600 kjoules par kilo) et leur humidit.
Les cendres qui ne contiennent plus qu'environ 0,4 X de car-
bone, et dont le calibre est proche des schistes crus avant trai-
tement, sont utillsables pour la fabrication d'Slements en be ton.
Ellea peuvent egalement servir d'ingredient en cimenterie.
Les fumees liberies par la combustion ne contiennent que de
tres faibles quantites de poussieres et leur temperature est
abaissee a environ 110C. Leur evacuation 3 1'atmosphere par une
cherainee metallique ne pose aucun problerae d'ordre ecologique.
Ml
1.1 Introduction
Metallurgle Hoboken-Overpelt (M.H.O.) est generalement consideree

comme un des leaders mondiaux de la metallurgie des metaux non-fer-
reux.
Ses besoins en energie sont Sieves et representent, pour 1'ensem-
ble des ses trois usines, une depense annuelle de l'ordre de 2,5 mil-
liards de francs.
Elle a done ete amenee, depuis longtemps deja, a se pencher avec
attention sur les diverses procedes susceptibles d'economiser de
1'energie. D'ou l'interet qu'elle porte a la combustion de schistes
et autres produits auxquels leur tres bas pouvoir calorifique enleve
toute valeur.
La technique des lits fluidises y est appliquee depuis plus de 30
ans : le premier usage qui en fut fait portait, et porte encore d'ail-
leurs, sur l1elimination du soufre contenu dans les concentres utili-
ses pour l'approvisionnement des usines. Celle-ci se fait par un
traitement de grillage comportant la recuperation de la chaleur libe-
rie par la reaction.
Rappelons que les concentres fournis par la flottation des mine-
rals, sont des matieres tres fines < 100,0., qui, pour des raisons me-
tallurgiques, sont pelletisees avant traitement.
Ayant maitrise la technique des lits fluidises sur des materiaux
pelletises, M.H.O. a pousse l'application S d'autres materiaux simi-
laires en granulometrie et en pouvoir calorifique (ordre de 1050 kilo-
calories/kilo soit 4300 kilojoule).
L'objet du present expose est de decrire le procede utilisS
actuellement a l'usine d'Overpelt, dans 1*installation de
demonstration C.E.E. 027, projet de 1980, et d'indiquer les resultats
deja acquis.
Pour des raisons budgetaires evidentes, une part appreciable de
l'ancienne installation a ete reprise, mais adaptee aux exigences du
jour. Certains de ses anciens elements, qui sont devenus inutiles,
auraient dti disparaltre mais leur suppression eut ete fort couteuse.
Leur presence pose, actuellement, quelques difficultes d'exploita-
tion.
Outre les economies intrinseques d'energie que 1'installation
permet de realiser, les 6 points suivants ont ete pris en
consideration pour l'etude du projet C.E.E. :
1) accroissement de la production de vapeur.
2) amelioration du rendement thermique.
3) traitement de produits encore plus pauvres que par le passe.
4) captage de l'SO- et l'SO, liberes par la combustion.
5) depoussierage pousse des fumees apres chaudiere.
6) tests de matieres diverses pour des amateurs potentiels du
procedg.
1.2 Description sommaire de 1'installation
Schematiquement, 1'installation se presente comme le montre la

figure I en annexe, tandis que la figure 2 est un exemple de realisa-
tion d'un echangeur de chaleur amont a installer au point 6 de la
figure 1.
142-
L'installation existante a et6 completee par" les elements
sulvants :
- les echangeurs de chaleur apres chaudiere.
- les filtres 3 poussieres.
- 1'Injection de chaux pulverulente pour capter S0 ? et SO,
apres la chaudiere, et abaisser le point de rosee des fumees.
- la salle de contrSle.
- les elements de chaudiere du lit.
- la manutention amont et aval, ainsi que les silos & cendres
calcineea.
De plus, 1'alimentation en eau de la chaudiere fortement agrandie
a neceasite le remplacement du dispositif automatique ancien par un
nouveau susceptible d'un plus fort debit.
Le four est caracterise par une section horizontale de 7 m ,
mesuree au niveau de la grille, et une hauteur libre de 4.6 m au-
dessus du lit fluidise maintenu a une epaisseur de + 1 m.
La charge de base de 1'installation (le combustible) est consti-
tute par du schiste de calibre compris entre 0 et 12 mm, en provenance
directe de la laverie des charbonnages de Waterschei et traite tel
qu'il est recu. II est charge humide a 10 - 12 Z d'H-O par une
chute laterale (figure 2 ) .
Les cendres sont extraites en continu au niveau de la grille par
un extracteur mecanique; ceci permet d'eviter 1'accumulation de gros-
ses partlcules dans les couches inferieures du lit.
L'air de fluidisation est divise en deux circuits, chacun rece-
vant a peu pres la moitie du d^bit total. Chaque circuit est pourvu
d'un recuperateur de chaleur independant : celui d'amont preleve sa
chaleur sur les cendres du combustible, et le second, a l'aval de la
chaudiere, sur les fumees avant leur arrivee a la cheminee.
Une chaudiere du type Laraont et des tubes en epingle installed
dans le lit alimentent une turbine a contre-pression d'une puissance
electrique de 3.3 MW. La vapeur produite a 42 bars et 410C/415C y
est detendue a 3.5 bars pour assurer les besoins de chauffage de
l'uslne a zinc.
1.3 Commentaires sur les 6 points pris en consideration pour l'gtude
1) Acroiseinent de_la rdctionjte vapeur

La technique de placement d'Elements vaporisateurs dans le
lit fluidise est bien connue dans le secteur des metaux non-fer-
reux et plusieurs chaudieres fonctionnent selon ce systeme. Aussi
l'anclenne chaudiere a-t-elle ete completee par un jeu de tubes
dans le lit.
Des problemes de mise au point du nouveau circuit d'alimenta-
tion en eau obligent de llmiter temporairement la production de
vapeur de 1'Installation a 10 tonnes par heure.
L'equilibre du bilan thermique du lit fluidise, ou la tempe-
rature ne peut pas dSpasser le point de fusion des cendres, ne
peut Stre maintenu actuellement que moyennant un trop fort exces
d'air.
Celui-ci sera raraene & une valeur plus raisonnable (15 3
25 X) des que les elements de chaudiere places dans le lit seront
en fonctionneraent regulier. A ce moment, en effet, 1'exces de ca-
lories sera absorbe in situ, la surface de chauffe des elements en
epingle permettant d'assurer 45 Z de la vaporisation totale de la
chaudiere.
-143-
2) Ameliôation_du rendement thermiU
Far le passe, la temperature des fumees a la sortie de la

chaudiere, etait abaissee a + 230C.
Actuellement, grace aux echangeurs places apr&s chaudiere,
cette temperature est ramenee a 110C. La perte de chaleur a la
cheminee est done deja pratiquement divisee par deux. Grace a la
reduction de l'exces d'air, elle sera a nouveau divisee par 1,8
lorque les elements vaporisateurs dans le lit pourront fonctionner
normalement.
3) TraJLteme_nt_de^ rd_uits_encore_ ^lus_pauvres_qîe_pa^r_le_ j>as
Autrefois les schistes recus etaient ceux rejetSs par des

lavoirs de Houthalen, puis Beringen et Zolder. Un kilo de produit
sec avait un pouvoir calorifique de 1.000 a 1.100 kilocalories ou
4.200 a 4.500 kilojoules.
Actuellement, en vertu de nos accords avec K.S. (Kerapense
Steenkolenmijnen), e'est la mine de Waterschei, dont la laverie
est des plus modernes, quiapprovisionne Overpelt : le produit a
frequemment plus de 83 % de cendres (ecarts compris entre 80 et
86 %) et son pouvoir calorifique, par kilo de matiere seche, tombe
a 700 900 kilocalories (ou 2.900 a 3.800 kilojoules). Son hu
midite moyenne est de 10 a 12 % mais elle peut atteindre 15/18 %,
d'apres les conditions atmospheriques.
Les calories recuperees dans les fumees aprs chaudiere, as
surent le prechauffage complementaire de l'air de combustion n
cessaire.au traitement de ce produit vraiment tres pauvre. La
temperature de l'air injecte depasse en moyenne les 300C.
II est possible non seulement de traiter des produits trSs
riches en cendres, mais aussi d'obtenir des cendres dont la teneur
en carbone residuel n'excede guere 0,3 %, preuve d'un excellent
rendement de combustion.
Les cendres extraites du lit passent directement dans
l'echangeur de chaleur install^ sous la grille de fluidisation oil
elles sont refroidles jusqu'a + 55C, avec bien entendu comme co
rollaire le prechauffage de l'air de fluidisation, l'echangeur
ayant un rendement thermique superieur a 95 %.
4) atg d^'S02_ et S03
L'amelioration du rendement thermique obtenue par un abaisse

ment de la temperature des fumees a 1'entree de la cheminee,
aurait pu crSer des difficultes, dues a 1'apparition de condensa
tions acides, dans le systeme de depoussierage des fumees apres
chaudiere. Pour les eviter, il a ete necessaire d'abaisser le
point de rosee des fumees par elimination aussi poussee que pos
sible par des anhydrides de soufre, S0, et surtout SO,.
Thôriquement 1'installation permet de fixer ces anhydrides
en deux endroits :
a) dans le lit fluidise sur la matiere qui y est en
fluidisation.
b) entre la sortie de la chaudiere et 1*entrSe de l'Schangeur
de chaleur aval.
144
a) Dans le lit fluidisg
Lea schlstes calclnSs contiennent, selon leur origine, de

3 a 6 X CaO + MgO : 11 y a done 13 une source "naturelle" de
neutralisation. De plus, vu la granuloraetrie de 0/12 mm du
produit traite, environ 80 X de celui-cl est extralt dlrecte-
ment du lit, apres un temps de sejour moyen d'l/2 h & 3/4 h.
Le rapport stoechlometrique depasse largement les 2 a 2,5 con-
siders corame necessaires pour une bonne fixation du soufre.
Avec les schlstes, 11 n'a pas encore ete necessaire jusqu'icl
d'ajouter, selon une technique connue, du calcaire ou de la
dolomie dans le lit.
b) Dans le circuit des fumees
L'amelloratlon du rendement thermique global, citee plus

haut, a ete obtenue en abaissant la temperature des fumees 9
1'entree de la chemlnee jusqu'environ 110C par le placement de
l'echangeur de chaleur apres chaudiere.
C'est cet echangeur qu'11 y avait lieu de proteger contre
la condensation des fumees en velllant a abaisser leur point de
rosee par elimination poussee du SO,.
En effet dans la chaudiere, la reaction exothermique
SO, + 1/2 0, ^ . SO, a tendance 1 evoluer vers la for-
mation d'SO, puisque les fumees y sont en cours de refroidis-
sement. Corarae cette evolution est favorisee par la presence du
metal des tubes de la chaudiere qui constitue une masse irapor-
tante de catalyseur pour cette reaction, c'itait done surtout
le SO, qu'll importait de fixer pour abaisser le point de
rosee.
Cecl a ete realise a plus de 85 X par injection de chaux
pulverulente dans une tuyere orientee a contre-courant a la
sortie de la chaudiere. Cette injection est assuree par un
tres leger dibit d'air qui repartit en une heure une addition
de chaux de 40 kg dans un debit de gaz de 20.000 Nn .
II eOt ete lllusoire de compter ici sur le CaO et le MgO
des poussieres emportees pour fixer les anhydrides de soufre,
le temps de contact etant trop court pour ces reactifs I action
relativement lente.
Des mesures seront faites ulterleurement vlsant & mieux
preclser quelle fraction du SO. a ete transformee en SO,,
et a reduire si possible l'addltion de chaux sans sacrifier
pour autant le rendement d'elimination du SO,.
5) Deoîsî^ra_ge_ ^ e ^ fumees apres chaudiere
L'abalssement de la temperature des fumees a l'entr&e de

la chemlnee et son corollaire obligatoire, l'abalssement du
point de rosee de celles-cl ont permis 1'utilisation de flitres
a poussieres a manches rSalisees en dralon. Les poussieres qui
n'ont pas ete deposees naturelleraent dans la chaudiere, sont
captees dans deux flitres fonctionnant en parallele, et les fu-
mees a la chemlnee en contiendront moins de 10 mg/Nm lorsque
la raise au point du flitre sera achevee.
-145-
6) TejL'iL JLU. 5atiS^rs_aiite <jue_ les 5.ch,ltes du l.avir ^ e
Waterschel
Grace a une automatisation poussee, les interventions ma-

nuelleS ont ete reduites au strict minimum. En dehors des ma-
nutentions sur les plaines de stockage et des operations d'en-
tretien, 1'unite entiere fonctionne avec un seul operateur.
Comme parametres d'asservissement, 11 faut noter la modulation
du debit de combustible qui regie la temperature de combustion
et la perte de charge du lit qui est maintenue constante par
adaptation automatique de la vitesse d'extraction des cendres.
Cette simplicity apparente n'empeche pas que de grands
ecarts de comportement existent entre des schistes de provenan-
ces diverses, d'une part, et les combustibles de nature diffe-
rente, d'autre part. Parmi ces derniers il y a lieu de ranger
les lignites, les tourbes, les dechets domestiques etc.
De tels ecarts peuvent etre corriges par :
- la modification des plages de reglage des facteurs
prScites.
- la modification du positionnement des elements
surchauffeurs dans la chaudiere.
- 1'adaptation de la granulometrie du produit a traiter.
- le recyclage eventuel de residus calibres.
- la modification du debit d'air souffle par unite de
surface de grille.
Tout ceci constitue bien entendu un ensemble de donnees
experimentales, et des unites nouvelles ne peuvent etre concues
qu'a la lumiere de tests a l'echelle industrielle. Certains
ont deja ete faits et d'autres sont prevus pour les prochains
mois.
XI va de soi que la taille, la souplesse et le degre de
perfectionnement de l1installation de M.H.O. 3 Overpelt repon-
dent parfaitement a cette fin, les tests sur des matieres spe-
ciales etant un des objectifs importants sinon prioritaires de
la demonstration.
Conclusions
Le four a schistes installe a l'usine a zinc de M.H.O. a Overpelt

fonctionne selon un procede original. Concu a l'origine pour les seuls
besoins metallurgiques de M.H.O., il a pu, grace a l'appui de le
C.E.E., Stre adapte a d'autres fins. II permet la combustion de pro-
duits a tres bas po.uvoir calorifique, tels que les schistes et d'autres
combustibles qu'il est pratiquement impossible de traiter par les pro-
cedes conventionnels.
La capacite de 1'installation d'Overpelt a ete volontairement li-
mitee a la saturation de la centrale electrique de l'usine. II est
possible toutefois de prevoir des unites plus vastes, puisque la con-
ception des appareils utilises se prete a des augmentations d'echelle.
Les possibilites d'extension ne sont toutefois pas illimitees. II
faudra done determiner un module optimum qui permettra par la suite et
par simple juxtaposition de modules identiques de construire un com-
plexe de grande taille.
146-
METALLURGIE HOBOKEN-OVERPELT
Combustion de schistes, schema de principe
1 Tr ernes a co*bu*1M e$
2 Exlraclrurs doseurs
3 Almentateur dc la charge
4 Lit flutdise de coM bustmn
5 Lit flwdue dc rechauffage
dc I'air dc combustion
6 Prechauffeur d * r aownt
7 Chaudtere
6 C|done I eventuel I
9 Oepoussiereur
Q W Prechauffeur d'air aval
11 Venhlateur detraction
12 themmee
13 Ventrlateurs d'ar de fluidisation
K Transporters a cendres
FIGURE I
METALLURGIE HOBOKEN - OVERPELT
SCHEMA DU FOUR ET DU SYSTEM E

DE RECUPERATION DE CHALEUR
CQMSUSTI1LCS
FIGURE I I
PÂAAA/W*f
CALC NC FROIO
-147-
DISCUSSION
Question - Mr. Sirchis, CEC :
Vous avez parte de difficultes d'immerger des tubes a eau
dans Le lit. Pourriez-vous nous dire en quoi elles consistent?
Answer Les difficultes resident dans le systeme d'alimentation en

eau. Nous avons prevu une alimentation automatique et c'est
elle qui ne fonctionne pas comme prevu. De toute facon ce
n'est qu'une question de semaines et si le constructeur ne
parvient pas a regler la question, nous le ferons nous-memes.
Question - Mr. Ehringer, CEC :

Mr. De Keyzer, vous avez dit que les autres schistes sont
utilises eventuellement dans la cimenterie. Est-ce que vous
avez une experience dans ce domaine ou est-ce que c'est une
supposition?
J'ai pose cette question parce que dans le cadre du program-
me de recherche sur.les economies d'energie, nous avons des
programmes ou des projets qui etudient en particulier I'uti-
lisation des cendres de hauts fcurneaux dans les cimenteries.
Les travaux sont en cours et je crois utile de vous en infor-
mer.
Answer Tous les schistes sont differents. Leur composition et Leur

comportement different. Si les schistes conviennent pour etre
brules, il faut voir si la composition convient pour la ci-
menterie, mais il y a des cas en Belgique qui sont positifs a
ce sujet.
Les scories des hauts fourneaux peuvent egalement etre uti-
lisees et en plus elles ont la particularity de colorer le
ciment. II n 1 entre pas dans nos intentions a Metallurgie
Hoboken de nous lancer dans telles recherches mais nous som-
mes tout a fait disposes a eventuellement faire un essai sur
les schistes. II suffit d'avoir un millier de tonnes de
schistes pour faire un essai d'une duree de 3 a 4 jours.
Question - Mr. Willis, N.C.B. :

I would like to put a question on the economy. I would like
to know if studies have been done to find out if there is a
certain optimum with respect to the purification of these
schists in advance to the process of combustion?
Answer Les schistes sont des matieres de peu de valeur. Les frais
les plus importants sont les frais operatoires et le trans-
port. Nous considerons qu'on ne peui pas transporter les
schistes sur une distance de plus de lOOkn par exemple. II
faut implanter le four a schistes au centre d'une source
d'alimentation. La valeur des cendres est un element tout a
fait negligeable. C'est surtout la calorie qui doit payer le
procede.
148-
COMBUSTION OF CARBONACEOUS WASHERY SHALES IN FLUIDIZED BED
Author : R. PUFF
Contract Number EE-037-81
Duration : 24 mois
Head of project S. DELESSARD
Contractor : CERCHAR
Address : Station Experimentale de Marienau

F-57600 FOR B A C H
SUMMARY :
The fluidized bed combustion of low calorific value fuels such as washery
and flottation carbonaceous shales or oil shales had been studied long
years ago in France at the Centre d'Etudes et Recherches des Charbonnages
de France (CERCHAR), which had acquired an important theoretical and
practical experience.
A process, characterized by the use of a grid of original design and

the procedure of plant control and operation, has been studied.
Combustion tests with a variety of carbonaceous and oil shales have

been carried out on pilot units operated at the "Station experimentale
de Marienau" of the CERCHAR and promising results have been obtained.
Following these encouraging tests, CHARBONNAGES DE FRANCE have decided

to construct a unit with a grid of 18 m2 cross-section within the
Drocourt washery where it will burn carbonaceous shales of net calorific
value of 2.500 to 5.000 kJAg to produce from 15 to 20 t steam/hr
(30 bars 300 * C) to be fed to a coking plant and chemical works.
This project has obtained a financial aid of the European Communities.
The design study has been committed to CREUSOT-LOIRE and construction

will start early in 1983; it will come into operation in the beginning
of 1984.
-149-
I. INTRODUCTION
La technique de combustion en lit fluidise fait actuellement l'objet

d1 efforts importants de recherche, notamment appliquee a la combustion
de charbons, en particulier en GrandeBretayne, en Republique Federale
d'Allemagne et aux EtatsUnis.
Cela etant, la combustion de combustibles a faible pourvoir calori

fique, tels que schistes houillers de lavoir et de flottation ou
schistes bitumineux, a ete etudiee depuis de nombreuses annees en
France au CERCHAR, qui a acquis une experience theorique et pratique
(voir references biblioyraphiques en annexe) :
dans le domaine de la caracterisation et de la combustion des

charbons, des schistes et des dechets industriels divers
dans les etudes sur maquettes chaudes de la fluidisation a pression

atmospherique (0,2 et 0,8 m2 de section de grille), avec applica
tion
. a 1'incineration des boues et combustibles a faible pouvoir

calorifique, notamment schistes houillers, bitumieux et dechets
industriels divers
. au sechage et prechauffage des charbons
dans les applications industrielles de la fluidisation a la pression

atmospherique (procedes et materiels brevetes) :
2
. a 1'incineration des boues (appareils en service de 9 m
de section)
. au sechage et au prechauffage du charbon (appareils en service
de 100 t/heure procede COALTEK).
Les Charbonnages de France ont decide de construire une installation

dans le but de demontrer la possibility de bruler des schistes de
lavoir de tres faible pouvoir calorifique (2,5 a 5 MJ/kg), destinee
a assurer la production de vapeur (~100.000 a 150.000 t/an), rem
placant en tout ou partie celle produite actuellement par la com
bustion de gaz de fours a coke.
Le gaz ainsi deplace pourra trouver une meilleure valorisation en

centrale thermique, en remplacement du fioul de soutien ou d'appoint
ou dans 1'industrie chimique pour la synthese.
150
Che implantation sur le site des Usines de DROCOURT des Houilleres
du Bassin du Nerd et du Pas-de-Calais (HBNPC) a ete retenue pour la
realisation de ce projet.
2. DESCRIPTION DU PROJET
Le CERCHAR a acquis, a partir d'etudes sur la fabrication de semi-

coke par pyrolyse du charbon, un tres important savoir-faire en
matiere de fluidisation.
Ce savoir-faire a ete valorise dans le domaine de l'incineration et

de la calcination; plusieurs brevets ont ete pris, portant princi-
palement sur la grille "a pyramides et a siphons". Des fluidiseurs
ont ete vendus sous licence CERCHAR, en particulier 2 calcinateurs
de 3,4 m de diametre, installes dans la raffinerie de BRAZI
(Roumanie).
Pendant la periode ecoulee, le CERCHAR a eu 1'occasion de prouver

qu'il etait possible de bruler correctement des combustibles tres
cendreux - schistes de 1avoir et schistes bitumieux - dans des
foyers de ce type; des essais comparatifs avec d'autres techniques
conduisent affirmer que le foyer a lit fluidise est le mieux
adapte a ce yenre de produits; les essais en lit fluidise ont ete
effectues, soit sur une petite unite de 0,8 nr de section, soit sur
un pilote de 0,2 m2.
Les Charbonnages de France ont done decide, a la suite de ces essais

encourageants, de realiser une unite de demonstration repondant aux
besoins d'un site industriel.
2.1 Presentation du site
La cokerie de DROCOURT prepare actuellement de la vapeur

(30 bars - 300 *C) par combustion de gaz de cokerie dans trois
chaudieres a charbon a grille cheminante, modifiees, ayant
chacune une capacity de production de 20 t/h.
Une seule chaudiere assure la totalite de la production pendant

que la seconde est en reserve et la troisieme en revision.
Sur le site de la cokerie de DROCOURT se trouve egalement un

lavoir place, par rapport a la cokerie, a 1'oppose de la chauf-
fer le, a 500 m environ a vol d'oiseau.
-151-
Ce lavoir peut traiter environ 12.000 t/j de charbon brut et
produire 6.000 t/j de schistes, se dcomposant de la faon
suivante :
Schistes Teneurs en
Proportions cendres
< 0,5 mm (flottation) ~10 % 72 %

0,5 - 10 mm (fines) ~24 % ~83 %
10 150 mm (grains) <~66 % ~88 %
Ces schistes sont mis sur terril ou en lagune.
2.2 Presentation du projet
Les installations existantes sont concues, soit pour bruler des

schistes de lavoir ou bitumieux relativement riches (PCI ^.
4,5 MJ/kg), soigneusement calibres, soit pour bruler dans un
lit fluidise auxiliaire des schistes de flottation de grande
finesse (<. 200^) et de pouvoir calorifique eleve (PCI sur sec
de 7 a 10 MJ/kg, H,0 ^ 50 %).
L'installation propose sera affecte a la combustion d'une

large gamme de schistes de lavoir simplement concasses (0-10,
0-15 mm) sans preparation particuliere (ni broyage, ni homo-
geneisation) et de pouvoir calorifique faible, compris entre
2,5 et 5 MJAg, impossibles a bruler par toute autre technique.
L'installation comprendra :
2
- le calcinateur a lit fluidise, de diametre 4,8 m (18 m de
section de grille) et ses cyclones depoussiereurs
- la chaudiere de recuperation de la chaleur sensible
des fumees
- le depoussierage final.
Uhe unit de rcupration de la chaleur sensible des cendres lui

sera adjointe; les calculs et essais, faits au stade du pilote,
ont en effet montr tout l'intrt de cette rcupration pour
les schistes bas PCI, dont par nature la teneur en cendres
est leve.
152-
Cctte installation ne sera pas une "installation coumerciale"
mais un prototype de mise au point a 1'echelle industrielle
d'un procede etudie sur pilote.
Elle necessitera vraisemblablement des adaptations, c'est pour-

quoi elle fonctionnera, dans un premier temps, en parallele avec
les chaudieres existantes qui pourront assurer, le cas echeant,
la releve.
2.3 Choix de 1'implantation de 1'installation
II etait possible d'implanter 1'installation, soit pres du

lavoir, soit pres de la chaufferie.
Le site du lavoir a ete retenu pour les raisons suivantes :
a) la place est suffisante
b) les silos a charbon et a schistes, ainsi que les installations

de manutention existent. II suffit de les renover et de les
completer pour alimenter le calcinateur
c) les schistes calcines pourront etre mis directement au

terril par les moyens existants en service actuellement
pour les schistes bruts
d) le cout des liaisons eau et vapeur avec la chaufferie a gaz

est inferieur au cout d'installation d'un stockage adapte
au niveau de cette chaufferie
e) le transport des schistes bruts du lavoir vers la chaufferie

est ainsi evite, de meme que celui des schistes calcines
entre la chaufferie et le terril
DESCRIPTION DU PROGRAMME DE TRAVAIL
3.1 Objeetifs de la demonstration
La conception originale de la grille de fluidisation et le mode

de reglaqe automatique du fluidiseur permettent de bien mai-
triser la fluidisation, et done le regime thermique, aussi
bien pendant les operations de mise en route et d'arret que
dans les conditions normales de marche. Les problernes sont
d'autant plus difficiles a resoudre que la section du fluidiseur
est grande et qu'il s'agit de tralter des produits sensibles
aux prises en masse conme les schistes houillers.
153-
Les objectifs de la demonstration sont :
- demontrer, au niveau industriel, la inise en fluidisation et

la combustion de schistes grossiers et de faible pouvoir
calorifique
- optimiser et simplifier les composants de 1'unite, notamment

la grille de fluidisation
- etudier la souplesse de fonctionnement
- mettre au point un mode de conduite automatique, y compris

pour le demarrage et 1'arret de 1'installation
- optimiser la recuperation des chaleurs sensibles par re-

froidissement des cendres extraites du lit fluidise
Le but de l'ensemble du projet est de maitriser ces techniques

pour des installations importantes, susceptibles de fournir
des puissances suprieures 30 MW.
Par ailleurs, il est important de montrer la possibility d'uti-

liser, dans de telles installations, des produits standards non
particulierement prepares et d'analyser les caracteristigues de
fonctionnement en vue d'aboutir a une definition d'equipements
appropries.
3.2 Deroulement chronoloqique du projet
Les travaux sont prevus selon un calendrier s'etalant sur

trois annees et demie :
Etudes et construction
Debut : l61" Janvier 1982

Fin : 30 septembre 1983
Duree :21 mois
Campagnes d'essais et mesures
Debut : l ^ octobre 1983

Fin : 30 juin 1983
Duree : 21 mois
154-
DESCRIPTION DU PROCEDE
4.1 Fonctionnement du procede
L'installation sera concue sur la base d'un reacteur de fluidi-

sation, equipe d'une grille CERCHAR "pyramides et siphons" de 18
L'installation doit, en principe, traiter des schistes 0,5-10 m

directement preleves a la sortie du lavoir sans broyage prealabl
Ces produits auront une teneur en cendres comprise entre 86,5 %
(2,5 MJ/ky sec) et 80 % (5 MJ/kg sec) et une humidit de l'ordre
de 8 %.
Selon le pouvoir calorifique de schistes et compte tenu des con-

ditions operatoires retenues :
Temperature de combustion : 850 - 950 " C

Vitesse de fluidisation : 2,0-2,5 m/s
Exces d'air : 20 - 25 %
le debit de schistes bruts sera compris entre 12 et 28 t/h.
La figure 1 presente un schema de principe de 1'installation.
Les schistes houillers 0-10 mm sont diriges vers deux tremies

de stockage de 500 tonnes.
S'lls sont lnsuffisamment riches ou trop humides pour assurer

une combustion autothermique, il est possible de leur incorporer
des mixtes de lavoir (charbon a environ 50 - 60 % de cendres*.
Ayres deferraillage, les schistes passent au travers d'un crible

d^ protection avant d'etre delivres dans le calcinateur au moyen
d'une double vis a ruban a vitesse variable disposee au centre
de la voute du four.
Compte tenu du dclin de la production de charbons du bassin

Nord - Pas-de-Calais, la production du lavoir de DROCOURT en
fines 0,5-10 mm risque de devenir insuffisante pour alimenter
le calcinateur; un poste de concassaye des produits 10 -150 m
est done necessaire. Au besoin, il sera possible d'enrichir les
schistes en deplacant le seuil de coupure au lavoir, ou en leur
ajoutant des "mixtes de lavoir", produits de pouvoir calorifique
super ieur 6 MJ/kg.
-135-
Le calcinateur, d'une hauteur de 9 m, est revetu interieurement
d'une double couche d'isolant et de refractaire, de qualite
anti-abrasion dans sa partie inferieure. Son diametre est de 6 m
dans sa partie superieure (zone de revanche) et de 4,80 m au ni-
veau' du lit fluidise.
2
La grille de fluidisation de diametre 4,80 m (18 m ) est consti-
tuee de 61 pyramides a base hexagonale a siphons alimentes par
un ventilateur de 37.000 m 3 N/h sous 3.000 mm CE.
Deux bruleurs de 10 kj/h, fonctionnant au grisou (59 % de methane)

facilitent les demarrages. 31 pyramides sont alimentees en gaz
tandis que les 30 autres sont alimentees en eau.
Les cendres grossieres extraites en continu du lit fluidise par

un col de debordement lateral pourront etre refroidies a 200 C
dans un tube tournant de-3 m de diametre et 18,5 m de long, pro-
duisant de 1 'air chaud a 800 C.
Les fumees issues du calcinateur sont depoussierees dans deux

cyclones en parallele de 2,50 m de diametre et de 8,40 m de
hauteur. Les cendres fines seront refroidies a l'eau.
Un dispositif plus performant permettant la recuperation de la

chaleur sensible des cendres grossieres et fines est a 1'etude.
2
the chaudiere de 36 m de section et de 20 m de hauteur, equipee
d'un dispositif de grenaillage pour le nettoyage des tubes, pro-
duit de la vapeur 30 bars - 300 C a partir des gaz issus du
calcinateur et du dispositif de recuperation sur cendres gros-
sieres et eventuellement fines.
Les fumees subissent un depoussierage final dans un filtre a

manches en feutre de teflon (1.130 m^ de surface filtrante) avant
d'etre rejetees a 1'atmosphere via un ventilateur exhausteur et
une cheminee de 45 m de hauteur.
Les cendres refroidies sont collectees et dirigees vers le terril.
La marche de 1'installation est controlee et commandee par un

ensemble a base d'automates programmables avec ecrans de controle.
-156-
4.2 Production de vapeur
Le tableau ci-dessous donne la production de vapeur 30 bars -

300 C en tonnes/heures selon la Vitesse de fluidisation et la
qualite des schistes et ce pour une temperature de combustion
de 900 "C (voir figure 2 ) .
Recuperation Sur Sur cendres grossieres

cenclres grossieres et fines
^\^ Vitesse 2,0 2,5 2.0 2,5
^"X. m/s
Schistes^s.
MJAg ^ \
2.5 12,5 15,4 14,0 17,5

3,2 12,4 15,2 13,6 17,0
5,0 11,6 14,2 12,3 15,3
Deux cas sont distingues
- Recuperation de chaleur sur les seules cendres grossieres

extraites du lit fluidise (representant de 13 a 26 % de la
chaleur non recuperable) et sur les fumees de combustion.
La chaleur sensible des cendres fines (2,7 a 7,6.%) n'etant
pas recuperee.
- Recuperation sur la totalite des cendres.
La figure 3 donne a titre d'exemple le bilan thermique et mas-

sique etabli pour un schiste a 3,75 MJAg sec, avec recuperation
sur les seules cendres grossieres.
4.3 Programme de mesure
L'installation sera equipee de moyens de mesure des debits mas-

siques et gazeux indispensables, des moyens de mesure des tem-
peratures et de pression, ainsi que d'appareils d'analyse auto-
matique des gaz (notamment C02, CO, O2, SO2) et de controle des
rejets de particules (opacimetre).
157-
Les caracteristiques des produits seront etablies en laboratoire.
Des mesures ponctuelles seront realisees pour repondre a des

points particuliers, par exemple : rejet de poussieres par
sondage isocinetiques.
Ces differentes mesures permettront de controler les performances

de 1'unite, notamment :
- efficacite de combustion
- rendement thermique
- respect de 1'environnement, etc.
5. CONCLUSIONS
Des essais sur maquettes chaudes ont determine les Charbonnages de

France a construire, avec l'aide financiere des Communautes Euro-
peennes, une unite de demonstration qui brulera les schistes 0 -10 mm
du lavoir de DROCOURT (Houiilieres du Bassin du Nord - Pas-de-Calais)
pour produire la vapeur necessaire a la cokerie et au complexe chi-
mique voisins (12 a 17 t/h de vapeur 30 bars - 300 C).
L'economie d'energie attendue sera superieure a 10.000 TEP/an en

marche continue avec recuperation de la chaleur sensible de la totalite
des cendres et production d1environ 17 t/h de vapeur pour un fonction-
nement de 8.000 heures/an.
Le gaz de cokerie ainsi deplace pourra trouver une meilleure valorisa-

tion en centrale thermique en remplacement du fioul de soutien ou
d'appoint ou dans l'industrie chimique pour la synthese.
Ce projet permettra de mettre en evidence les potentialibilites de la

grille CERCHAR (souplesse de fonctionnement, regulation, possibilites
d1 arrets et de demarrage) ainsi que la faisabilite industrielle d'uni-
tes de grande taille fonctionnant de facon automatique.
II est etroitrement lie au developpement de chaudieres a lit fluidise

de cendres seches pour la combustion de charbons cendreux et soufres
entrepris par le CERCHAR, CREUSOT-LOIRE et l'INSTITUT FRANCAIS DU
PETROLE, avec l'appui financier du Ministere de l'industrie. En effet,
ces organismes se proposent d'etudier les structures optimales de chau-
dieres a lit dense et construisent, dans un premier temps, sur la
Plate-Forme Nationale d'Essais des Charbons de Mazingarbe (Pas-de-Calais),
un prototype a lit dense de 3 m2 de section de grille.
-158-
BIBLIOGRAPHIE
P . FOCH e t J . GEOFFROY
Developpement de l a c a r b o n i s a t i o n par f l u i d i s a t i o n a l a S t a t i o n
experimentale de Marienau
31ne Coiigres I n t e r n a t i o n a l de C h i m e I n d u s t r i e l l e
LIEGE, Septembre 1958
S . DELESSARD e t R. PUFF
Applications i n d u s t r i e l l e s da t r a i t e m e n t en l i t f l u i d i s e
Sechagebroyaye de materiaux en g r a i n s e t c a l c i n a t i o n de
combustibles pauvres
lOeme Journes d ' Etudes s u r l e s Flammes
Mars 1980
R. MARCELLINI e t S . DELESSARD
La combustion en couche f l u i d i s e des boues d ' e p u r a t i o n
Journee de 1'Eau
PARIS, Decembre 1970
R. CHAUVIN
Les s c h i s t e s h o u i l l e r s , source de materiaux pour l a
c o n s t r u c t i o n e t l e genie c i v i l
CCE Journees d ' i n f o r m a t i o n
LUXEMBOURG, Septembre 1979
P . DUMOUTET.et S . DELESSARD
Combustion de s c h i s t e s bitumieux l o r r a i n s en l i t fluidise
9eme Journees d'Etudes sur l e s Flammes
Octobre 1977
S . DELESSARD, R. PUFF, J . C . K ITA

V a l o r i s a t i o n de s c h i s t e s h o u i l l e r s
Journees europeennes sur l a F l u i d i s a t i o n
INSTITUT de Gnie Chimique
TOULOUSE, Septembre 1981
S . DELESSARD, R. PUFF, J . C . K ITA

Combustion of carboneous s h a l e s by t h e CERCHAR FBC p r o c e s s
The DROCOURT P l a n t
The Seventh I n t e r n a t i o n a l Conference on F l u i d i z e d Bed Combustion
PHILADELPIA, USA, October 1982
159
l . C a l e i n a t e u r a l i t riuldiBo
2 . Cyclones 10. Bandea peseuaea
3. Chaudlere de re'cupe'ration 11. Mferreillear
ft. Plltre a mancha 12. Trlmle & achlatea ealcinea
5 . Cheminle
6 . Tube t o u m a n t r a f r o i d l a a e u r
7. Vis refroldlaseuse
8 . Tre'miea a a c h l s t e s
9. Mlxtes
FIG. 1 Schema d!ensemble de 1'unite de demonstration de DROCOURT
Vapour J O b*ra/3O0*C
tonnea/heure
I
autotharwlcite"
Injection dc gaa I Injection d'eau
ou de nlxtea . 5 c h l B t e , 29.9 t/h
Joax 580 "3"/"
HLrteax *
Sehiatta" 16.2 t/h
9.5 /
* I I a'aglt de* tonnage de eohletea brut* a 8 % d'eau

(S) Ht'oUpe'rateur aur totallt< dta eeodrea Ideal pour alxtea pria a 6 % d'oau.
(T) Tub* Toumant >ur cendrea groaalhrea
aeulea
-J 1 II L_
FIG. 2 Production de vapeur 30 b/300C du calcinateur DROCOURT
160
Co*but*on d* s c b l i u * houlllara
)7S0kJ/k) i i c
HO'C
2.S/i
70% iir * n xei
Fortoa
fin** particuloo
P-rt#o
rondonont
tub* touraant
Production do vapour Portoa

(30 b . n - 30O"C)
' rondoaont ehaudlor*
U . 9 t/h
Rondoaont global 57,J %
FIG. 3 - Production de vapeur - Calcinateur de DROCOURT
DISCUSSION
Question - Mr. Sirchis, CEC :

Etant donne La temperature du Lit, temperature que vous avez
qualifiee d'auto-thermicite, y a-t-iL des risques d'auto-ex-
tinction du lit et dans ce cas, comment avez-vous prevu d'eva-
cuer la masse chaude du combustible?
Answer Nos essais sur maquettes chaudes n'ont jamais mis en evidence
de risques d'extinction du lit aux temperatures inferieures a
850 C auxquelles nous avons travaille. Si la temperature de-
vait descendre en-dessous de cette limite inferieure, nous
avons prevu I'injection d'un combustible gazeux a la base de
des pyramides qui se comporteront alors comme des bruleurs
Ainsi, nous pourrons remonter la temperature a une limite
acceptable. Si neannoins nous sommes amenes a evacuer du so-
lide pour une raison quelconque, la grille de fluidisation est
munie d'un debordement utilise, en marche normale, pour reti-
rer les cendres calcinees en les faisant passer dans le dis-
positif de recuperation de la chaleur sensible, et done parti-
ciper a la production de vapeur.
161-
TOTAL PRECALCINATION BY CIRCULATING FLUID BED
Authors : Dr. Albert FOLLIOT
Contract number EE-035-81
Duration : 24 months
Head of project P. JAVELLE
) LURGI UMWELT UND CHEMOTECHNIK

Contractor : GIE ) CREUSOT-LOIRE
) LAFARGE COPPEE
Address : CREUSOT-LQIRE Entreprises

B.P. 107
F - 71103 Chalons-sur-Saone Cedex
SUMMARY
Vforld cement consumption increases at the same rate as world population.

It will step in the next twenty years from 900 M m.t per year to 1.400
M m.t per year.
The development in this industry, of high performance processes, suitable

for low-grade fuel valorisation, is explained by the increasing price of
fossil fuels, which account for half of direct expenses in cement
manufacture.
So the contractors (LURGI UMWELT UND CHEMOTECHNIK, CREUSOT-LOIRE and

LAFARGE COPPEE) study the use of a recirculating fluid-bed as calcinator.
This reactor will use fuels of high ash contents or low grindability
(coal shales, vegetal refuse) to supply up to seventy percent of the
process thermal balance.
The recirculating fluid-bed, developped by LURGI, is used for Alumina

calcination and coal shales combustion. After conclusive precalcination
test of cement raw meal at a 2 m.t per day rate, the development of a
1.250 m.t per day prototype was decided.
This prototype, object of the present contract, will be set in the

PORT-LA-NOUVELLE Cement plant (Aude) of CItffiNTS LAFARGE FRANCE.
It will use coal shales from a neighbouring coal dump. The use of
vegetal refuses is also foreseen.
-162-
1 - L'EVOLUTION TECHNIQUE PES CIMENTERIES
1 . 1 . Lea procedes par vole seche
Jusqu'a 11 y a une trentalne d'annees, l e four r o t a t i f e t a l t l e

seul instrument de c l i n k e r i s a t i o n u t i l i s e , la plupart du temps
en vole humlde, e t en vole seche lorsqu'on manquait d'eau.
La raret de l ' e n e r g i e apres la 2eme Guerre Mondlale, puis son

rencherlssement a partir de 1974, ont assure la suprematie de la
vole seche qui consomme normalement 800 thermies a la tonne de
c l i n k e r , contre 1300 pour la vole humlde.
Les echanges dans l e four r o t a t i f se faisant a travers la seule

surface du talus de matiere sont e f f i c a c e s dans la zone de c l i n -
ker i s a t ion, parce q u ' i l s mettent en jeu l e rayonnement de la
flararae, a raison de la puissance quatrieme de la temperature
absolue. Par contre, en dessous de 900, ou l e rayonnement de-
c r o l t considerablement, la seule convection prevaut e t l ' e c h a n -
ge thermique, trs f a i b l e , contraint a un allongement exagere du
four.
En 1956 environ, HUMBOLDT a propose un echangeur a transport

pneumatique, conslstant a i n j e c t e r la matiere pulverulente dans
une conduite, traversee par l e s gaz sortant du four, l a ma t i e re
entratnee e s t lnstantanement en e q u i l i b r e de temperature avec
l e s gaz chauds - la surface d'echange par convection etant de
l'ordre du metre carrS par gramme - puis separee des gaz dans
un cyclone ; on place a i n s i en s e r i e 4 ou meme 5 t e l s d i s p o s i -
t i f s , e t a la s o r t i e du dernier, la ma t i e r e entre dans l e four.
1.2. La prScalclnatlon
Environ 1970, un nouveau prog res a ete accompli, permettant un

nouveau raccourcisseraent du four r o t a t i f , en operant la decar-
bonatation dans un rSacteur f i x e , alimente separement en com-
bustible et en a i r de combustion, s o i t par prelevement d'une
partie des gaz du four, fonctionnant dans ce cas avec un fort
exctSs d ' a i r , s o i t par une conduite s p e c l a l e amenant l ' a i r chaud
du refroidis8eur.
1.3. L ' u t l l i s a t l o n des combustibles bon marche
La composition du c l i n k e r , chaux, s i l i c e , alumine, f e r , permet

l ' u t l l i s a t l o n de combustibles tres pauvres, parce que leurs cen-
dres sont t r e s gnralement formees des mimes mineraux. Leur
f a i b l e pouvoir calorifique pose cependant des problemes de com-
bustion et l e s recherches de nouveaux procedes doivent e t r e
o r i e n t e e s dans l e sens d'une combustion complete e t a i s e e .
-163-
1.4. Recherches des conditions optimales
Un reacteur optimal pour l'utilisation de combustible faible

cinetique de combustion, suppose simultanement une turbulence
raisonnablement elevee et un long temps de sejour. Gertains
reacteurs flash assurent de hauts niveaux de turbulence, mais
des temps de sejour tres faibles, de l'ordre de la seconde.
Par ailleurs, les reacteurs a lit fluidise classique assurent
quant a eux, un long temps de sjour, mais une productivity
faible sur les produits trs fins. L'actuel dveloppement est
orient vers un reacteur permettant d'assurer simultanement un
temps de sejour long, de l'ordre de la dizaine de minutes, une
turbulence raisonnable et une productivite elevee.
2 LA TECHNIQUE LCL
2.1. Optimisation du lit fluidise
Depuls plus de 50 ans, les lits fluidlses ont ete developpes

comrae reacteurs chimiques pour les particules fines (< 3 mm)
dans la plupart des secteurs de 1'Industrie.
Pour des particules de taille donnee en fonction de la Vites

se des gaz, le comportement du systeme gaz + matire est repre
sents sur la figure 1 que nous commentons :
L'expansion du lit fluidise crot en meme temps que la Vites

se moyenne du gaz de fluidisation. Ce faisant, le type de flui
disation evolue et passe du lit fluidise classique avec une sur
face definie au lit fluidise en circulation, puis au racteur
a transfert pneumatique.
Le lit fluidise en circulation (reacteur B) est employe pour

des particules de 50 a 800 microns. Ce reacteur presente, comrae
on peut le voir sur la figure 1, les plus grandes vitesses de
glissement gazmatire pour une granulomtrie donnee. Ceci est
du a la charge eleve de ma tie re dans la zone super ieure du
reacteur, d'oii re suite une aggregation des particules qui di
minue sensiblement l'effet d'entranement par le gaz.
La figure 2 presente la structure du lit fluidise circulant.

Ce reacteur se compose de trois elments : le fluidiseur, le
cyclone de recirculation, le joint fluidise assurant le retour
des produits cyclones au fluidiseur et 1'extraction des produits
finis.
La partie basse du fluidiseur comporte une grille par laquelle

est injectee une fraction de l'air de combustion
qui assure la fluidisation en lit dense
dans la partie inferieure du reacteur. L'air complementaire
necessaire a la combuscion est introduit au tiers de la hauteur
du reacteur et assure 1'expansion et le transport de la raatie
re precedemment fluidisee. La matiere en trainee dans la partie
164
haute du reacteur e s t separee dans le cylcone de r e c i r c u l a t i o n . Le
j o i n t f l u i d i s e a la base de ce cyclone retourne la majeure part de
la matiere au reacteur f l u i d i s e , et simultanement, assure l ' e x -
taction reguliere du produit t r a i t e .
La matiere 2 t r a i l e r , a i n s i que le combustible n e c e s s a i r e , e s t

introduite lateralement dans la partie basse du f l u i d i s e u r .
On peut ajuster la charge circulante e t par 13, l e temps de

sejour, done de traitement de la matiere en jouant sur la
vanne d ' e x t r a c t i o n .
2 . 2 . La c a l c i n a t i o n de l'alumine
Le principe du pro cede precedent a f a i t l ' o b j e t de travaux de

la part de la Metallgesellschaft a FRANCFORT, 11 y a plus de
20 ans. La cooperation qui a ' e s t instauree par la s u i t e entre
lea deux s o c l e t e s LURGI e t VEREINIGTEN ALUMINIUMWEKKEN (VAW)
devait conduire aux premieres r e a l i s a t i o n s a l ' e c h e l l e indus-
t r i e l l e , en passant par une i n s t a l l a t i o n p i l o t e a l ' e c h e l l e
aemi-industrielle.
Pour la c a l c i n a t i o n de l'alumine, une temperature de l'ordre de

1100C e s t necessaire. Par la maltrise a quelques degres pres
de c e t t e temperature, partout dans l e reacteur, ce procede a
permis d'ameliorer de facon determinante e t l e s c a r a c t e r i s -
tiques c r i s t a l l i n e s de l'alumine deshydratee, et la consomma-
tion d'energie par rapport au systerae de c a l c i n a t i o n c l a s s i -
que, l e four r o t a t i f .
A ce jour, plus de 20 u n i t e s , avec des capacltes atteignant

1500 tonnes, jour , ont e te i n s t a l l e e s .
2 . 3 . Application 3 la fabrication du ciment
Selon ce qui a ete d i t au chapitre 1, la q u a s i - t o t a l i t e de la

consommation thermique re s u i t e de la decarbo natation du c a l c a i -
re. Cette reaction, monovariante, se produit pour l e C03Ca pur,
a la temperature constante de 890C sous la pression atmosphe-
rique. Un reacteur d l i t f l u i d i s e , dont la c a r a c t e r l s t i q u e e s t
d'assurer des transferts thermlques iroportants a temperature
constante, e s t bien adapte d une t e l l e reaction.
Des contacts entre le Groupe de Recherches assocle de LAFARGE-

COPPEE et CREUSOT-LOIRE ENTREPRISES ont e t e pris avec LURGI
aes 1979.
Une experimentation d'une semaine, portant sur 10 tonnes de

cru de cimenterie, a te r e a l i s e e en mars 1980 dans la s t a t i o n
d ' e s s a l s LURGI a FRANKFURT. On a obtenu a ce stade, une decar-
bo natation a plus de 98 X & des temperatures inferieures &
900C d'une part, et d'autre part, une marche s a t i s f a i s a n t e du
reacteur jusqu'a 1050C.
165-
Compte tenu de la possibility d'utiliser dans l'usine LAFARGE
de PORT-LA-NOUVELLE, a titre de matiere d'appoint des schistes
charbonneux, dechets de lavage des mines voisines, la dcision
a t prise d'adjoindre la tour DOPOL existante un prototype
de prcalcinateur LCL. Ce dveloppement en cimenterie du lit
fluidise circulant doit permettre de promouvoir la substitution
de combustibles degrads a des combustibles classiques et d'etu-
dier des possibilities de superprecalcination (precalcination
totale et surchauffe de la matiere avant introduction au clinke-
riseur).
L'OPERATION PILOTE DE PORT-LA-NOUVELLE
3.1. Objectifs
Situee en bordure de mer dans la region LANGUEDOC-ROUSSILLON,

l'usine de PORT-LA-NOUVELLE constitue la plus recente realisa-
tion de la Societe des CIMENTS LAFARGE - FRANCE.
Les objectifs que l'on se propose de realiser sont d'ordre mul-
tiple.
Sur le plan procd, caractrisation du nouveau systeme de pre-

calcination par ses performances sur le plan thermique et son
aptitude a 1'utilisation de combustibles solides bon marche a
faible pouvoir calorifique. Les premieres experimentations con-
cerneront l'emploi de schistes houillers de faible valeur calori-
fique, en provenance des terrils du Bassin Houiller des Cevennes.
Elles seront prolongees par 1'utilisation de dechets vegetaux et
eventuellement de rejets menagers.
Sur le plan exploitation, determination des avantages apportes

par le procede LCL, notamment pour ce qui concerne :
- Les couts de fabrication,

- la capacite de production,
- la regularity du clinker.
Sur le plan technologique, optimisation des equipements (Meca-

nique - Refractaires - Controles et Regulation).
La mise en service de 1'installation LCL est prevue pour Octobre

1982. Apres unephase experlmentale de l'ordre de 12 Mois, il
devrait done etre possible de disposer dans un an des elements
permettant d'etablir le dossier de faisabilite du procede, puis
d'engager l'etape de developpement industriel.
3.2. Descriptif
Mise en service en 1970, la cimenterie de PORT-LA-NOUVELLE uti-

lise un procede de cuisson par voie seche avec prechauffage de
farine dans une tour comportant 4 etages de cyclones echangeurs.
La production de l'usine est actuellement de 1200 tonnes par
jour de clinker.
-166-
La ligne de culsson e x i s t a n t e comporte :
- Un prechauffeur DOPOL a 4 etages de cyclones doubles,

- Un four POLYSIUS sur 3 appuis de 0 4 n x 65 m,
- Un r e f r o i d i s s e u r a g r i l l e FULLER 0 2,75 x 16,925 m.
Le p l l o t e de precalcination e s t imp Ian te 3 la base de la tour

de prechauf fage, entre l e four r o t a t l f et l e dernier etage de
cyclones echangeurs. La d i s p o s i t i o n de la tour e x i s t a n t e ne
nous permettant pas de mettre l e LCL normalement en charge sur
l e four, nous avons ete contraints de l e placer au sol et de
remonter l e produit precalcine par un c i r c u i t de gaz u t i l i s a n t
l e s fumees du four. L 1 ensemble des nouveaux equipements e s t
supports par une nouvelle tour en be ton de 35 metre de hauteur.
II comprend (Voir figure 3) :
1. Le reacteur a l i t f l u l d i s e proprement d i t e t son cyclone de

depousslerage complementaire.
2. La conduite de derivation des fumees s o r t i e four, qui assure

la reprise des matieres precalcinees a la s o r t i e du reacteur
LCL, e t leur retour au cyclone d'alimentation du four.
3. La chambre de detente et la conduite d'air t e r t i a i r e qui

assurent 1'alimentation du reacteur en a i r chaud poussiereux.
4. Le c i r c u i t de preparation e t de transport de l ' a i r propre de

f l u i d i s a t i o n - Ce c i r c u i t comporte :
. Une conduite d'air quaternaire chaud preleve au r e f r o i d i s -

seur et son cyclone de depousslerage.
. Un changeur de chaleur a i r poussiereux-air propre.
. Une conduite d'air de f l u i d i s a t i o n surpresse et rechauffe.
5. Les v e n t i l a t e u r s et surpresseurs necessaires au bon fonction-

neraent du procgdS.
3 . 3 . Programme experimental
Apres une periode de l'ordre de 1 mo i s consacree 3 la mise en

service des equipements :
. Essais de fonctionnement 3 vide et en charge,

. Etalonnage des appareils (Pression - Debit - Temperatures-
Analyseurs de gaz).
Les e s s a i s prvus pendant l e s 12 premiers mo i s de marche de

1 ' I n s t a l l a t i o n LCL de l ' u s i n e de PORT-LA-NOUVELLE comprendront :
167-
A. Recherches des conditions optimales de marche du precalcina-
teur LCL.
(Duree 3 mo is - Combustible charbon) .
Controle de la temperature et optimisation des quant ites

d'air chaud provenant du refroidisseur a clinker.
b. Recherches de la repartition optimale entre combustibles

four et precalcinateur.
(Duree 1 mo is Combustible charbon + schistes) .
c. Campagne de mesures sur les limites de la superprecalcination.

(Puree 3 mo is Combustible charbon).
d. Adaptation des meilleures conditions de marche du precalcina-

teur a 1'utilisation de schistes houillers.
(Puree 5 mois - Apres mise en service d'un circuit special
de broyage des schistes).
Influence de la granulometrie du schiste sur les conditions

combustion et la qualite du clinker.
Bilan d'ensemble.
Ce programme sera conduit sous la responsabilite du partenaire

GIE/PLN, en liaison etroite avec le CERILH qui en etablira, sous
sa responsabilite, le rapport portant notamment sur le bilan
energetique d'ensemble, ainsi que sur les parametres d'exploita-
tion industrielle.
3.4. Resultats escomptes
Le resultat le plus immediat doit etre une economie substan-

tielle sur le cout energetique de la fabrication du ciment gra-
ce au remplacement de quantites importantes de charbon par des
schistes houillers en provenance du Centre-Midi.
Compte tenu de la composition chimique du calcaire disponible

dans la carriere de l'usine de PORT-LA-NOUVELLE, de 1'analyse
des cendres et du pouvoir calorifique de 2000 kcal/kg de ces
schistes, 1'utilisation de ce combustible de substitution de-
vrait representer de l'ordre de 32 % des besoins therraiques
necessaires a la fabrication du clinker.
Ceci perraettra une economie d'environ 50 kg de charbon par tonne

de clinker, soit plus de 10 000 T.E.P. par an.
Compte tenu du prix de revient de ces schistes, 1'operation

devrait degager une economie de 3 500 000 F.F. par an sur les
frais de fabrication.
Les essais programmes permettront d'autre part, de tester l'aug-

mentation de capacite potentielle resultant d'une telle instal-
lation, augmentation devant etre particulierement elevee compte
tenu des performances attendues du procede.
-168-
L ' i n s t a l l a t i o n p i l o t e permettra, dans un staJe u l t e r i e u r , l ' u t l -
l l s a t l o n de d i f f e r e n t s residus combustibles actuellement de v a -
leur d'usage n u l l e .
PERSPECTIVES DE DEVELOPPEMENT
4 . 1 . Le procede LCL : La precalclnatlon par combustibles non

conventlonnels
Deux atouts permettent au procede LCL de se placer pour la v a l o -

r i s a t i o n de combustibles non conventlonnels :
La p o s s i b i l i t y de regler l e temps de sejour de la ma t i e re et du

combustible dans le reacteur entre 5 e t 20 minutes environ pour
l e temps de sejour moyen.
L'aptitude a prendre en charge dans la zone inferieure du reac-

teur en l i t f l u i d i s e dense, des p a r t i c u l e s de 1 a 2 mm, voire
plus, pour des materlaux de f a i b l e d e n s i t e .
A i n s i , nous comptons developper l e precalclnateur LCL pour la

v a l o r i s a t i o n thermique de dechets, notamment d'origine v e g e t a l e ,
d i f f i c i l e m e n t broyables et qui ne peuvent e t r e u t i l i s e s ni dans
l e s bruleurs c l a s s i q u e s , ni dans l e s reacteurs type f l a s h . Pour
ces combustibles, la p o s s i b i l i t y d'incorporer a temperature mo-
deree (env 800C) , l e s res id as de combustion au cru de cimen-
t e r i e resoud egalement l e s problemes l i e s 3 la f u s i b i l i t e des
cendres.
Nous comptons egalement developper l e LCL pour l'emploi de com-

b u s t i b l e s a haute teneur en cendres, t e l s l e s s c h i s t e s h o u i l l e r s ,
dont la cinetique lente de combustion l i m i t e l'emploi dans l e s
bruleurs c l a s s i q u e s . Pour ces combustibles, l'abondance des cen-
dres ne pose pas dans ce cas, de problemes de d i s p o s i t i o n des
r e j e t s , ceux-cl etant integres au clinker ; t o u t e f o i s , en fonc-
tion des carrieres disponibles et de la composition des cendres,
le taux de subsltutlon des s c h i s t e s aux combustibles nobles peut
etre l i m i t e .
4 . 2 . Extension du l i t f l u i d i s e a La cMnkerlsation
L'aptitude du rSacteur LCL a surchauffer la matiere precalcinee

a plus de 1000C (Cf 2 . 3 . ) permet de r e a l i s e r au niveau d'un
racteur statique une part iraportante des echanges normalement
effectues dans le four de c l i n k e r i s a t i o n .
Dans une t e l l e u t i l i s a t i o n du LCL, couple a un four r o t a t i f ,

c e l u i - c i pourrait encore etre raccourci, au point de pouvoir
reposer sur deux appuis seulement. Ceci entralnerait un a l l e g e -
ment s e n s i b l e du four et de sa mecanique.
Par a i l l e u r s , l e Groupement envisage d ' a s s o c i e r la superprecal-

clnation (precalclnatlon + surchauffe) a un reacteur de c l i n k e -
r i s a t i o n statique en l i t f l u i d i s e .
169-
La mise au point d'un procede entierement s t a t i q u e de c l i n k e r i -
s a t i o n p e r r a e t t r a i t evidemment de marquer deux avances d e c i s i v e s
dans l e domaine de l'econoraie d ' e n e r g i e e t dans- l e domaine des
i n v e s t i s s emen t s .
jj: VITESSE RELATIVE GAZ MATIERE POUR

DIFFERENTS REACTEURS SOLIDE GAZ
<=>GAS
--SOLIDS
LEASING
SOLIDS
THROUGHPUT
MFANÔI ins
VELOCITY INCREASING EXPANSION
FIG.2:
CALCINATEUR A
LIT FLUIDISE
CIRCULANT L.C.L.
SECONDARY AIR
PREHEATED
RAW MEAL,
COAL.SHAL^ = CALCINED MEAL

TO KILN
PRIMARY
AIR
-170-
Vers sommat'
A Boihe cUt-cnVc
\? Y *
fOU.tR
D I M . en suspension
da calcine
Air flmdisohon
FIG. 3
DISCUSSION
Question - Mr. Robinson, Deborah FLuidised Combustion Ltd. :

. I have a question related to this subject that I would like
clarification of and I am sure many people in the audience
would also like to hear this. In fact, I have split the
question into three parts. They all interconnect really to
try to arrive on the conclusion.
- First, what is the thermal capacity of the plant being
installed?
- Secondly, what is the proposed oxygene content in the flue
gaz of the reactor cycle?
. - Thirdly, what is the expected electrical power consumption
for the reactor section of the plant?
Answer La capacite du four est de 1.200 tonnes. La consommation

thermique normale est de 800 thermies/tonne pour la totalite
du processus,dont 500 thermies dans le calcinateur. Ce qui
fait done par jour, 500 calories par kg.
La teneur en oxygene ne depasse jamais 1% comme e'est I'habi-
tude des cimentiers.
D'autre part, la consommation electrique est estimee environ
a 3 ou 4 kWh a la tonne.
Question - Mr. Fortuin, NE0M :

I would like to ask if a comparison was made in this project
between the circulating fluidised bed and the conventional
fluidised bed and secondly I would like to know if the final
decision was taken based on several proposals which were
gained from industry or was it only Lurgi which was in the
picture in this particular case?
Answer La finesse de la matiere crue de ciment, qui s'etale dans une

echelle entre 5 a 100 microns avec la majeure part dans les
30 a 40 microns, et aucune particule superieure a 100 microns,
condamne le lit classique, parce que les vitesses admissibles
des fluidisations avec des particules aussi petites sont tres
faibles. Par consequent, si vous avez tres peu de gaz vous ne
pouvez mettre que tres peu de combustible.
D'autre part, nous sommes interesses au systeme Lurgi a cause
de I'homogenite des temperatures et surtout pour I'experience
acquise a'vec I'alumine dont la dimension de grain est de
I'ordre de 30 microns, c.a.d. tout a fait analogue a celle du
ciment.
172-
FLUIDIZED BED BOILER USED IN THE PRODUCTION OF BRICKS
Authors : B.G.H. WEBERS
Contract number : EE-126-80
Head of project : B.G.H. WEBERS
Contractor : DE HAZELAAR
Address : Saeffelderstraat 10
N L - 6104 KONINGBOSCH
SUMMARY
In September 1981 a contract was signed for the construction of an AFBC-

bollerplant In the brick-works 'De Hazelaar' In the Netherlands.
Hot commissioning of t h i s bollerplant started at the begin of August 1982.

Next October the f i r s t operating experiences are expected.
The AFBC-boller has a rated capacity of 10 tons steam per hour and produces
saturated steam at 16 bar ( g ) .
The b o i l e r c o n s i s t s of a watertube combustor chamber and a convectlve s h e l l

b o i l e r . In the combustor chamber a freeboardzone, providing adequate residence
time, Is present above the fluidlzed bed.
The steam Is used as the direct heating medium i n the brick-production

process. The AFBC-boller i s intended to serve the p l a n t ' s t o t a l steam needs
and In t h i s way to s u b s t i t u t e heavy f u e l - o i l by c o a l . This implies high
demands on a v a i l a b i l i t y , load-following and part-load capacity. Emission of
S02 w i l l be reduced s u b s t a n t i a l l y by adding limestone to the c o a l . The s o l i d
residue of the combustion i s u t i l i z e d within the brlckproductlonprocess.
Particulate emission i s controlled by use of a f a b r i c - f i l t e r .
The production schedule, depending on the building market, shows a two or three
s h i f t - o p e r a t i o n during five days per week. An e x i s t i n g s h e l l b o i l e r , fired
with heavy f u e l - o i l and of the same capacity as the AFBC-boiler, w i l l
serve as back-up f a c i l i t y .
173-
The purpose of t h i s project i s the production of steam by using coal instead
of fuel o i l i n an i n d u s t r i a l environment and by doing s o , to demonstrate the
commercial v i a b i l i t y of AFBC-plants in the range of about 5 through 25 MW
thermal.
The project received financial support from the European Communities and the
Dutch Government, under the energy saving- and fuel d i v e r s i f i c a t i o n
programmes.
On behalf of the demonstration, the AFBC-plants performance w i l l be

registered during the f i r s t three years of operation, for which purpose
additional instrumentation i s i n s t a l l e d and a performance r e g i s t r a t i o n
program w i l l be executed.
In t h i s registrationprogram s p e c i a l a t t e n t i o n w i l l be paid to the t e c h -
n i c a l , including the environmental, and the economical r e s u l t s , which w i l l be
disclosed and used for the encouraging of follow-up-projects.
1. Introduction
The AFBC project 'De Hazelaar' is intended to serve as an example for

commercial FBC application in the range of 5 up to 25 MWth.
The purpose of the project is industrial steam raising by substituting coal
for heavy fuel oil in an industrial environment under commercial con-
ditions and in an environmentally acceptable way. As it concerns the
first industrial AFBC-application of this kind and magnitude in the
Netherlands, potential risks with respect to technical economical perfor-
mance will be involved. Financial and technical support is therefore
necessary as a compensation and an incentive bonus.
The results of the industrial performance of 'De Hazelaar' will be
disclosed and published in order to promote intended follow up projects.
2* 2E2ie.llf.EiRi22
The purpose of the project is to produce steam at reduced (energy)costs,
by substituting coal for heavy fuel oil in an environmentally acceptable
way.
-174-
If t h i s objective can be net In a commercial plant and i n an i n d u s t r i a l
environment the application of AFBC plants in the range of 5 through 25 MW
thermal i s demonstrated s u c c e s s f u l l y . This w i l l encourage follow up pro-
j e c t s of comparable s i z e s , provided that proper marketing e x i s t s and
government support Is available i f necessary.
Calculations show that, under the assumptions made, which are l i s t e d i n
table 2 , the investment in AFBC i s p r o f i t a b l e . If coal use i s considered,
the choice of AFBC i s e s p e c i a l l y favoured i f stringent emission standards
i n connection with coal use are required.
The undersecretary of the environmental protection department i n the

Netherlands, however, has recently recommended l e s s stringent S02~emission
guidelines for i n d u s t r i a l coalplants for the time being u n t i l 1990. As a
consequence, the major incentive for the AFBC-option i s lacking for short
term decisionmaking. However, fuel v e r s a t i l i t y and prospective lower
operating c o s t s could make the AFBC option viable i n the Netherlands in
comparison with conventional coal options.
The value of the demonstratlonproject at 'De Hazelaar' i s anyhow that in
I t s c l a s s an adequate reference Is stated and that the r e s u l t s w i l l become
available of t h i s commercial AFBC-plant. This enables a b e t t e r technical
and economical evaluation, in view of i n v e s t i g a t i n g marketing-
possibilities for follow up projects.
A l l relevant operating data and r e s u l t s of 'De Hazelaar' w i l l be

recorded, evaluated an d i s c l o s e d . Technical aspects, including environ-
mental and economical aspects, w i l l be emphasized. In order to perform
the demonstration, additional instrumentation i s i n s t a l l e d and additional
measurements w i l l be carried out.
The demonstrationprogram w i l l cover the f i r s t three years of operation.
During the f i r s t year attention w i l l be necessary, as a n t i c i p a t e d , for
eventual technical problems related to teething troubles. After solving
these problems, the technical-economical demonstration w i l l be carried
out.
In view of the financial r i s k s , related to the newness of the technology,
financial and technical support was necessary for the project to be
realized.
-175-
The project received support i n view of these r i s k s on account of the o i l
saving and fuel d i v e r s i f i c a t i o n policy of the Committee of the European
Communities under the energy savingprogram and from the Dutch Government
under the National Dutch Coalprogram.
The Committee provided a risk-bearing loan for a part of the construction
c o s t s of the plant and furnished a grant for additional instrumentation
and measurements which are necessary on behalf of the demonstration.
The Dutch government gave additional subvention, apart from structural
s u b s i d i e s , for a part of the construction c o s t s of the plant.
The project was founded sepember 1981 when spadework was succesfully
rounded off with the order by 'De Hazelaar' for the delivery of the
plant.
Hot commissioning started at the beginning of August 1982. The transfer
of the plant w i l l be issued at the end of September.
The b o i l e r has now, 6 weeks a f t e r the f i r s t operation on coal began, pro-
duced about 1500 tons of steam. The plant i s i n operation during one
s h i f t daily up to now, and can provide almost a l l the necessary pro-
cesssteam already.
In the beginning there have been some technical problems with respect to
the coal and limestone transport-equipment which seems to be overcome by
now. Within three months a f t e r the transfer of the plant, a guarantee
performance testprogram w i l l be carried out in co-operation with the
supplier.
3 . Processdescription
The AFBC project i s r e a l i s e d i n the sand-lime brick plant 'De Hazelaar'

which i s located i n the municipality of Echt near Koningsbosch, i n the
south-eastern province Limburg of the Netherlands.
The brickplant i s a modern f a c i l i t y with the capability of manufacturing
bricks in current s i z e s and also elements of much larger s i z e s .
The productionschedule depends on the building market and shows a two - or
three - s h i f t operation during five days a week.
The e x i s t i n g boilerhouse had to be pulled down partly and rebuilt in
larger dimensions. Three o i l storage tanks were moved and two e x i s t i n g
o i l f i r e d b o i l e r s had to be removed.
-176-
The new plant (see scheme 1 , 2) includes an AFBCboller with economiser,
baghouse f i l t e r , flue gas stack, feedwater treatment, coal and limestone
storage bunkers and Intermediate hoppers, and transport and handlingequlp
ment for c o a l , limestone and s o l i d residue. The b o i l e r produces saturated
steam at 18 bar (g) which i s the direct heating source in the brick manu
facturing process. As a result of contamination in the process the steam
condensate cannot be u t i l i z e d as b o i l e r make up water.
The AFBC b o i l e r Is intended to serve the plant's t o t a l steamconsumptlon.
This implies high demands on a v a i l a b i l i t y , load following and partload
capacity. An e x i s t i n g four year old s h e l l b o i l e r fired with heavy fuel
o i l and of the same capacity as the AFBCboiler w i l l serve as standby
facility.
The AFBCplant w i l l operate f u l l y automatically except for s t a r t u p and

shut down procedures.
The rated steam capacity i s 10 t / h on f u l l load with a aonsumption of
some 1,2 t / h steam c o a l . Minimum load Is achieved with a turn down r a t i o
of 1 4.
Imported steamcoal of regular trade quality w i l l be used normally though

different q u a l i t i e s , namely high ash, low v o l a t i l e , high sulfur c o a l , w i l l
possibly be u t i l i z e d in exploring relations with technical and economical
performance. The coal w i l l be processed with just a minor treatment in
crushing the larger p a r t i c l e s , which are present in the feed, to below
30 mm s i z e .
Limestone is used as a sorbent for SO2 by which 80Z sulfur retention is
aimed at. The limestone is a product from nearby German or Belgium
quarries and will be supplied In the required size distribution.
Coal and limestone are supplied by lorries and are tipped and separately
stored in a covered underground bunker with some 80 hours full load
storage capacity.
Fly ash, a mixture of coal ash, unburnt carbon and spent sorbent, Is
separated on different places from the flue gas.
Particulates are ultimately separated in a baghouse separator with high
efficiency, resulting in an emission of definitely less than 50 g/GJ.
Nuisance act regulations, which have to be met, impose a maximum particulate
emission of 20 g/GJ.
177
The solid residue will be utilized as a raw material, in fact as a sand
substitute, in the brick manufacturing process. Since the quantity of
8olid residu is relatively small, about 0,5% m/m of the total raw
material throughput, special batches of bricks with larger portions of
residue will be manufactured in order, either to establish the maximum
allowable ratio, or to find the specifications to be met for substantial
substitution of either sand or lime.
The performance of the AFBC plant and the utilization of the residue will
be registered on behalf of the demonstration.
Performance registration program (an outline)

1. For the benefit of the demonstration a performance registration
program of industrial nature will be performed.
Basic objectives of the demonstration are:
- to evaluate the technical and economical merits of 'De Hazelaar'
AFBC-plant;
- to disclose the basic data for the benefit of the commercialisation
of intended succeeding projects;
- to serve as a reference for potential users;
- to investigate the utilization of the solid residue in building
materials;
- to evaluate the price performance relation with respect to environ-
ment conservation;
- to learn how to handle the solid materials, inherent to coalfiring.
2. Survey of the workingprogram and instrumentation.
a. Additional provisions and instrumentation are:

- sampling points in the flue gas and solid residue flows;
- flow measurement of coal and limestone feed;
- oxygenanalyser;
- SO2 and particulate - analyzer.
-178-
b. The working prograa Is b r i e f l y reviewed hereafter:
- a hazard and operabilltystudy w i l l be performed i n co-operation wit
the supplier. This study i s necessary i n view of generating a deepe
understanding in the aspects of safety and r e l i a b i l i t y with
respect to AFBC-technology i n general, and 'De Hazelaar' AFBC-
plant specifically.
Operating instructions and Instrumental safeguards, supplementary
t o the e x i s t i n g provisions, w i l l eventually be introduced i f con-
sidered necessary.
- A logbook of the performance during normal operation w i l l be kept.
The c o l l e c t e d operating data form the raw material to be used i n
evaluating the boilerplants performance.
After checking the basic data and i s o l a t i n g unreliable data, a
record w i l l be made of:
* processconditions;
* consumption of coal, limestone, light up fuel, heavy fuel oil
(stand-by boiler), feedwater, power;
* combustion air;
* steamproduction;
* solid residue production;
* emission of SO2, 00, OO2, particulates;
* coal- and limestone-analyses;
* plantstartups, shutdowns and failures.
A survey will be made every month of the operating and running costs.
- Performance tests.
In order to establish an independent data base on the performance,
comprehensive measurements and analyses under reproducable conditions
will be carried out at regular intervals. Performance will be
registrated at maximum continuous load, at minimum load and as far as
necessary with the use of different qualities of coal or limestone.
- Exploratory testing.
Exploratory tests will facilitate the Interpretation of performance
tests and the comprehension of environmental effects. These tests
will be performed occasionally.
179-
- Boiler inspections.
An insp"".tion program will be worked out on parts of the boiler-
construction and components, which have to stand heavy duty, in order
to establish the up to date condition and the life expectancy.
- Dynamic behaviour.
Load following capability depends on the design and on actual
processparameters. Timeconstants and response rates will be deter-
mined in relation with processconditions.
- Utilization of solid residue.

Investigation will be necessary as a support for possible processing
problems in the brickplant. Product quality will be determined in
relation with ashparameters.
Economy
Basically the rate of return on investments depends on the investment

costs either total or marginal, and the net energycost savings. Besides
coal and oil prices and additional operating costs, the utilization of
the boiler plant is an important factor.
'De Hazelaar' will normally operate two or three shifts daily during 5
days per week and 45 weeks per year.
The average steam consumption at full load conditions of the brick plant
depends on the scheme of switching of the autoclaves. At 'De Hazelaar'
about 80% of the maximum continuous boiler rating is expected as an
average.
The anticipated overall boiler utilization over a 10 years evaluation
period will vary between 33% and 50%.
Assuming a 50% utilization, about 3000 t/a of heavy fuel oil will be
substituted.
Under the existing circumstances if the building market the boiler utili-
zation will be close to 33%. In the next future the effect of the exe-
cuted reorganisation in the sand-lime brickindustrie and the growing
market share of 'de Hazelaar' will lead to improved utilization.
180-
In table 2. the return on investment is calculated based upon a 10 year
evaluation period and constant prices (no inflation). An Internal rate of
return on investment of 15Z (before taxes) or 11Z (after taxes) was found
which can be considered modest.
The above mentioned rate of return is related to the total investment
costs. For this project however, substantial grants and loans were allo-
cated, so that the rate of return for the owner and operator of the plant
can be substantially higher if the plant operates well.
6. Parties Involved.
'De Hazelaar' BV is the principal and operator of the AFBC plant and
of the performance registration program. The order for a turn key dell-
very was awarded to a Dutch bollermaker Standard Fasel BV, a subsidiary com-
pany of Standard Kessel AG which cooperates with Thljssen AG in the
development of AFBC-technology.
The Committee of the European Communities and the Dutch Government
granted substantial financial support.
The performance registration program will be carried out by several Dutch
subcontractors under responsability of 'De Hazelaar'.
NEOM BV, a subsidiary company of the Dutch government, assisted in the
realization and will supervise the execution of the demonstration.
-181
TABLE 1
Design data AFBC plant 'De Hazelaar'
Performance
Steam production (max. cont. rate) 10 t/h
Turn down ratio lf4
Steampressure 18 bar (g)
Steamtemperature (saturated) 210C
Thermal efficiency (m.e.r.) 85,5%
SO2 retention 80%
Fly ash emission 50 mg/m^
Fuel
Steamcoal 1,2 t/h
LHV* 25,1 GJ/t
Ash* 12%
Moisture* (Total) 8%
Volatiles 28%
Sulfur 1%
Size distribution 0-1 mm <: 10%
30-50 mm L 20%
* Ash and moisture included
Operating conditions
. Excess air 25%
Combustion temperature (m.c.r.) 950C
Flue gas temperature (ex eco) 180C
Ca/S (mole ratio) 3
182-
TABLE 2
Economic evaluation (10 year period)
Basic assumptions (basis 1981)
Availability 90Z (50Z, 1 s t . year)

Savings in heavy fuel oil 3000 t o n s / a
Fuel oil prices (41,5 GJ/t) f 550/ton
Real increase in oil price 3Z p . a .
Coal price (25,1 GJ/t) f 225/ton
Real Increase in coal price 1,5Z p . a .
Limestone cost f 25/ton
Residue disposal p.m.
Power consumption 100 KW
Maintenance costs 3Z of investment
Operating costs f 60.000 p . a .
Electricity costs (real increase 2Z p.a.) f 0,15/KWh
Inflation NO
Internal rate of return at constant prices (1RR)
IRR
Total capital, before taxes 15Z
Total capital, after taxes HZ
183-
Steam
ooocooooooo
_i
\ oooooo oooooY
OOOO OOO yf\
iNh.
p
startup
burner
&
_l
vM
ll u
waterwall
11
III 1 !
.1 ll ll
.I
n il;!' 1 ' 1 ! ,
Fluidised bed
combustion ,chamb|f>|:
Coarse grit : rIn; il',,

i ', ill
>.<>.;
II1
,<)\iv
Figure 1 - Atmospheric Fluidised Bed Boiler "De Hazelaar"

Flue gas
fresh water
1. cosl Bunker 10. flue gas stsck

3. llaestope bunker 11. ash container
3. c o s l hoppers 12. start-up burner
4. limestone hoppers 13. softener
r
5. AFBC chaaber 14. deaerator
6. shell boiler
7. economlser
8. cyclone
P. baghouse filter
Combustion.,
air
Figure 2
[ Fluidised Bed
A Freeboard
fll 1st Tube pass
jy 2nd Tube pass
Y Economizer
o
11
1000 JD
T1
i-l (1)
800 P u
i-l
m
>W h
600 >.Flue gas <l) i-H
C) nt
b

U-lh
00 h (1)
3 j=
water/steam en 4-1
T-H
5 200 II 4J
01 0)
u S z
g. 0
E
01
1 I 1 E" 10 20 30 i.0 50 60
H
Figure 3 Temperature path Figure 4 Net thermal rate (MW)
DISCUSSION
Question Mr. Puff, Cerchar :

Pourquoi n'aton pas retenu pour cette chaudiere La concep
tion de chaudiere a tubes d'eau et pourquoi aton pris une
chaudiere composite avec tubes de fumee? Estce que cette
chaudiere composite a tubes de fumee peut etre adaptee a
d'autres installations pour des besoins en vapeur haute pres
sion?
Answer The concept is a combination of water cooled tubes and a con

ventional boiler and I think it is the optimum combination
with respect to the combustion aspects. You can provide a
deep bed in which the combustion efficiency and a good com
mercial viability can be achieved in'comparison with a com
plete shell type boiler.
Question Mr. Puff, Cerchar :

La chaudiere Konig Ludwig est un lit profond a 100% tubes
d'eau. Y atil d'autres raisons que cette raison de com
bustion optimale p.e. une question de couts qui ont fait
preferer une chaudiere composite?
186
Answer We think that it is the best combination for this smaller
boiler type from an economical point of view and with respect
to the combustion characteristics.

I would like to comment to Mr. Webers that this design of boi-
ler is of course very similar to the Babcock Fluid Burned Boi-
ler which is being developed in the UK between the Babcock and
the National Coal Board and we have found the advantage to be
one of cost. There is a pressure limitation but in this appli-
cation it is clearly a very good boiler to apply. I would like
to ask Mr. Webers for his experience on the reliability of the
system and also any information he has on the turn down cha-
racteristics of the boiler?
Answer The boiler has been operated for about 200 hours and in the
beginning there were some troubles but these were not caused
by bad design.
About the turn down ratio, it is intended to turn down the
boiler to a capacity of about 25% and I must say with preli-
minary experiences that this high range of turn down ratio
indeed can be reached.
Question - Mr. Knobbout, expert :

What is the dimension of that fluidised bed?
What is the type of coal you are using and is it easy to buy
or do you have to mix it up with different types of coal?
Have you already any idea about efficiency on desulturisation?
Answer The type of coal which will be used is a normal trade, a re-
gular trade quality coal with about 1 to 1.5% of sulfur con-
tent and besides has a distribution between 0 and 30 mm with
a small fraction above 30 mm.
It is a coal we expect to be on the market in large quanti-
ties.
We have no experience at the moment on desulfurisation.
187-
MULTIFUEL FLUIDIZED BED
EXPLOITING INDIGENOUS SOLID FUELS .
Author : Peter O'BRIEN
Head of project : Peter O'BRIEN
Contractor : KERRY COOPERATIVE CREAMERIES LTD.
Address : Prince's Street

IRL - TRALEE Co. KERRY
SUMMARY
The conversion from oil based energy in an industrial process for

strategic and economic reasons using a fast circulating fluidized bed
is described. It is aimed not only at displacing oil with solid fuel
to the maximum extent, but to provide maximum fuel flexibility by
using a variety of widely differing fuels from coal as industrial
smalls to fuels of domestic origin such as peat, wood and coal shale.
The plant, a single multisolid fluidized bed of the Battelle/Foster

Wheeler type is designed to operate on various blends of these fuels
and it is aimed to demonstrate the versatility of the unit in using
these fuels efficiently.
At the same time, the process requires a steam service with a very
fast dynamic load response and a high level of turndown. The plant
has been designed to satisfy this requirement and it is aimed to
demonstrate this capability.
-188-
1. FUEL FLEXIBILITY
For strategic and economic reasons, it was required
to convert an industrial steam raising plant from
1001 oil firing to a combustion process which would
utilize not just coal but alternative (non-coal)
fuels.
1.1 Coal
Extensive research was conducted into the
practical and longterm availability of high
and lower grade imported fuels as well as the
prospects for development of potential fuels
of domestic origin. It was concluded that an
industrial-scale user in a small country such
as Ireland, which has to import virtually all
its coal needs would be subject to significant
supply interruptions, quality variations or
price disadvantages in the longer term. This
vulnerability is increased where there is a
need to burn a specific and particularly a
higher grade coal. It was calculated that
there would be outstanding benefits both
commercially and technically in having a
combustion process which was flexible enough
to utilize a wide, and if possible, the full
range of coals and that it could well be
justified to invest in such a capability.
While pulverized coal combustion offers
reasonable versatility on coal use FBC offered
even more and is not at an economic disadvantage
in the industrial boiler range.
1.2 Peat
At the same time, there is a significant
local low-grade fuel resource in the form of
peat. The locality is rich in peat bogs but
these are poorly developed and likely to remain
so until an efficient harvesting technology is
applied and a large stable market is available
locally. This market must not only be local
with a large base but be capable of absorbing
the inherent fluctuations in quality and supply
which arise due to the wet Kerry weather. Only
an industrial-scale consumer could provide such
a scale market but the fluctuations mentioned
are technically unacceptable to any conventional
furnace. While research identified a suitable
economic harvesting technique a new type of
boiler capable of coping with such conditions was
required.
189-
What we wanted was a furnace which could operate
efficiently on sod peat and its normal quality
variations but which could remain in operation
ev.en when peat supplies became erratic as they
undoubtedly would from time to time. Since
this furnace/boiler is applied in a manufact-
uring process environment where steam is only
a service which must be available as required
at all times, any shortfall in peat must always
be compensated by another fuel. This other fuel
is best chosen as coal as it is the most easily
available in practice. Thus a furnace is needed
which can operate on both coal and peat and
over a wide range of blends since the peat is
notoriously erratic and its storage is problem-
atic. This is a particularly difficult
requirement as coal and peat have very different
combustion characteristics, and for conventional
combustion systems the necessary furnace design
for one would be incompatible with the other.
Substantial research into conventional and

developing furnaces led to the conclusion that
not just FBC but fast circulating FBC (second
generation) was the best and perhaps only
available technology to yield this degree of
flexibility.
1.3 Other Indigenous Fuels
At the same time, there are other low grade fuels
available regionally.
In common with many other countries and indeed
more so than many, Ireland has a significant
outfall of waste wood for which there is no
developed market. Additionally, there is
potentially a much larger supply in the future.
It was calculated that this, like peat, is of
economic and strategic interest. Wood is thus
seen as an alternative or supplementary fuel to
peat.
Additionally, there are coal shales and other
problem coals, none of which are economically
exploitable using conventional methods of
combustion but which could be handled via FBC.
Thus again, by creating a viable industrial
market for these coals, it becomes feasible to
exploit otherwise idle low-grade reserves which
can act as import substitutes.
-190-
1.4 Opportunity/Miscellaneous Fuels
It is desired to have technical and commercial
access to the 'opportunity' fuels which can
become available nationally or internationally
even though for limited periods e.g. petroleum
coke. These of course would not be burnable in
conventional furnaces. Testwork to date has
highlighted second generation FBC as the most
efficient means of burning this unreactive
material.
1.5 Summary
Research led to the conviction that we should
invest in the system that gave the widest
possible choice of fuels in terms of technical
capability to handle and burn them efficiently
and that this would realise a worthwhile
strategic and commercial advantage.
FUEL FLEXIBILITY - DESIGN AND DEMONSTRATION

Having identified the fast circulating FBC as the
most appropriate process, the multisolid fluidized
bed (MSFBC) developed by Battelle/Foster Wheeler
appeared the most promising and this was chosen as
the demonstration means or plant. The design data
has been generated from extensive pilot-scale work
at Columbus, Ohio to larger scale plants and
empirical data from the most recent - an industrial-
scale steam generator built for Conoco in Texas to
about half the size of that to be built by Kerry -
has been applied in designing this demonstration
plant.
It is our intention to demonstrate how this process

can efficiently cope with a wide variety of
technically different (and often incompatible)
fuels over a wide range of blends. In specifying
this range it was decided to regard coal as the
base fuel, it being the most available and most
developed. It is designed then to handle up to
1001 coal. For logistical and technical reasons 301
coal was chosen as the minimum blend. Normally a
blend will consist of coal plus one other fuel, the
prime supplement being peat, though it could also be
wood or any of the nominated alternatives.
-191-
It is of course no coincidence that peat, in common
with other 'agricultural' harvests, is a seasonal
crop and thus becomes available at the same time
when th,e milk processing season and steaming
requirements reach their peak. Thus there is no
difficulty anticipated in being able to supply up
to 70$ of the fuel blend (in heat terms) as peat
even at maximum output. H owever, as peat does not
store easily, buffer storage capacity will be limited
and this may mean wide variations covering much of
the blend range 0-704 over cycles of a few days. It
is intended to demonstrate how the MSFBC can
efficiently cope with such variations in blend as
well as in fuel quality.
Outside of the peat season or perhaps even in season,

wood is seen as a likely replacement for peat. The
plant is so designed to cope with either peat or
wood which are in any event very similar in
combustion terms. Also the opportunity fuels etc.
will be purchased and used as determined by price
and availability prevailing from season to season.
It is aimed to demonstrate how flexibly the plant
can technically exploit such opportunities.
Finally to maintain maximum versatility, it was

required to burn coals, etc. with minimum
preparation - e.g. in the as received condition -
and to avoid precrushing, etc. Of the fast
circulating fluidized beds, the MSFBC because of
its dense bed held most promise here allowing
(coal) feed sizes up to at least 2in (50mm) to be
used. It is intended to demonstrate this
capability.
PROCESS REQUIREMENTS
The new steam raising facility is to be located in a
large centralised milk processing complex whose
business is to manufacture a wide variety of high
added value food products. It consists of a number
of relatively independent processing units each with
its individual requirement for process steam.
However, quality and output of the end product is of
paramount importance and steam is used very much as
a service toward that end.
192-
3.1 Load Response
Operations are such that steam demand can be
highly variable in each sector. Large steam
flows can be demanded or shut off quickly. If
product quality is to be maintained, very fast
boiler response is necessary to service
(positive) demand. Thus, steam demand is
fluctuating and dynamic load response demanded
is high"
3.2 Turndown
Additionally milk processing in Ireland is
highly seasonal with a high plateau for about
four months in summer, a low trough in winter
and a relatively steep increase or decline in
spring or autumn respectively. Thus, in
addition to the inherent requirement for (short-
term) low turndowns associated with a fluctuat-
ing load, there is the additional seasonal
burden which calls for turndowns below 10$ in
mid-winter. Now, demand is at such a low level
and so erratic in mid-winter that it becomes
expedient to shut down any large boiler for
annual maintenance at that time. But it has
been calculated that there is a substantial
amount of operating time outside of such a
shutdown where turndowns down to at least 15%
of MCR are required. Thus, there is a definite
requirement for very high turndown.
3.3 Tackling Requirements
For Load Response and Turndown
It is planned that the existing oil fired plant
much of which is relatively new, will be retained
as a strategic standby in the case of non-
operation of the MSFBC - e.g. during annual
shutdown or otherwise.
This is entirely satisfactory as this represents

capacity for virtually 1001 of peak steam
requirements. With such strategic standby
available, it is permissible that the MSFBC
boiler be a single unit.
Now, the requirements for load response and
turndown initially seem so demanding as to so
warrant that a number of units be employed on an
on/off sequence modulation. However, capital
costs alone preclude such a luxury. It is
necessary that a single unit be employed. The
implications of failing to achieve the required
levels are wide-ranging and justified significant
on-site research to define and quantify the
problem.
-193-
Empirical research showed that a dynamic load
response capacity to cope with load changes
aobve 10% per minute was definitely required a
significant amount of the time. It was deter
mined that response capacity of at least 104 per
minute was necessary and that up to 154 per
minute was desirable.
Failure to achieve these response rates on a

fast upturn (steam demand) would result in the
need for the standby oil boilers to be maintained
on continuous warm standby with a sensitive and
sophisticated overriding control enabling their
rapid firing in the event of a shortfall in steam
supply. In effect, this would be costly, complex
to operate and result in significantly reduced
oil displacement. On the other hand, failure to
achieve a fast turndown (steam shedding) would
result in frequent lifting of safety valves with
according losses,.wear and noise.
At the same time, failure to achieve stable

operation of the solidfired boiler at sustained
low turndowns would again result in enforced
firing of the standby oil boiler with substantial
reduction in oil displacement and therefore
savings, etc.
Now, our requirements for fuel flexibility
warranted use of FBC but the first generation
bubbling bed type would not attain the performance
sought for process service. However, fast
circulating FBC inherently held good promise of
achieving it and of these it was felt that the
Battelle/Foster Wheeler MSFBC was the most capable
for the reasons set out below.
In fact, the plant has been designed for load

response exceeding 104 per minute and up to 154
per minute and to achieve turndown on automatic
operation down to 154 of MCR with the possib
ility of reaching lower on manual control. It
is intended to demonstrate this process capabil
ity.
Finally, the inherent capability of the MSFBC to

control emissions, particularly nitrogeneous and
sulphurous, is of immense interest and it is
aimed to evaluate this aspect also.
194
4. DESCRIPTION OF THE BOILER PLANT
4.1 General
The MSFB system proposed for this project is a
variant of 'circulating' or 'fast' fluid bed
technology developed by the Battelle Institute
and engineered by FWPP. The system was
considered to be that best suited to meet the
requirements of Kerry Coop in providing steam
generating plant for a multifuel capability
and was also able to satisfy the steam demands
of the plant for operating over a wide load
turndown range and maintaining a rapid load
response capability. The main feature of the
system is that by incorporating a fixed 'bubbling'
bed and a superimposed entrained bed with the
firing taking place in a combustion chamber under
isothermal conditions, the heat transfer and
combustion processes are effectively separated.
The unit is designed to produce 105,0001b/hr
(47,620kg/hr) of saturated steam at 24 Bar G on
the principal fuels which are coal and coal/peat
mixes.
4. 2 Combustion System
All combustion of fuel takes place in the
refractory lined combustor vessel which does not
contain any heat transfer surface. The combustor
vessel operates as a fast or circulating bed, in
which the bed material, consisting of fine
particles, is transported vertically upwards
through the vessel by the combustion air and flue
gas flow. After leaving the top of the combustor
vessel, the fine bed solids are separated from
the flue gas flow by high efficiency hot gas
cyclones. The separated solids leave the bottom
of the cyclones, pass through the external heat
exchanger, and are returned to the bottom of the
combustor vessel to complete the solids circul
ation system, via a nonmechanical 'Lvalve' and
pipe arrangement, which controls the flow of
solids.
The combustor vessel is divided into two zones
with different cross sectional areas:
1. A lower reducing zone in which primary
combustion air is introduced at superficial
fluidizing velocity of about 7m/s. The
quantity of air being less than half that
required for stoichiometric combustion of
the fuel.
195
2. An upper oxidising zone in which the air
required to complete the combustion is
introduced through nozzles in the vessel
wall. This zone normally operates at a
superficial fluidizing velocity of about
9m/s.
In addition to the fine circulating bed material
the lower reducing zone contains a 'bubbling1
fluidized bed of large particles, with a size
range between 6.5mm and 13mm. This dense bed
has the effect of greatly increasing the
residence time of the fine bed particles of
ash and inert material, ensuring the maximum
heat transfer to the particles and maximising the
combustion efficiency of the system. The dense
bed is violently turbulent and ensures good
mixing, reducing the requirements for multiple
fuel feed points and in addition, allowing for
fuel being bed in lump form (50mm), thus
reducing the requirements for fuel preparation.
When firing a fuel with a high sulphur content,
crushed limestone is substituted for the fine
inert material and because of the good mixing and
high residence time in the system, desulphuris-
ation of the flue gas can be achieved in an
effective and efficient manner.
After leaving the combustor, the hot gas is
passed through a high efficienty cyclone of
conventional design which splits the solids
stream from the high temperature flue gases which
are then ducted to a natural circulating
convective boiler system to generate a proportion
of the total steam required.
Final heat recovery is accomplished by the use of

an economiser located down-stream of the boiler
and flue gas clean-up to reduce particulate
emissions to an acceptable level is effected by
the use of a multicyclone grit collector.
The combustion air/flue gas system operates on
the balanced draught principle with an induced
draught fan, located down-stream of the final
gas clean-up equipment and discharging flue
gases to the chimney.
-196-
The hot solids separated out in the cyclone
pass to the external heat exchanger (EHE) in
which the heat is removed from the fine
circulating bed particles in a low velocity
fluid bed containing heat transfer surfaces
returning to the combustor through the 'L-valve'
solids return system. The EHE tube bundle
generates approximately half of the steam
output of the plant.
The low fluidizing velocity in the EHE with no
combustion in the bed reduces any risk of
corrosion/erosion of the heat transfer
surfaces. The fluidizing air is set at a
constant mass flow and as boiler load varies,
the rate of solids recycle is varied by cont-
rolling the 'L-valves'. This then causes the
EHE bed temperature to change which in turn
alters the rate of heat transfer to the tube
bundle, and consequently varies the amount of
heat extracted. Thus, the combustor optimum
operating temperature can be maintained at a
constant value over a wide load range.
The EHE is divided into two sections to provide
'hot' solids recycle from an uncooled section
and 'cold' solids from the heat exchange
section.
Variation of the cold solids recycle rate can
achieve a load turndown range in the order of
6:1 under automatic control whilst maintaining
a constant combustor operating temperature. By
reducing the combustor operating temperature
and adopting changes in control philosophy,
significant further load reductions can be
achieved.
4.3 Fuel, Ash and Bed Material Handling
In order to provide a high degree of flexi-
bility in terms of fuel quality, moisture
content and sizing, a mechanical fuel handling
system has been prepared. Peat in sod form is
delivered by front end loader to an intake
hopper where it is carried by a fully enclosed
'en masse' chain conveyor to a crusher unit
which reduces the peat size to about 50mm
maximum. The crusher discharges into a common
coal/peat 'en masse' elevator which empties
into a 10,600ft C300m ) peat storage silo.
197-
Coal of SOmm-zero size is delivered by tipper
truck to a ground bunker from where it is
carried by a fully enclosed 'en masse' chain
conveyor directly to the common coal/peat
elevator which also empties into a 300m coal
storage silo.
From the coal and peat storage silos, the two
independent fuel handling systems are identical
apart from machine sizes and capacities. Each
storage silo has a flat bottom and fuel discharge
is effected from the centre by a screw which
rotates around the full floor of the silo as
well as about its own axis. Enclosed 'en masse'
elevators take the fuel from the silos and empty
into 800ft (23m ) surge hoppers. The surge
hoppers have conical bottoms and fuel flow is
ensured by a circular bin.discharger which
comprises a steel shaft with a number of rake
arms attached, and which rotates around the
hopper bottom as well as about its own axis.
The surge hoppers discharge into drag chain
feeders which have variable speed controlled by
the steam demand. The fuel feeders discharge
into the fuel inlet pipes to the combustor.
The ash handling system is a pneumatic lean
phase vacuum system which collects ash and fine
bed material, as appropriate, from the EHE drain,
economiser hoppers and multicyclone hoppers as
shown on Fig. 1. The..pneumatic system discharges
into a 3,500ft (100m ) ash storage silo from
where it is removed by truck after ash
conditioning.
PROGRAMME
Planned start-up is March 1984. The evaluation
programme will continue until late 1985. During
this period we will examine fuel flexibility with
different fuels and a range of blends, boiler
efficiency on different fuels and blends and
turndowns, carbon utilizations, load response
capacities at different fuel blends, turndown
capacity on different fuel blends, power
consumptions, environmental aspects , wear rates
and maintenance requirements.
198-
_MAI STEAM
K M Ml
mil* mm* " TO PROCESS
SH0Tl
3 Off-[HE
CMCUUTIM PUMPS
4 I STAKOIY I
*o? INDUCED
OIAUCHT rAN
VACUUM I
_FIITE UNIT
CONDITIONS OF OPERATION
FUIl COAl ADO PEAT
O I I E I EVAPORATION - 41*20 k|/k IIOS.000 Ik/kr)
STOP VALVE PnSSUAE - 24 kir t | 3S0 p u l l
STEAM TEMPEIATUIE - 224C i l l , 1415'F H I I
FEEOUATEft TEMPEIATUIE - tO'C IMO'FI
EICESS All - !/.
STACK CAS TEMPEMTUIE - 112' C ( J24*FI
C C.V Of COAl - 21.794 k j / k | | 11.500 k t i / l k ) ASH COIOITIONEI
EfflCIEUCr-COAl n u n c - Vil'h oil CCV
COAl COISUHPTIOI - SIIOk|/k 111.420 Ik/k I
COMIUSTOI TEMPEIATUIE - tOO'C lUSO'fl
FIG. 1 - KERRY CO-OP SYSTEM DIAGRAM

DISCUSSION

r would like to ask Mr. O'Brien why he chooses this parti-
cular fast fluid technology. Is it for commercial or for
technical reasons?
If it is technical can he comment? He mentioned the high re-
sidence time characteristics and the fuel size characteristics
and possibly lower erosion. I was wondering whether there are
any others. I would say that we think that he is right because
the National Coal Board in the UK has signed an agreement with
Foster Wheeler for the development of the system in the UK,
but I had been interested in the expansion of the technical
views that Mr. O'Brien has on the differences of the diffe-
rent technologies.
Answer There is always a commercial element in any choice and that

was not the primary stimulus. The primary considerations were
technical and I would, stress that while we did Look at alter-
native systems our particular process requirement for a very
fast low response and a very high degree of turn down were
probably dominent in the early stages. Fuel flexibility can
be achieved by many of the processes being developed or on
offer but I do believe however that this process does give us
a slightly wider fuel flexibility in the addition of the dense
bed which is a trick if you like. It is additional to the fact
that circulating bed is being offered by others and I think
that it does in fact give us the possibility of slightly in-
crease turn down and higher fuel efficiency over a wider ran-
ge of varrying operating conditions. There are other features
such as the separation of the combustion from the heat trans-
fer and also the corrosion and erosion problems.
Question - Mr. Folliot, Creusot-Loire :

- What is the moisture content of the peat being used?
- What is the sizing of the sods and what is the advantage of
the extrusion?
- What is the calorific value of the peat?
- What is the importance of the seasonality of the peat?
Answer - The moisture content of the peat is highly variable and is

Largely dependent on weather conditions. It is expected to
fluctuate between 35% moisture and 50% moisture with an
average of about 42.5%.
- One of the major breakthroughs for us in recent times was
the development of a peat harvesting technology which could
economically exploit the many but relatively small and topo-
graphically difficult bogs in Kerry. This development uti-
lized a small tractor driven turf cutter which produces an
extruded sod. This sod is typically 100 mm diameter and per-
haps 200 mm long when first cut. One of it's main advantages
is that it dries quickly and in doing so will shrink to per-
haps 50-60 mm diameter. Now although the fluidised bed it-
self could probably handle such sods in their original dried
-200-
condition, it is desirable for storage and handling purposes
for these sods to be crushed to a 50 mn maximum
dimension. This is achieved by an inline peat crusher
operating on the intake and before the storage bin.
The calorific value of the sods is such that for an average
moisture content, 3.67 ts turf equal 1 t HFO.
The factory in which the boiler operates, being a food pro
cessing unit, has its seasonal peak in the summertime. In
the wintertime on the other hand, production is at only a
fraction of peak output.
Fortunately, the production of peat which is again an agri
culturally related activity, also peaks in the summertime.
Since for technical and logistical reasons sod turf does not
store easily, plentiful supplies of turf are available only
in the summer and autumn. As most steaming occurs in sum
mer there results a potential for a high level of peat usage
in lieu of coal.
Question Mr. Willis, N.C.B. :

What will be the price of the peat?
Answer We have calculated that the total cost of extraction, hand

ling and transport will be significantly lesser than the cost
of imported coal per unit of useful heat. The relationship
between price and cost depends on the extent to which we as
a company or as a cooperative become corporately involved in
the business of peat production. The price also depends on
quality and particularly the level of contained moisture.
Overall, however, we believe that the price of peat will com
pare favourably with coal.
Question Mr. Knobbout, expert :

Did you consider also a storage system of steam in the
design?
Answer You are asking whether we looked at the alternative of steam

accumulators and indeed we did and no matter what design of
accumulator we thought of it was uneconomic. It did not give
us a competitive answer which allowed Fosters to look toward
this type of boiler.
201
FLUIDIZED BED COMBUSTION OF WASTE ACID TARS
Author : Eddie ROBINSON
Head of project Eddie ROBINSON
Contractor : DEBORAH FLUIDIZED COMBUSTION LTD.
Address : 6 Davy Drive

N.W. Industrial Estate
Peterlee
Co. Durham SR 8 2 JF
UNITED KINDOM
SUMMARY
This paper briefly describes the development, design and operation

of a circulating fluidized bed boiler plant. The plant installed
at Avonmouth, England, produces process steam from the combustion
of waste acid tars and waste oils. It avoids-the cost of waste
disposal and reduces the works' fuel bill for the production of
process steam. The results from two years successful operation
together with a brief specification are included.
202-
1.0 SUMMARY
This paper briefly describes the development, design, and the operation
of a circulating fluidlsed bed boiler plant. The plant installed at
Avonmouth, England produces process steam from the combustion of waste
acid tars and waste oils. It avoids the cost of waste disposal and
reduces the works' fuel bill for the production of process steam. The
results from two years successful operation together with a brief
specification are included.
2.0 INTRODUCTION
Throughout the chemical process and allied industries as production

increases so do the energy requirements and arisings of waste materials.
To produce this energy and dispose of the waste, manufacturers invariably
use the means which are proved to be most economical. In some cases,
Incentives to explore alternative means of producing energy, or utilisation
of the waste by "ln-house" processing do not exist. However, considering
the growing economic constraints and the realisation that our energy
resources are by no means inexhaustible, there is an increasing awareness
of the need to make full use of all waste materials since today's waste
could be tomorrow's fuel.
At the present time the predominant waste disposal routes in use are
tipping on authorised landfill sites and combustion.
Since the impetus given to development of new technologies depends

primarily upon economics, tipping has been the most popular route,
because of its low first cost. However, it is in the area of combustion
where the recovery of potential thermal energy from the waste now yields
superior rewards.
Prejudices against combustion which in the past may have existed due to
comparatively high capital cost and equipment availability have now
been overcome by the vigorous technical development of new combustion
techniques. The technique which is causing the greatest impact is the
application of Fluidlsed Bed Combustion.
3.0 FLUIDISED BED COMBUSTION (F.B.C.)
In conventional fluidlsed bed combustion systems (see Fig. 1) air passes

upward through a mass of finely divided particles. At low velocity the
particles are not disturbed, but as the air supply increases the particles
become suspended and Increasing velocity gives rise to bubble formation
and vigorous turbulence. The suspended mass has the same characteristics
as a boiling fluid, hence the term fluidlsed. Ideally the fluidlsed bed
when hot provides the necessary requirements for efficient combustion,
i.e. Uniform temperature, Turbulence, Residence time. However, as the
air passes vertically through the bed lateral mixing is limited, giving
rise to problems of poor fuel distribution, uneven temperatures,
stratified gas streams, elutriation, and ash removal. Consequently the
combustion of fuels, especially volatile liquids and fine matter, requires
numerous feed points to achieve satisfactory burn-out and uniform
temperatures. Also when burning materials which have extraneous incom-
bustible ash, the base of the bed needs to be fitted with several ash
extraction points.
-203-
4.0 THE CIRCULATING FLUIDISED BED (C.F.B.)
To overcome these problems a "second generation" of fluidised bed was

conceived and patented in 1969/70. It is from this patent that the
Deborah'Fluidised Bed (DFB) is derived (Fig. 2 ) . In the circulating
bed the feedstocks, gases and bed material are circulated in a roughly
elliptical path, created by differential air flow supplied through
separate plenums and an inclined distributor. This produces controlled
lateral mixing and vertical turbulence, which is not practical in a
conventional bed. A novel feature of the combustion chamber is the
inwardly inclined wall above the bed which confines circulation within
the chamber and controls bed expansion at high and low firing rates.
Further air injection may be used below the deflector to assist in
combustion and circulation.
Entry for liquid fuels is via one or two simple open pipes without
nozzles or atomisation, and solid fuels only require a single feed
point. Extraneous ash gravitates to the ash outlet facilitating
removal. The process gives rise to the following advantages:- rapid
distribution of fuel, uniform temperature, high once through combustion
efficiency, elutriation minimised, multifuel capability, simple fuel
feed and ash extraction systems.
5.0 DEBORAH FLUIDISED BED BOILER
The DFB Boiler was developed during four years intensive research and
finally patented in 1978. The experience gained from two incinerators,
a cold model and a converted boiler produced data for the prototype
boiler installed at Avonmouth.
An important feature is that the boiler has been specifically designed

for the fluidised bed, optimising operation and avoiding an ungainly
configuration. The design shown in Fig. 3 brings the membrane water-
tube walls into direct contact with the fluidised bed so that expansion
of the bed, under all operating conditions, exposes the correct amount
of heat transfer surface for temperature control without a requirement
for additional in-bed tubes. The resulting combustion chamber design
produces a high specific capacity, i.e. 1.5 Mff/m3. Some 1.7 tonnes of
bed material provides a comparatively large heat sink, which can hold
temperature for long periods during shut down enabling a fast restart.
Above the combustion zone is a 2m high freeboard which acts as an
expansion zone for elutriated material, the majority of which falls
back into the fluidised bed after losing its entrainment velocity,
disengagement being assisted by membrane wall geometry producing an ideal
gas flow path. The hot gases flow through two vertical passes of a
convection bank, causing additional solids removal which is classified
and recycled back to the bed, the fines being discarded.
The design provides the following advantages:
(a) The DFB Boiler will burn any type of fuel efficiently and
automatically, provided the G.C.V. is not less than 11,650 kJ/kg
-204-
(b) There ia no requirement to use additional heat transfer
surfaces iamersed in the bed to control temperature.
Boiler output is produced by absorption through the con-
tainment walls only.
(c) Control of output is simply effected by varying the air and

fuel inputs within the range of 2.0 to 1.
(d) When slumped the bed retains its heat for long periods,
reducing the requirement for start up fuel and maximising
response.
(e) Automatic de-ashing of the bed is carried out on load with

minimal effect on output.
6.0 AV0NM0UTH PLANT
The plant installed at Avonmouth includes a single module DFB water -

tube boiler, suitable for continuous operation.
Normal boiler operation involves the production of approximately

3,000 kg/h of steam from the combustion of acid tars and waste oils.
The main purpose of the plant is to dispose of the tars and produce
steam. Control of fuel'inputs, therefore, depends upon the production
and subsequent burning rate of the acid tar.
From a knowledge of tar production and its calorific value, the input
to the boiler is set manually to suit requirements. This usually
provides about 60% of the required heat input. Because all steam
produced is taken, the controlling parameter is bed temperature. The
remainder of heat input is provided by waste oils, the rate of which
is controlled automatically to maintain the required bed temperature
and steam production.
Boiler start-up from cold requires the use of two conventional oil
burners. One pre-heats the combustion/fluidising air being supplied to
the boiler, the other is situated above and its flame directed onto
the surface of the bed material. These two burners increase the
temperature of the bed particles to a point when gas oil can be in-
jected directly into the bed (c. 300C). Increasing the gas oil rate
raises bed temperatures to 600C when the waste oils can be supplied
Instead of gas oil. At 650C the pre-heat burner is switched off. At
700C acid tars can be introduced, provided that the stack liner
temperature is sufficiently high. An interlock will prevent tar
being supplied to the boiler if the stack skin temperature is not
above the dew point of the flue gas. This is to avoid excessive
corrosion caused by condensation of sulphuric acid from the sulphur
compounds formed during combustion.
At 7S0C the overhead burner is switched off. The tar burning rate is
then fixed and bed temperatures maintained around 800C on waste oil.
-205-
Start-up from cold to continuous on load steam production at 17 bar
gauge takes about 4 hours.
Plant status is displayed on a mimic installed in the main control

panel. ,
6.1 Feed System
The primary factor in the efficient combustion of liquids in a

fluidised bed is to ensure intimate mixing of the liquid fuel and
the air, since its residence time is short due to rapid vaporisation.
To avoid "glueing" or coking the bed, the combustion reactants must
be thoroughly mixed with the whole of the bed material as soon as
possible.
In conventional fluidised bed systems attempts to achieve this

condition are made by installing many injectors, some quite elaborate,
across the whole area of the air distributor.
In the DFB the inherent vigorous mixing and turbulence removes the
liquid away from its injection point immediately, causing rapid
diffusion throughout the bed. This permits a minimum number of
injection points to be used and only requires a relatively crude
form of fuel gun. Liquid fuels are virtually poured into the bed
through open bore pipes, acid tar through 12mm bore tubes and other
less viscous liquids through 6mm bore tubes.
This system maintains the mechanical simplicity of a fluidised bed,

and allows flexible and reliable boiler operation even when feeding
liquids containing suspended solids.
6.2 Acid Tar Combustion
Our client at Avonmouth, Tenneco Organics Ltd., is involved in the

re-refining of waste lubricating oils, particularly from the auto-
motive industry. Ash contents of waste oil from petrol powered .
engines average about 1-2%, with a lead content representing 35-
50% of the total ash. One of the re-refining processes employed
is to contact the waste oils with sulphuric acid to precipitate
metal bearing compounds, leaving an acidic sludge with the following
representative analysis. (See Table I)
This is difficult, if not environmentally impossible to burn in

conventional combustion systems, because of the lead vaporisation
in hot regions of the furnace.
Deborah Fluidised Combustion Ltd. carried out investigations into

acid tar burning which revealed that very little vaporisation of
the lead occurred at temperatures below 900C, however, beyond
this a dramatic increase occurred. Therefore, uniform combustion
temperatures below 900C are of paramount importance when considering
this type of waste as a fuel. (See Fig. 5)
-206-
The DFB satisfies this prerequisite by providing an isothermal
environment of _ 10C throughout the whole bed volume (1800mm
long z 900mm wide x 760mm deep). This control parameter becomes
more important following boiler operation of more than about 2
weeks on acid tar, because of a coating which forms on the starter
bed material (see Table II), and if temperatures anywhere in the
system exceed the fusion point of the coating the condition
manifests itself as friable agglomerations of bed material.
It is necessary, therefore, that efficient combustion takes place

within the bed where the environment is relatively easy to control,
rather than have a situation where freeboard combustion can produce
significant temperature differentials.
6.3 Emissions
The measurable lead content in the flue gases is present as

particles submicron in size, which pass through the high efficiency
multicell cyclone. The quantity, however, is still acceptable and
within the limits imposed by the Alkali Inspectorate. (See Table
III).
CONCLUSION
The circulating fluidised bed water-tube boiler plant at Avonmouth

burns petrochem. waste tars/oils at high efficiency and within stack
emission limits. The plant operates continuously as a process plant
within the context of high standards of engineering, control and safety
demanded in refineries. Since it has operated commercially for 2 years
it can be considered a successful demonstration of energy conservation.
A typical cost benefit analysis indicates a payback of less than 2 years.
(See Table IV).
As further evidence that the project has been successful Deborah are now
involved in the supply of a similar plant to Technipetrol of Italy. The
versatility of the system is also being demonstrated by three further
contracts involving the combustion of tyres, sawdust and coal.
-207-
TABLE
ANALYSIS OF ACID TAR
Carbon C 34.90% w/w

Hydrogen H 5.50
Oxygen 0 6.35
Nitrogen N 0.30
Sulphur S 2.12
Sulphuric Acid H2S04 . 37.80
Moisture H20 9.00
Ash 4.03
100.00
Lead in Ash 25 to 50% w/w

Viscosity @ 70F 5,000cP
Specific Gravity 1.3
Calorific Value 18,648 k J A g
TABLE I I
Particle Size Starter Bed Used Bed
Greater than 2mm 0 2%

Less than 2mm, greater than 1.68mm 0 6%
Less than 1.68mm, greater than
1.41mm 13% 36%
1.0mm 43% ' 47%
0.85mm 15% 8%
Less than 0.85mm 29% 1%
-208-
T A B L E I I I
Emission Standard Set Plant Performance
Particulates 0.48g/m3 0.34g/m3

Lead 0.12g/m3 0.1g/m3
S0 2 114kg/h 63.8kg/h
3
S03 0.12g/m not detectable
Plume Ringelmann 1 within limit
T A B L E I V
Cost B enefit Analysis (1982)
Cost
Plant Installed 268,000
Annual Operating Cost (labour, maintenance,
utilities) 37,000
TOTAL 295,000
Benefits
Disposal Cost Avoided (2500 tonnes 6 50/
tonne) 125,000
Reduction In Fuel Bill (1080 t.o.e.
12.5p/l) 143,617
TOTAL 268,617
Payback 295,000
268,617 1.1 years
209-
Liquid/Solid Liquid/Solid
Fuels Fuels
Figure 1 - Conventional Bed
Fine Solid
Start-up
Fuel
I
Burner
'^ / 'St f I ill 'J I / s

L Solid
' Fuel
Liquid V* " ^T"_ '

Fuel "
y
--!_
Ash/Bed*
s
Figure 2 - C i r c u l a t i n g F l u i d i s e d Bed
-210-
Figure 3 -
Water Tube
Boiler
Figure 4 -
Sections through
Boiler from
Figure 3
Figure 5 -
Graph illustrating
Extent of Lead (Pb)
Vapourisation at
Various Temperatures
oo 700 aoo OO 1000 1100

Tnpvratur* c
211-
BRIEF SPECIFICATION AND PERFORMANCE DATA FOR CIRCULATING
FLUIDISED BED FIRED STEAM RAISING PLANT AT AVONMOUTH, ENGLAND.
Feed Stocks , Waste Acid Tars from oil rerefining

process and waste oils.
Characteristics Tars: G.C.V.s between 12,000 and 22,000kJAg

Sulphuric Acid content not more
than 50% w/w
Ash content not more than 10% w/w
S.G. 1.1 to 1.3
Oils: G.C.V.s between 27,000 and 45,000k J A g
H2O content not more than 30% w/w
Ash content not more than 5% w/w
Combustion Process Circulating Fluidised Bed
Optimum Thermal Capacity 10,000,000 Btu/hr 2.93 IW
Maximum Thermal Capacity 12,000,000 Btu/hr 3.517 MW
Minimum Thermal Capacity 6,000,000 Btu/hr 1.758 MW
Maximum Burning Rates Specified Tar 13.6 Tonnes/day

Specified Oils 9.0 Tonnes/day
Combustion Air Supply Forced draft fan 63 nm3/min 1650mm w.g.
Air Distribution Via three separate plenums and tangential

injection. Plenums are fitted with pre-
cision drilled stainless steel distributor
plates and stainless steel woven wire mesh
as sand trap.
Fluidised Bed Dimensions Length 6' 0" (1800mm)

Width 3' 0" ( 900mm)
Depth 2' 6". ( 700mm)
Mass 1.7 Tonnes (1700kg)
Bed Material Derived from BS63 single sized road stone

Basalt group size range ^.2" (2mm) to T^g"
(1.5mm) and tested to BS812.
Bed Make-Up Automatic bed addition by bed sensors via

pneumatic transport from hopper
Bed Re-Cycle Automatic bed re-cycle after discharge from

boiler momentum separator via vibratory
classifier and pneumatic transport.
Boiler Specially designed membrane wall water-tube

boiler (patent no. 1604998) includes integral
soot blowers and momentum separator.
212-
Boiler Dimensions Length 14' 0" (4500mm)
Width 6' 6" (2000mm)
Height 16' 0" (4900mm)
Dry Weight 11 Tonnes (11,000kg)
Gross Weight 16 Tonnes (16,000kg)
Boiler Capacity Max. steaming rate 10,0001b/hr 4545kg/hr

Min. steaming rate 4,0001b/hr 1818kg/hr
Steam conditions 250p.s.i.g. 17.24 bar
Blowdown 6.6% steaming rate
Overall efficiency 75% to 80% depending
on fuel
Start Up To Full Load Time From cold 4 hours

After weekend shutdown 2 hours
Feed Water Quality Total hardness as CaC0 3 ppm 0

De-Alkallsed Caustic Alkalinity as CaCO* ppm 300
Total Alkalinity as CaCO* ppm 700
Sulphite as Na 2 S0 3 ppm 40
Phosphate as Na2P0 4 ppm 50-100
T.D.S. ppm 3000
Flue Gas Handling Multicell Cyclone 98% efficiency on 15 micron

discharging to sealed hopper
I.D. Fan 140 am3/min 6 -150mm
Start-Up Burners Two, dual fired oil or natural gas
Burner Capacity Preheat 500,000 Btu/hr 146.5 KW

Overbed 1,000,000 Btu/hr 293 KW
Start-Up Fuel Required From cold 272 litres 35 sec oil

or 98 therms natural gas
Control It Instrumentation Semi-automatic control via comprehensive

instrumentation, panel includes mimic dis-
play with full visual and audible alarm
annunciator system providing failsafe
operation to petrochemical plant standards.
Additional safety features included as
recommended under N.F.P.A. codes of
practice.
Utilities Used Bed material 1 kg/hr 6 M.C.R.

Electricity 45 KW/hr 0 M.C.R.
Labour 0.25 manhours/hour
Maintenance
required 1200 manhours/year
-213-
ATMOSPHERIC FLUIDIZED BED COMBUSTION BOILER
FOR ELECTRIC POWER GENERATION BY COAL
Authors : G. BERTONI, L. GUIDI
Contract number EE-190/80
Head of project G. BERTONI
Contractor : ANSALDO
Divisione Breda Generazione Vapore
Address : Viale Sarca, 336

I - 00126 MILANO
SUMMARY :
Fluidised bed combustion technology has the potential to overcome

or reduce some problems related to coal utilization in stoker-fired
or pulverized coal-fired boilers for electric power generation.
ANSALDO and F. TOSI are carrying out, in cooperation with ENEL and
with the technical support of Fluidised Combustion Contractors Ltd,
a demonstration project for an atmospheric fluidised bed boiler
able to process high sulphur content coals, such as Sulcis coal
(Sardinia).
The boiler produces 80 t/h of superheated steam, at 475 C and 59
bar, to feed a steam turbine-generator of about 15 MWe.
The plant, in addition to the AFBC boiler, consists of the main
following subsystems : coal-limestone feeding, air feeding, start-
up, gas clean-up, fines recycling, ash withdrawal and control system.
In this project special attertion is paid to plant reliability and
controllability to ensure adeguate load following comparable with
other types of boilers. An integral part of the project is a two
year experimental operation to demonstrate the boiler performance
when feeding different coal-limestone mixtures and the reliability
of boiler critical components and auxiliaries.
-214-
1. INTRODUCTION
Fluidised bed combustion technology is among the most important

innovations in combustion in the last decade.
In a fluidised bed combustor the bed generally comprises particles
of coal, ash and sulphur acceptor (limestone or dolomite) and is
fluidized by the combustion air.
Coal is feed in a crushed form underbed or screened overbed.
Limestone, or dolomite, gives coal the ignition energy and enables
sulphur to be retained as sulphate, reducing SO emission to the
atmosphere.
The bed operating temperature is in the range 750900 C depending
on coal/limestone reactivity and on ash fusibility; if high sulphur
removal is required the temperature is about 850 C.
Heattransfer tubes are immersed in bed to remove the heat from
coal combustion and to control the bed temperature.
Fluidised beds can operate either at atmospheric pressure or at
elevated pressure.
Fluidised bed combustion technology has many advantages over conventio
nal combustion systems mainly :
the performance of a fluidised combustor is relatively insensitive

to the type of fuel used due to the high thermal inertia of
the bed and the absence of flame stability problems
high sulphur content coals can be utilized without recourse to

stack gas scrubbing, because desulphurization is achieved within
the bed; NO x emissions are reduced due to the low combustion
temperature
reduction of boiler capital cost due to higher and more uniform

rate of heat transfer to surface immersed in the bed and reduced
fuel preparation.
2. THE DEMONSTRATION PROJECT
The project that ANSALDO and F. TOSI are carrying out, in cooperation
with ENEL and with the technical support of Fluidised Combustion
Contractors Ltd, involves the design, construction and operation
of an atmospheric fluidised bed boiler, with a capacity of 80 t/h
of superheated steam at 475 C and 59 bar, able to process high
sulphur content coals (such as Sulcis).
The boiler that will be sited in the existing Porto Vesme Power
Station, owned by ENEL and local to the Sulcis coal mine, will feed
a steamgenerator of 15 MWe. B ecause of the industrial size of
the boiler, the project is an important step for development and
commercialization of fluidised bed combustion technology.
213
An integral part of the demonstration project is a two year campaign
of test and experimental operation necessary to scale up the fluidised
bed boiler to a larger capacity as required by the utilities.
The main objectives of the experimental operation are to demonstrate
the utilisation of coal FBC technology for electric power generation,
to check the correctness of the design, to check the reliability
of components, to verify the plant performance, to train personnel
in the operation of a fluidised bed boiler.
During the experimental operation a special attention will be paid
to the performance of critical components; for this purpose the
arrangment of a superheater tube loop in the fluid bed and the coal-
limestone feed by spreaders are planned.
The project is under contract with the Commission of the European
Communities.
The plant commissioning is expected to be completed in 1984.
3. PLANT DESCRIPTION
3.1 General description
The Porto Vesme fluidised bed boiler installation (fig. 1 ) , seen

in air and gas flow sequence, consists of a secondary and a primary
air fan, a combustor divided into four fluidised bed modules complete
with evaporator heating surface, a freedboard, a rising convection
pass with secondary and primary superheater and a downward pass
with primary super heater and economiser.
Gases leaving the boiler pass through a mechanical dust collector,
a primary and secondary air heater, a baghouse filter, an ID fan
and the stack.
The coal, limestone and air supplies to the fluid bed modules are
arranged as four separate streams. From a storage bunker, via
a feeder, the 0-38 mm coal is supplied to a mill which serves the
dual purpose of drying and crushing the coal to below 6350 microns.
The limestone is supplied to the turret of each mill, via a screw
feeder from separate limestone bunker outlets, where it is mixed
with the coal product.-
From each mill coal and limestone are pneumatically conveyed to
the fuel feed nozzles in the modules.
Elutriated material, collected in the mechanical collector and the
baghouse, is reinjected into fluid bed combustor via a feed point
in each of the four modules, or rejected to the ash collection system.
The spent bed material is cooled in _a screw 'conveyor. Two coolers
are provided, each accepts two ash off-take points from two modules.
216-
3.2 Fluidised Bed Combustor
2
The fluidised bed combustor, about 40 m , consists of four modules;
each module has its own windbox/plenum chamber and distributor,
although the bed itself is undivided. The distributor consists
oi d water cooled membrane tube floor on which air nozzles, fuel
feed nozzles, and ash off-takes are fixed.
The fluidising air is introduced into the bed from the abutted windbox/
plenum chamber via a series of holes in the air nozzles. The positioning
of the air holes facilitates the formation of a static layer of
bed material which acts as an insulant to prevent overheating of
the floor.
Each module has eight fuel feed points, ensuring even distribution
of coal over the bed area. Bed height is kept constant by sequentially
discharging spent bed material from the modules; one ash off-take
is provided in each module for this duty.
In the expanded condition the bed operates at a depth of about 1.2
m whilst running at a fluidising velocity of 2,5 m/s at MCR conditions.
The bed operating temperature can be varied between 750 C and 900
C to facilita'e boiler turndown, the MCR design temperature has
been selected at 850 C.
In the bed, evaporator tubes are provided to extract heat and maintain
the bed at the desired operating temperature; within the tube bank
space is provided for the future addition of a superheater tube
loop.
Each bed module is equipped with thermocouple sets which enable
bed temperature to be monitored at different heights.
The freeboard is constructed from membrane walls and is partially

refractory lined to prevent overcooling of the off gases.
The freeboard -oof necks in prior to the flue gases entering the
boiler convection pass.
3.3 Coal and Limestone Supply a/stem
The coal and limestone supply system consists of four coal and four
limestone bunkers each with its own outlets and feeder, four mills,
each mill having four outlets each subsequently feeding vertical
two-way splitters, which finally feed the coal and limestone mixture
via a total of 32 coal supply points into the fluidised bed to ensure
even distribution of the fuel within the bed.
The storage capacity of the coal and limestone bunkers are equivalent
to 18 hours operation of the boiler at MCR.
The feeders deliver coal and limestone at a controlled rate into
the milling plant. Hot air from the air heater is supplied via the
primary air fan into the mill for drying and transportation of the
coal and limestone.
-217-
The mills are of the modified Babcock "E" mill type which have been
specifically designed for this purpose. Each coal supply point is
complete with automatic purging facilities to prevent potential
blockage from fall back of bed material.
3.4 Air Feeding system
Two independent air feeding systems, each with a FD fan, ducts and
the relevant control equipments, are provided.
The primary air system is for coal and limestone dryinc and transporta-
tion through the mills; th<* secondary air system is for bed fluidisation
and coal combustion.
The airheater is of tubular type with flue gases flowing inside
the tubes.
The primary air, delivered by the FD fan, is heated in the hottest
upper part of the airheater and then is fed to the mills; the primary
air temperature is controlled by an attemperation system.
During plant start up, when the airheater is cold, the primary air
is heated to the right temperature by means of burners arranged
inside the ducts upstream of the mills.
The secondary air is delivered by the FD fan in the lower part of
the airheater and then to the windbox of each fluidized bed module.
At the inlet of each module a flowmeter and control valve are provided
to allow independent control fo secondary air flow to the modules.
3.5 Flue Gas Cleaning and Ash Discharge system
After the flue gased have left the boiler they enter a mechanical
dust collector, where the greater part of the fly ash is removed.
The partially cleaned gas then enters the air heater in which the
flue gas temperature is reduced to 170 C whilst preheating the
primary and secondary air.
Final gas clean up is effected in the bag. filters after which the
flue gas leaves the system via the ID fan and the stack.
Ash collected in the mechanical collector and the baghouse is routed
to a holding vessel prior to reinjection into che fluid bed, or
disposal.
Spent material is automatically withdrawn from the bed on a timed
sequence basis. This material is cooled in two water cooled screw
conveyors where the temperature is reduce for disposal.
3.6 Startup system
The fluidised bed boiler is started-up by means of four oil burners

located in the freedboard walls and inclined below the horizontal.
Burners have the capability of being modulated.
Each burner is a self contained package including fan, ignitor,
actuator and controls.
-218-
4. FB C BOILER LOAD CONTROL AND STARTUP
4.1 B oiler Load Control
A comprehensive control system has the function of allowing fully

automatic operation of the boiler over the full load range whilst
providing monitoring and alarm facilities to protect the process
and plant.
In the Porto Vesme boiler the fluid bed consists of a conventional
waterwall furnace enclosure with four identical fluid bed sections
or troughs mounted across the base of the combustor. Water filled
boiler tubes immersed in the fluid bed material allow a high level
of heat transfer from the bed to the water. The rate of steam genera
tion is therefore a function of the bed surface area in contact
with the tubes and of the temperature of the bed.
Further steam generation aiiQ superheating are obtained by convection
and radiation in the freeboard and convective sections of the ooiler.
At full load all four beds of the boiler will operate at a stable
temperature of around 850 C (fig. 2 ) , where the heat released
by burning coal with the fluidising air in the bed will balance
the heat being delivered to the bed, tubes and the boiler. Reduction
of steam output in the range approximately 75% to 100% vrn will
be obtained by reducing the bed temperature by adjustment of the fi
ring rate. Below 75% MCR the bed temperature using 4 beds will be too
low for correct working of the fluid bed and so load reduction will
be obtained by slumping one bed. This will have the effect of reducing
the heat transfer area and combustion region by 25% and hence will allow
an output of 75% MCR to be obtained at 850 C using 3 beds. Adjustment
of boiler output by means of bed temperature change will then be
possible in the range of approximately 55% to 80%. By slumping 2
beds load control range my be extended down to around 35% MCR and
using only 1 bed fluidised the boiler may be operated in the 25%
35% MCR range.
The task of the combustor load control system is to ensure that
at all times the boiler load demand is satisfied with the fluid
bed working in the correct temperature range and with the correct
number of beds in service. The system will provide fully automatic
control of the bed temperature in response to steam demand and
will automatically carry out fluidising and slumping sequences for
each bed as required by the steam load and bed temperature conditions.
The overall response rate of the boiler will depend on a number
of factors associated with the fluid bed controls. There are two
primary performance features to be considered :
(a) the rate of load change when changing bed temperatures with
a fixed number of beds
(b) the rate of load change when fluidising or slumping an additional

bed.
219
The response of the boiler output to bed temperature changes will
be optimised by use of a tuned temperature controller. This will
enable the bed temperature to be changed rapidly by using large
transient changes in the fuel feed rates.
Overfiring capacity will be built into the feed system for each
bed to improve the response rate. In general an increase rate
of 3% MCR per minute is expected.
To deal with the problem of maintaining a steady load change rate
when slumping or fluidising a bed a special supervisory control
system has been developed. This system is designed to allow a
change in the number of beds to be anticipated. The sequencing
actions required to add or remove a bed from service are then carried
out in a coordinated manner to ensure the minimum of disturbance
to the boiler output.
4.2 FBC boiler cold start-up
The FBC boiler start up from cold is based on the following philosophy
(fig. 3 ) .
Two beds will be heated simultaneously from cold with overbed oil
burners. In this way, the freeboard ^xit temperature can be main-
tained at a safe limit to protect the dry superheater tubing while
the circulating system and steam drum can be heated at a constant
rate producing a drum temperature rise of approximately 2.5 degrees
Centigrade per minute.
The drum pressure will be allowed to rise to 3.5 bar at which time
10% of the MCR steam flow will be passing to the superheater.
One bed will now be slumped while the other is established with
. coal firing.
10% steam flow to the superheater will be maintained while the drum,
pressure rises to design saturation pressure.
The remaining beds will be started according to turbine steam demand.
Each bed will be fluidised and mixed with the adjacent operating
bed(s), the oil burners located above the beds will be used to overfire
the fluidised beds to maintain safe ignition temperature and to
control the bed temperature rise back up to minimum bed operating
temperature.
For the Porto Vesme ' boiler the sequence for cold start up is as
follows :
1. two beds will be fluidised at a superficial velocity of 0.6

m/s
2. overbed oil burners will be ignited and heat output controlled

to produce a drum temperature rise of 2.5 degrees Centigrade
per minute and a freeboard exit temperature less than 480 C
3. when drum pressure reaches 3.5 bar, 8000 kg/hr of steam will
be passing to the superheater
-220-
4. slump one bed
5. increase burner firing rate on remaining fluidised bed to increase

bed temperature to safe coal ignition temperature
6. increase air flow rate to design MCR mass flow rate in order
to maintain stoichiometric conditions
7. burner firing rate will maintain bed material temperature at

ignition temperature
8. start coal feeders at 50% design speed and allow coal inventory
in the bed to increase to a level where the combustion reaction
rate is sufficient to increase the bed temperature
9. allow feeders to operate for 5 minutes and check for increases

in bed temperature. I f bed temperature has not increased then
shut down feeders and increase burner firing rate - wait 5 minutes
and re-start coal feeders
10. if bed temperature is increasing then put bed under automatic

control
11. control system will set feeder for minimum bed operating conditions
and ramp down oil burner firing rate to bring bed temperature
smoothly to equilibrium to avoid 'overshoot' on temperature
and the possibility of temperature oscillation
12. bed temperature will be controlled to maintain steam flow while

drum temperature/pressure is allowed to rise to design temperature/
pressure
5. COAL AND LI MESTONE PROPERTI ES
As with any process, plant design and performance is dependent upon

the characteristics of the feed materials, although this dependence
is less for FBC than for more conventional comoustion systems. The
processes of combustion and sulphur retention depend upon the properties
of the coal and limestone respectively, whilst the properties of
their mineral matter affect both fluidisation and attrition/elutriation.
During the preliminary design stage of this plant three coals and
two limestones have been considered as potential feed material.
A limited examination for their properties with regard to FBC was
therefore carried out.
5.1 Coals
Typical analysis of the coals are given in table I .

Sulcis no. 3 appears to be a high ash bituminous coal, whereas Sulcis
>>nd East Kentucky coal lay further towards the lignitous end of
the coal spectrum. Thus all three coals should be reasonably reactive
especially the Sulcis and East Kentucky coals.
221-
The initial deformation temperature of the Sulcis coals are significan-
tly higher than the maximum bed temperature of just over 900 C
at which it is proposed to operate this paint. Thus the ash should
be free of any. tendency to stickiness or sinter.
It has not been possible to examine the ash size distribution or
its resistance to attrition due to the difficulty in obtaining a
representative sample. However, the quality of the bed material
will largely be determined by the substantial quantity of limestone
that will be required for sulphur retention.
The relatively high CaO level in the Sulcis ash clarly contributes
to the significant proportion of sulphur retained by this ash, as
indicated by its high SO 3 content. In fact, about 28% of the sulphur
is retained, thus significantly reducing the limestone requirements.
The ash from Sulcis no. 3 coal appears to retain about 15% of the
sulphur.
5.2 Limestones
Limestones vary considerably in their capacity to asborb sulphur,

probably as a result of the individual quality of their porosities
following calcination, ftowever, as yet there appears to be no way
of predicting a limestor.a's reactivity from a chemical and physical
analysis alone.
For this project two limestones, local to the Sulcis mine, were
characterised in a test facility of NCB Coal Research Establishment;
the degree and speed of sulphation of the limestone were measured
and the results compared directly with standard limestones which
have previously been both characterised and used in FBC. Alternatively,
the results and test conditions could be used to evaluate the limestone
reactivity by application of a mathematical model developed for sulphur
retention.
The analysis of limestones is given in Table I. Under the test condi-
tions, the degree of sulphation of limestones A and B were similar at
23.9% and 23% respectively.
Both the A and B limestones appear to be fairly reactive and, with fines
recycle, a Ca : S ratio of no more than 2 should be required for 90% re-
duction in SO2 emission. In the case of Sulcis coal, where the coal
ash retains some sulphur, less limestone should be required.
Finally, the calcined limestones are subjects to a simple standard at-
trition test. The results of this test were used to compare the
samples' resistance to attrition with that of standard FBC materials.
Both limestones were found to be fairly resistant to attrition and hence
will not only maintain the bed, but will require a significant bed
drain.
-222-
TABLE I : COAL AND LIMESTONE ANALYSIS (X W/W)
COAL LIMESTONE
Sulcis Sulcis East A B

no. 3 Kentucky
Carbon 51.2 37.0 65.2 CaO 54.6 52.2

Hydrogen 3.8 2.5 4.6 MgO 0.1 2.2
Oxygen 9.8 6.6 9.6 co2 42.6 43.1
Nitrogen 1.3 0.7 1.0 Others 2.7 2.5
Sulphur 5.2 5.5 3.2
Moisture 7.9 12.0 7.9
Ash 20.7 35.7 8.5
G.C.V. (kcal/kg) 4903 3521 6650

Hardgrove index 48 51
Ash :
Si0 2 31.6 36.8

A1 2 0 3 13.6 10.8
Fe 2 0 3 12.5 27.6
CaO 15.8 22.4
S03 17.8 5.8
initial deform.
temp.(reducing 1150 1100
atm) C
-223-
6. PORTO VESHE FBC BOILER PERFORMANCE DATA
The main performance data of Porto Vesme FBC boiler are given in table II.
TABLE II : PORTO VESME FBC BOILER PERFORMANCE DATA
LOAD * 100 80 60
Superheater steam flow t/h 80 64 48

Superheater steam outlet temperature C 475 475 475
Superheater steam outlet pressure bar 59 59 59
Feedwater temperature C 190 187 185
Coal flow rate (Sulcis) t/h 11.4 9.2 6.9
Limestone flow t/h. 2.7 2.2 3
Number of beds in service 4 4 3
Fluidising air velocity m/s 2.42 1.84 1.86
Fluidising air flow t/h 78.2 59.6 44.8
Excess air % 20 20 20
Boiler efficiency on GCV % 87.5 87.0 87.1
Bed temperature C 875 805 812

Gas temperature : C
at freeboard exit C 851 785 745
at primary airheater inlet C 318 297 278
at secondary airheater outled C 172 163 155
Air temperature : C
primary air at airheater inlet C 38 38 38
primary air at airheater outled C 287 266 252
secondary air to FD fan inlet C 38 38 38
secondary air at airheater outlet C 235 227 218
Heat to immersed bed tubes MW 22.3 19.6 15.1
Heat to enclosure + radiation MW 12.3 10.0 7.9
Heat to SH & Economiser & airheater MW 22.3 16.1 11.4
Total heat MW 56.9 45.7 34.4
REFERENCE :
- Fluidised Combustion Contractors Ltd : technical papers

- NCB Coal Research Establishment : limestone characterisation report
1) mil n KONMCN. K T COUfCTOI
2) CON. u DWT C a i f C T I K NOTO
)) LIKSTOC KCYCU mm two
FUIIDIZO I Q nOUU 10) Ml AM* AIKATU
S) STEM* D M 11) SECOOMV MMEATBt
6) SUfWCATfB 12) IMHOUX FILTB
Figure 1 - FBC Boiler Schematic

20 40 60
FREEBOARD GAS EXIT TEWERATURE VS. LOAD
N NUMBER OF BEOS I N SERVI CE
BED TDFEM TURE VS. LOAD
N/S
3.0
FLUIDIZINS VELOCI TY VS. LOAD
Figure 2 - Turndown Curves of FBC Boiler
226-
Bed Temp
I i i
90 100 110 120 130
Time I M ins)
1 Bed Fluidised
Air Flow increased to

maintain Stoichiometry
i
Start Coal Feeders
Ramp Burners to avoid

Temperature Oversrtoot
Bed Temperature automatically
controlled to limit
Orum Temp rise
Figure 3 - FBC Boiler Typical Cold Start Details
227-
Figure 4 Atmospheric Fluidised B ed Combustion Boiler
General Arrangement
1 o 111 ^ \
11
s : 1
' i t
mi o III
w
\A
DISCUSSION
Question - Mr. Furfari, CEC :

Vorrei domandare al signor Bertoni se I'impianto e gia pre-
visto, se la costruzione e decisa e se ci pub dare in-
formazioni sul tasso di desolforazione del carbone?
Answer Attualmente il progetto e nella fase di progettazione di det-

taglio cioe sono in corso di esecuzione i disegni di dettaglio
che consentiranno I'approvvigionamento dei materiali. L'im-
pianto e previsto essere completato entro il 1984. Riteniamo
di iniziare ad ordinare i materiali o alia fine di quest'anno
o molto piu probabiImente all'inizio dell'anno prossimo.
Riguardo alia desolforazione nel letto posso dire che con un
rapporto molare calcio/zolfo di 2 con ricircolo dei fini e
prevista una ritenzione dello zolfo del 90% con carbone Sul-
cis lavato, in quanto le reattivita del limestone e delle
ceneri stesse del carbone sono tali che consentono una
addizione, un'aggiunta di calcare ridotta.
Question - Chairman :
Let us suppose that your demonstration is successful that
you get, I hope, rather a high conversion efficiency and that
the equipment is capable of load variation. Do you think that
certain installation at the sort of size that you are con-
structing would have an important future for many electricity
producers in the Community?
Answer ALlora vorrei ripetere quanto ho detto prima e cioe che non
ritengo che questo impianto da 15 HW possa trovare una grossa
diffusione nel contesto europeo perche le dimensioni non sono
quelle normalmente adottate negli impianti termo-elettrici o
almeno in quelli di grossa potenzialita dove attualmente i
gruppi partono da almeno 160 HW per arrivare a 660 MW come
singola unita. Lascio da parte gliautoproduttori per i quali
il discorso e diverse Per le grosse utilities ritengo che
questo progetto dimostrativo rappresenti un passo in avanti
per poter poi fare uno scale-up a dimensioni che una volta
modulate e quindi messi insieme diversi moduli possano arri-
vare alle taglie di cui vi ho pari ato prima.
Tuttavia, la dimostrazione una volta effettuata consentira
I'applicazione in usi di produzione di energia elettrica di
entita limitate, diciamo dai 15 ai 50 MW elettrici, un grosso
prodotto utilizzabile in campo industriale, almeno nel contes-
to delta Comunita Europea .
Certainly in the size of 15 to 50 megawatts there are a lot
of plants which today use oil products and I would hope that
even in those sizes such a plant if it straight forward in
operation it should find quite a number of applications?
-229-
Comment - Mr. Willis, N.C.B.
Mr. Chairman, You did ask about the market potential for Mr.
Bertoni larger project and I should have perhaps stressed
that this project is aimed for the industrial power generation
market and we see its application in the range of perhaps
15-20 megawatts up to probably 60 MW thermal.
You see this is mainly for indust rial production of heat and
power but in the Community we bel ieve that a lot still re-
mains to be done to encourage the use of district heating,
This is already successfully appl ied in certain countries of
the Community notably in Germany and Danemark. Now we hope
that several Community countries and several districts in
Community countries will start wi th relatively small systems,
relatively small networks for dis trict heating. We should al-
so be using coal wherever we can and there I would see a pos-
sibility for this size of unit an d that I think needs to be
more in mind.
Comment - Mr. Willis :

You touch a subject near to my hart because for two years I
managed the Nothingham Scheme which is the largest British
Scheme we have and this boiler would indeed be suitable for
that sort of application. I did not really make the point but
one of the weaknesses of the fluidised bed boiler is that to
achieve the operation of the boiler in a simple way there is
a significant blow down of water from the primary and second
separators. Of course much to see PHD schemes present no pro-
blem what so ever can be used for heaters and preheaters and
district heating to take off in a significant way would be a
most suitable boiler for that particular application.
Comment - Mr. Graf, Lurgi :

Just one point to your comment. There is a project, in Ger-
many, Duisburg, where an installation shall be built of up to
96 MW electric. There is a district heating and firing coal.
There are some projects, one of these is the 96 MW electric
and if you go down of course and you produce maybe 50% for
the district heating so this is going to be the contract and
it will be built within the next 2-3 years.
Comment - Chairman ':

Just for that information I might add just in this context
that the Commission has a series of studies being done for
us in each Member State about the situation on combined heat
and power and district heating. We b'elieve that on the basis
of these studies which have been done by experts in each in-
dividual Member State we shall get a very useful overview of
the situation and perhaps we shall be able to suggest some
guidelines for the future to extend this type of application.
230-
THE APPLICATION OF FLUIDIZED COMBUSTION OF COAL
TO A LOW COST. LARGE INDUSTRIAL WATER TUBE BOILER
INTEGRATED IN A COMBINED HEAT AND POWER SCHEiME
Authors : B.J. SMITH, D.M. WILLIS, J.M. TOPPER
Head of project : B.J. SMITH
Contractor : National Coal Board
Address : Hobart House

Grosvenor Place
UK-LONDON SW1X 7 AE
SUMMARY :
The National Coal Board in conjunction with ME Boilers Ltd. have been
responible for the design and construction of a fluidized bed coal
fired 30 MW coil boiler which is installaed at the British Steel
Corporation's River Don Works in Sheffield.
The design features of the boiler, and in particular of the fluidized

b"d and its controls are described. The unit was commissioned in mid
1981 and responsibility for operation passed to the NCB in November
1981. A resume is given of the early operational history. A number of
design improvements are in hand and solutions to a problem involving
bed coil rupture are identified.
It is concluded that design output has been met and that the demons-
tration will achieve its aims showing the scope for reducing the use
of industrial fuel oil by the development of the fluidized bed techno-
logy on this boiler.
-231-
1. INTRODUCTION
As part of the programme to demonstrate the commercial applicability

of its technology for fluidised bed combustion of industrially supplied
grades of coal, the National Coal Board has built a coil boiler of the
'once-through' type at the British Steel Corporation's River Don Works.
This boiler, supplied by M.E. Boilers Ltd., is unusual in that it is the
first 'once-through' boiler fired by a fluidised bed combustion unit with
coal as the fuel.
The output of the boiler is 10.1 kg/s of steam at 45.5 bar and 440C.
The system was originally designed for base load operation but following a
change in site use of steam, the design was revised to provide 6:1 turn-
down with rapid response to load changes. Thus, the steam output of the
boiler may be used for power generation via steam turbines or for process
uses, such as for steam hammers.
A schematic diagram of the installation is shown in Figure 1 and a

photograph in Figure 2. The fluidised bed is 5.03 m in diameter and is
divided into 6 segments which were designed for individual control to give
a 6:1 turndown. Improvements in the understanding of fluidised bed
performance elsewhere have shown that it will be practicable to obtain a
2:1 turndown on each bed and the unit is now to be modified so that it
operates with 3 pairs of beds rather than 6 individual ones.
About half of the heat release within the bed is removed by two
pancake coils. The remainder of the heat extracted is taken out of the
combustion gases in the convection sections, which include a superheater
section and an economiser.
Each of the 6 bed sections is equipped with a forced draught fan for
air supply, a coal feed rotary valve and drop tube and a supply system for
coke oven gas, which is used for start-up purposes as well as being a
second fuel.
Most of the ash produced is carried out of the boiler in the flue gas
and is captured by a multi-cell grit arrestor. Periodically, large
particles which have accumulated in each bed are removed by a bed cleaning
system particular to that bed.
The boiler is controlled by a programmable logic controller which will

permit automatic load following in the future.
The aims of the operating programme are to demonstrate the following:-
(a) That electric power can be generated from a combined heat and power
scheme using a fluidised bed fired coil boiler and with better thermal
efficiency than separate power and steam raising facilities would give.
232-
(b) That applying fluidised bed combustion of coal to large industrial
boilers is commercially viable.
(c) To provide data for the design of improved fluidised bed combustion
fired coil boilers.
(d) That a wide range of coals prepared to normal commercial standards

can be combusted efficiently.
(e) That programmable logic controllers can be successfully applied to

large fluidised bed boilers to give rapid response to load changes.
(f) Operability of the bed cleaning system.
(g) Environmental acceptability for U.K. conditions.
2. SPECIFICATION AND DESIGN PARAMETERS
The basis for the specification is the production requirement for

10.1 kg/s of steam at 45.5 bar and 440C from feed water at 82C. This
output is equivalent to a rating of 29.9 MW. Most of the main design
parameters are given in Table 1. A 1 MW test rig at the works of M.E.
Boilers Ltd. in Peterborough was used to aid and confirm the design
decisions taken with respect to combustion and heat transfer characteristics.
This test rig provided much useful data for the mathematical models of heat
and mass balances.
3. DESCRIPTION OF PLANT
3.1 Fluidised bed unit
Figure 3 shows the interior of the combustion chamber divided into 6

sections. Each of the refractory walls shown is 0.35 m thick and 2.13 m
tall. The annulus containing the fluidised bed is 5.03 m outer diameter
and 1.22 m inside diameter. The cross-section of the combustion chamber
starts to decrease towards the throat at a height 200 mm above the central
core.
The air distributor is shown in Figure 4. There are 225 standpipes in

each bed, each standpipe having 8 air escape slots. The standpipes are
rated to withstand 1000C, the base plate being insulated by a static layer
of sand. Each bed has an unbaffled plenum chamber for admitting the com-
bustion and fluidising air. Three ports are fitted from the base plate
through the plenum. One of these is for use with the bed cleaning system.
Bed material was originally specified as sand graded at 0.6 to 1.4 mm.
However, due to problems of availability at this grading the sand now used
is graded at 0.7 to 2.0 mm. The total weight of sand in the 6 beds is
approximately 14 tonnes.
The two in-bed pancake coils provide 60 m 2 of heat transfer surface

and were designed such that at Maximum Continuous Rating (10.1 kg/s
product steam) the inlet condition is 90% steam, 10% water by weight. The
static bed depth was designed at 400 mm above the standpipes as this is
just sufficient that, when at MCR, both coils are immersed and yet during
periods of slump neither coil is in the bed.
-233-
3.2 Ancillaries
The coal used is Allerton Bywater washed smalls. Its specification is

given in Table 1. Coal is transported from the British Steel bunker to the
boiler feed hoppers by a Macawber dense phase conveying system. There are
two boiler feed hoppers, each feeding 3 beds and each having a 5 tonne
capacity, which is equivalent to 2 hours full load operation. Level probes
within the hoppers activate the Macawber coal transfer system. Each bed is
fed via a rotary valve and drop tube arrangement.
Each forced draught air fan is rated to provide 11.35 tonnes/h of air.
A damper in the outlet duct of each fan allows turndown from the design
maximum output of 9.35 tonnes/h of air at 30% excess air. Further turn-
down to about one-third of maximum air requirement is possible. The single
induced draught fan is located between the grit arrestor and the stack. It
is used to maintain a balanced draught in the combustion chamber and to
remove flue gases to the stack. This fan is rated at 70.5 tonnes/h of gas
at 240C. A damper controls suction such that above bed pressure is main-
tained within 5 mm water gauge of the set point.
The start-up system employs coke-oven gas which is mixed with air in
the standpipes at the point of entry into the fluidised bed and it is
ignited by over bed burners. In view of the quantity of gas being fired in
each section the gas is supplied to the sections in steadily increasing
manner until full start-up flowrate has been achieved. Thus, the gas is
supplied first to the overbed igniter torches, then to the centre of the
bed at a low flowrate, then to the centre of the bed at full flow, and
finally to the periphery of the bed.
The grit arrestor is required to remove the majority of the coal ash
from the system though coarse grits from the first boiler pass are refired
directly with the coal. The grit arrestor installed is a multi-cell cyclone
with a dust collection hopper below. This is emptied by a rotary valve to
a drop tube, which takes the grit to sealed removable skips. The design
also incorporates facilities for refiring grit which drops out at the base
of the first boiler pass. This feature is shown in the schematic diagram
in Figure 1.
Accumulation of large particles in each bed can be counteracted by

removing bed material via one of the ducts running through the plenum into
a pneumatically operated vibrating screen. This uses a wedge wire con-
struction with a 3 mm gap. Oversized material is passed to a bin for
manual disposal whilst cleaned bed material is pneumatically returned into
the bed. This is one of the systems developed by the NCB for on load bed
cleaning.
3.3 Boiler
The boiler is an M.E. Boiler Ltd., 'once-through' water tube coil

adapted to fluidised bed firing by replacing most of the convective
evaporative section by a double pancake coil immersed in the bed. High
pressure preheated feed water passes through various sections to the bed
coil. The water/steam mixture from the bed is fed to primary and secondary
separators from which the water content is discharged. The steam is then
superheated in convective coils in the first pass.
-234-
The treated boiler feed water is drawn from the British Steel
Corporation's main boiler-house, monitored for conductivity and pumped for-
ward to the boiler by a feed pump having a discharge rating of 46.6 m3/h
at 90 bar. This pump has a limited turndown capability so that low rates
of supply are attained by recirculation of the excess water. The feed
water is heated indirectly by the boiler drain water in a high pressure
feed heater before entering the boiler coils. The water passes through the
economiser, intermediate, furnace and roof coils before being piped
externally to the evaporative coils in the fluidised bed. At this point the
water is expected to be 90% water and 10% steam by weight at full output
conditions. These percentages are reversed as the mixture enters the
primary separator.
The primary separator level control valve allows discharge to the

boiler drain system. Steam leaving the secondary separator should be
99.98% dry before it passes to the superheater inlet header, which is
located at the base of the boiler. Superheated steam passes to the
British Steel steam main at 440C and 45.5 bar.
The separator discharge system is designed to handle continuous

drainage at all boiler output levels and to treat this water prior to
disposal in the River Don. This water is fed to a vessel at 10.3 bar.
Residual water passes through more heat-exchangers before entering the
River at between 20 and 30C.
One of the advantages of the 'once-through' type of design is that

there is little possibility of a build up of dissolved solids in the water.
Thus, water treatment is not as critical as in more conventional designs.
4. PLANT CONTROL AND MONITORING
The control room is maintained under a slight positive pressure by use

of a fan fitted with an inlet filter. Most of the standard equipment is
housed in a panel supplied by Taylor Ltd., and this is linked to a program-
mable logic controller separately located, adjacent to the Taylor panel.
Temperatures are monitored throughout the bed and the boiler convective
sections. The bed temperature control thermocouple is one of a vertical
array of thermocouples in the centre of each bed. Thermocouples near the
edges of pach bed allow an assessment of temperature distribution. Metal
temperatures at the bed coil outer surfaces are now much more heavily
monitored than was initially intended (see Sections 5 and 6 ) . The super-
heater coil temperature is also monitored at two locations. Water/steam
temperature is monitored at a number of points through the coils.
Pressures are monitored through both the water/steam side and the
combustion side of the system. Pressure switches are used for a variety
of control functions and annunciator alarms. Pressure drop measurements
are used to indicate bed depth.
Calibrated orifice plates are used to monitor air and gas flow.
Coal flowrate is measured in terms of revolutions of the calibrated rotary
valves. Boiler feed water, water discharge to drain and cooling water
usage are also measured.
-235-
The programmable controller is supplied by Kratos and has a keyboard
and printer facility. The language employed is suitable for process control.
The unit has been programmed to cover safety procedures and many control
sequences while allowing a degree of manual control. Load following has
yet to be programmed into the unit.
Start-up is achieved by a series of manual actions which enable the

control sequences to operate in between them, providing that the necessary
pre-conditions are 'seen' by the controller.
The controller allows each bed section to be operated on gas or coal

provided that bed temperature is in excess of 700C. During operating on
coal, an air-to-fuel ratio controller maintains a 30% excess air condition.
A separate controller allows for manual trimming of coal feedrate to give
any bed temperature in the range 750-950C. At 950C coal feed is tem-
porarily interrupted and if bed temperature rises above 980 C a bed shut-
down is initiated. It is possible to slump beds as required, either to
shut down the unit or to simulate load following. It is also possible to
reactivate each bed after a period of being in the slumped state.
Because of the development nature of the unit there is considerably

more monitoring equipment than would normally be installed on a commercial
boiler. Apart from additional temperature and pressure measurements,
facilities are included for gas analysis, checking water conductivity at
various points through the coils, for dust and grit sampling and for
automatic data logging using an Apple computer. Analysis is carried out
via a link to the NCB mainframe computer at Cannock.
5. OPERATIONAL'HISTORY
Table 2 gives a summary of the operating times for the different beds.
As this indicates, preliminary trials began in May 1981. The equipment
was regarded as being sufficiently proven for the NCB to take over (from
M.E. Boilers) in November 1981. Since November 1981 a number of tests
have been carried out, some design modifications identified and 2 tube
failures have been experienced. These are dealt with below.
5.1 Test data
A number of performance trials have been undertaken. The summary of

one of them for full load operation is given as Table 3. This shows that
Maximum Rating can be attained but in this case at a lower efficiency than
had been anticipated. The discrepancy is primarily due to greater unburnt
carbon loss than had been expected. It has been shown that this situation
can be alleviated for lower load requirements by operating successfully
with a shallower bed and using slightly coarser sand.
5.2 Design modifications
Recurring difficulties have been experienced with coal feed blockages

which start from a build-up of fines on the end of the drop tubes.
Different drop tube materials have been tried with limited success, so
that it is proposed to try out a Sturtevant incremental feeder instead.
This device accurately injects small 'slugs' of coal. It is essential for
the success of this equipment that the 'slugs' be as large as possible to
simplify the measuring of it whilst at the same time not being too large to
-236-
upset the balance of combustion (causing smoke emission). Some development
work is therefore necessary.
Tests on the grit arrestor have shown the performance to be below

specification, possibly due to the gas flow pattern to the unit. Some
modifications have been made though it is possible that the unit may still
not meet requirements and further modifications may be necessary. The
system of removal of grits from the unit is also under review as the
present system is not acceptable in the long term.
The refractory walls acting as dividers between the beds have been
troublesome in certain respects. It is intended to change the design of
the spectacle plate (metal strips on either side of the bed coils as they
pass through the walls) and to reduce the number of wall sections above
the coils from 6 to 3 (see Section 1 for reasons for this) for when the
boiler is next fired.
A complete sand handling system is to be installed. It is intended

that sand will be delivered to site by tanker and transferred pneumatically
from the tanker into a reception hopper. Pneumatic transfer will also be
used to transfer sand from the main store to sand bunkers local to the
boiler. It is intended to automate the whole system so that a low
differential pressure in the fluidised bed will call for sand addition
from the local bunker, which will in turn be 'topped up 1 from the main
silo. This system will eliminate the need for manhandling of 50 kg bags.
5.3 Tube failures
From the outset it was realised that there might be difficulties at

the transition point from water to steam in the oncethrough design.
The original design envisaged that the mixture entering the bed coil,
a zone of high heat transfer, should be 10% by weight. Literature suggests
that the 'critical' mix for susceptibility to steam/water separation is
around 3% steam by weight though there was no information on steam/water
separation in flat bed coils with 100 mm diameter tube.
From a heat transfer point of view, the bed coil was recognised to be
a zone of high heat flux and the overall heat flux rates were well under
stood and confirmed by the 1 Mw test rig. The degree of variation around
the circumference of the coil was not however known so precisely and there
is much less recorded work on showing the effects of varying tube sizes,
different fluidising velocities and different particle sizes.
The heat transfer to the coil section above the bed was also less
precisely known as this depended upon the overbed temperature arising from
the combustion of the smalls coal whilst dropping on to the bed.
The first tube failure occurred at the end of November 1981 and
metallurgical analysis indicated that the crown of the coil had overheated
in the region of the failure. The defective length of tube was replaced,
more thermocouples were attached to the tube wall and the boiler was
recommissioned in February 1982.
Careful observation showed that the bed coil wall temperatures were
very sensitive to rate of flow and increase in the rate of flow by
237
increasing the water flow to drain at loads less than rating did reduce
metal wall temperature. However, a further more severe rupture occurred
on 29th April 1982, (see Figure 5) and the boiler has not been run since
whilst an in-depth analysis of the failures has been carried out.
A careful and systematic examination of the coil showed a number of

areas of metal wall thinning and metallurgical analysis confirmed that the
crown of the coil had overheated at these points due to inadequate heat
transfer to the steam/water mixture in the tube.
A number of experts, including research personnel of the Central

Electricity Generating Board, were consulted and the data obtained from
the plant was examined in some detail. It was concluded that at
conditions corresponding to water flows of less than full output the
quality of the steam/water mix at the entry to the coil is such as to give
steam/water separation and the formation of near stagnant bubbles of steam
at the crown of the coil. Increasing the water flow (as had been done
after the first failure) only moved the likely point of failure.
Bearing in mind that no recorded data exists for plant of this type
it was decided to carry out a major redesign and at the same time carry out
some model work using air/water mixtures in a clear tube.
6. REDESIGN
Because of the delays in obtaining new equipment the redesign of the

bed coil arrangement is in two phases. The first phase involved a change
in the flow arrangements so that the inlet coil is the top coil. During
start-up and low load conditions the top coil will be above the bed and in
a zone of lower heat flux. The bed will be temporarily operated with a
lower level whilst the metal temperatures are carefully monitored. A
slightly modified start-up is also to be used with the bed coils being
brought up to a steaming condition at a lower temperature than was used
previously. As a further aid in preventing steam/water separation,
turbulators are being fitted inside the tube in the first 6 laps. Again
no recorded experience is available on the use of turbulators in such large
tubes and decisions have been made on length of turbulator and their number
on scaled up information from other work and some limited modelling work.
Safety requirements on the plant have been increased to ensure that no
personnel can be harmed by further tube failure. This work will not be
completed until December and the boiler will not be available for running
until January.
The second phase will involve the introduction of a considerable

recirculation into the bed coil and increasing the ability of the
separators to deal with a much higher water flow. The recirculation pump
is on long delivery and will not be available until February or March.
If the model work and the preliminary runs with the modifications carried
out in the first phase indicate that separation is still likely in 100 mm
diameter tubes then a parallel pass of smaller diameter tubes will be
incorporated in the bed and indeed the design of such a system is now in
hand.
7. MODELLING
The cold model referred to previously consists of lengths of 100 mm
-238-
internal diameter clear tubes through which water can be pumped at varying
velocities and into which air can be injected.
This model will be used as a guide to indicate the velocities and

mixtures at which separation occurs as well as simulating the effect of
turbulation. The results will be indicative rather than providing any
absolute measure, but it is felt that they will be useful.
The results will be reported in the future.
8. C NCLUSIONS
(1) Early operations with the 30 MW boiler have shown that performance to
give the design output is possible. Thus, such units have a consider-
able potential for reducing dependence on oil firing for large
industrial boilers, using a design which is low in capital cost when
compared with conventional high pressure boilers with steam drums.
(2) The indications so far are that fluidised combustion technology can
be successfully incorporated into an economic combined heat and power
scheme.
(3) A number of relatively straightforward design improvements have been

identified. The only major problem so far has been the occurrence of
2 phase flow under certain conditions. However, short and long term
solutions for this have been identified which should result in pro-
longed satisfactory operation of the unit.
239-
IAELS 1 PARAMETERS USED IN THE DESIGN
Basis for calculations:

Coal type Allerton Bywater 19 mm washed snails
Coal analysis Moisture 11.556
Ash 7.4%
Volatile matter 31.3%
Fixed carbon 49.8%
Sulphur 1.71% (N.B. The chimney has been designed
assuming a 2.0% sulphur content).
Chlorine 0.34%
Gross calorific value, as fired 27,700 kJ/kg

Stoichiometric air requirements, (coal as fired) 8.94 kg air per kg coal
Boiler combustion efficiency 95%**
Combustibles loss (including CO loss of 1%) 5%**
Proportion of heat release in the fluidised bed 85%**
Proportion of heat release above the fluidised bed 10%**
Excess air level 30%*
Fluidised bed normal operating temperature 950C*
Total active area of fluidised bed (not including that
obscured by the dividing walls, taken as 225 mm thick) 16.1 m2*
Bed material 12/25 grade sand 600 um - 1400 um*
Steam/water temperature in the in-bed tubes.
inlet 280C
outlet 273C
Steam/water pressure in the in-bed tubes,
inlet 915 psig (64.1 bar)
outlet 820 psig (57.6 bar)
Turndown ratio 6:1
Flue gas temperature 193C
Calculated parameters
Fluidising velocity (@ 950C) 3.2 ra/s
Total combustion air requirements (@ 30% excess air) 56,200 kg/h
Coal feedrate 4835 kg/h
Combustion intensity (100% coal feedrate) 300 kg/m"2 h"1
Heat transferred to steam 29,900 kW
Boiler efficiency, by difference on gross C.V. 80.4%
Active area of inbed tubes 60 m 2
Total area of in-bed tubes (assuming 8.4 n 2
masked by walls, M.E. figure) 68.4 m 2
Heat to steam via in-bed tubes 13200 ktf
Proportion of total heat to steam via in-bed tubes 44%
Bed depth 380 - 400 mm
* a selected parameter
** estimated performance
-240-
TA3LE 2 a) OPERATING HISTORY (MAY 1981-OCT 1981)
Beds Operating Firing Times

Date Coal Sand Grit
1981 1 Fired Added Removed
1 2 3 4 6 Gas Coal
5
(mins) (mins) kg kg kg
Period 2 11:31 92:29 59200
May 13 1 124 5:56 2500 400
14 1 :39 2:07 1000 10003
20 / / 1:23 6:46 3600
22 / / / 1:52 11:30 9200 10003
Jun 1 1 :41 :25 200
2 / / :39 1:00 600
3 / / / / / 2:45 17:48 10700
4 / / / :44 3:40 2800 1100
9 / / / :53 17:35 12200
10 / 1 / 1:16 19:42 11600
11 / :15 6:00 4800 3806
Period 3 22:08 146:02 105580

Jul 14 / :30 1:49 1100
16 / / 1 / / / 5:46 14:51 12100
23 / / 1 / / / 7:29 41:10 32820 3720
Aug 3 6600
20 1 :36 5:22 4130 2020
21 1 :31 5:25 2570 750 1602
25 / / 1 / / / 4:03 54:14 43490 250
27 / 1 / / 3:13 23:11 9370 12040
Period 4 22:20 16510

Sep 11 / / 1:50 1460
15 / / 2:40 800
16 / 0:20 260
17 / / 1:50 1460
24 1 / 9:40 7730
25 / 1 6:00 4800
2 & 3 33:39 238:31 164780 2500 29738
-241-
TABLE 2(b) OPERATING HISTORY (NOV 1981-APR 1982)

Date Coal Sand Grit
1981 Fired Added Removed
1 2 3 4 5 6 Gas Coal
(hrs) (hrs) Tonnes kg kg
Period 5 666 450.0 11950 61340
Nov 2 / / / 46 25.6 1200 5520
3 / / /
3 / /
4 / 1 / / / 255 163.1 500 20800
5 / 1 / / /
6 / _/_./ / /
9 / 35 13.1 1600 2640
10 / / / /
23 / /
24 / / / / / / 232 161.1 7850 22240
25 / / / / / /
26 / / / /
Dec 2 / / / / / / 98 82.8 800 10140
3 / / / / / /
1982
Period 6 23 20 10.0
Feb 10 / 2
11 / 1
12 / 2
17 / / 2
18 / / 7 1
23 / 1 4
24 / 1 2 6
25 / 1 / 3 13
Period 7 334 258.0 6300
Mar 1 / 1 / /
2 / / 1 / / / 224 3400
3 / / 1 / / /
4 / / 1 / / /
9 / / /
10 / / 1 / / / 110 2900
11 / / / / /
-242-
TABLE 2(b) OPERATING HISTORY (NOV 1981-APR 1982) Cont.

Date Coal Sand Grit
1982 Fired Added Renoved
1 2 3 4 5 6 Gas Coal
(hrs) (hrs) Tonnes kg kg
Period 8 210 162.0 4750
Mar 15 / / / 9 800
17 / / / / 29 200
18 / / / 29 1800
24 / / 13 500
25 / / 20 500
30 / / / 29 500
31 / / / 26
Apr 1 / / / 25
27 / / 5 300
28 / / / 12
29 / / 13 200
Totals 1230 880.0
243
TABLE 3 DATA FROM ANALYSIS PROGRAMME FULL LOAD
RIVER DON ANALYSIS
Test Date 25.11.1981

Start 08.28 Hrs
Finish 09.11 Hrs
Duration .75 Hrs
Number of beds in operation
Coal Analysis:
Carbon 68 %
Hydrogen 4.3 %
Oxygen 7.2 %
Sulphur 1.78 %
Ash 9.2 %
Moisture 11.3 %
Calorific Value 27150 KJ/KG
Overhead Burner Off
Coal Flow Rate 5141 KG/Hr

Combustion Intensity 320 KG/Hr M2
Air Flow Rate, Bed 3 8820 KG/Hr
Fluidising Velocity, Bed 3 2.85 M/S
Excess Air in Bed 3 (Stoich. Method) 7.94 %
Combustor by Og Method 95.95 %
Average Temperature, Bed 3 921 Deg. C
Boiler Feedwater Rate 35847 KG/Hr

Steam Flow Rate 30332 KG/Hr
Drain Off Rate 4359 KG/Hr
Heat Input 38.77 MW

Heat Output to Steam 27.77 MW
Heat Output to Drain Off 1.38 MW
Heat Loss 3.57 %
Heat to Unburnt Solids (Total) 2.68 MW
Heat Loss 6.9 %
Boiler Efficiency 72 Of
Including Drain Off 75 %
-244-
INSTRUMENT * f C
CONTROL ROOM
ifll!H&tAft PAWti--
FIGURE 1 Schematic Diagram of the boiler installation

FIGURE 2 General view of boiler installation FIGURE 3 The combustion chamber
H?t3
$Vr-;'-
Wt
FIGURE 4 Single segment air distributor plate
FIGURE 5 Pipefracture
247-
USE OF FLUIDISED BED FURNACES IN STEEL HEAT TREATMENT
Authors : U.R. LENEL, M.A. MOORE
Head of project M.A. MOORE
Contractor : Fulmer Research Laboratories Ltd.
Address : Stoke Poges

Slough SL2 4QD
United Kingdom
SUMMARY
The energy consumption of a gas-fired fluidised bed heat treatment

furnace has been measured and compared with the measured energy
consumption of a gas-fired salt bath furnace and an electric-fixed
seal quench furnace. The measurements have been made on furnaces
employed in heat treatment of a variety of steel components at the
site of a commercial heat treatment company. Energy consumption of
the three furnaces has been metred and continuously recorded by a
data-logger over a period of several months. The results show
that the efficiencies of energy use by the salt bath and the
fluidised bed furnace are below 5 % while the efficiency of the
seal quench furnace approaches 20 %. The major factor affecting
furnace efficiency is material throughput, and efficiency is markedly
improved if furnaces are operated continuously with a high material
throughput. The performance of the fluidised bed equipment is not up
to specification and improvements in reliability through modification
of the equipment and operating procedures will be required before
any energy savings can be demonstrated.
-248-
1. INTRODUCTION
The objective of this project is to demonstrate the energy savings

obtainable by the use of fluidised bed furnaces in the heat treatment
of steels.
The production of steel components in many cases requires heat treat-

ment, which often includes heating to a temperature in the range 800 C
to 10OO C, holding at the specified temperature for times ranging from
a few minutes to several hours, followed by cooling. Several furnaces
suitable for such heat treatment cycles are available.
Fluidised bed heat treatment furnaces provide a direct replacement for

salt baths employing both cyanide and neutral salts for hardening of
steel components. They are also suitable for carburising, carbonitriding,
tempering and annealing/normalising treatments that are commonly performed
in seal-quench, muffle and pit furnaces. In the course of this project,
energy consumptions of a fluidised bed furnace, a salt bath and a seal
quench furnace have been monitored and compared.
The monitoring has been carried out at the site of Northern Beat Treatments
Ltd., Blackburn, which is a commercial heat treatment business specialising
in the heat treatment of a variety of steel components, e.g. for the automotive
industry. The energy consumption of furnaces is a matter of concern for
such a business, but the quality of the heat treated product is also
important and this must be taken into account when comparing furnaces.
FURNACES MONITORED
2.1. Salt Bath
A 25 inch diameter gas fired salt bath, operating with neutral salt
in the temperature range 800 C to 10O0 C or with cyanide salt at 850C,
has been monitored. The salt bath, shown in Figure 1, provides rapid
temperature control and high heat transfer rates but lacks capacity and
is messy and toxic. It is therefore used only for relatively small
batches of components for specific heat treatments, and at Northern Heat
Treatments is not operated continuously. The furnace is used on a daily
basis as necessary.
2.2. Seal Quench Furnace
A lOOkVA electric fired seal quench furnace, shown in Figure 2, has been
monitored. The furnace can heat treat several hundred kilograms of
steel in one batch. Heat transfer in the furnace is relatively poor
and the sealing of a protective (or carburising etc.) gas in the furnace
leads to problems during firing the furnace from cold, which takes
several days. The furnace is therefore run continuously, and every
effort is made to ensure continuous throughput of material.
-249-
.3 Fluidised Bed Furnace
An Apollo Beat Ltd. G76m Fireball Mark III Fluidised Bed Furnace, shown
In Figure 3, was Installed at the premises of Northern Heat Treatments Ltd
In August' 1981. The furnace Is natural gas fired and can operate with
a protective atmosphere gas. Due to difficulties in supply, the
recuperative hood, which recovers waste heat from the flue gases, has
not been installed, and other components have not operated satisfactorily.
These problems have caused considerable delay in commissioning of the
plant and even now the plant does not operate with 100% reliability,
particularly in terms of heating rate and product quality.
The fluidised bed is a versatile furnace which offers advantages of

relatively high capacity, rapid firing from cold, rapid temperature
control, ability to accept a range of atmospheres and good heat transfer.
It should therefore be possible to operate it in modes either similar to
salt baths (small batches, specific heat treatments) or similar to seal
quench furnaces (bulk heat treatment, semi-continuous loading). How-
ever, due to problems with reliability mentioned above and with sand
deposits (alumina) around the furnace, Northern Heat Treatments Ltd have
been reluctant to operate the furnace continuously on a commercial heat
treatment basis. The furnace has therefore been operated mainly on a
daily basis, heat treating a small number of batches of material during'
a day.
MONITORING EQUIPMENT
Energy consumption has been metered by commercially available gas and

electricity meters and has been continuously monitored by a data
logger. Records of heat treatments carried out are made by furnace
operators.
3.1. Meters
The gas consumption of the salt bath has been metered by a 1.5M
Dresser meter (visible in the back of Figure 1) and that of the
fluidised bed furnace by a slightly larger 3.0M Dresser meter. The
meters send pulses equivalent to a gas consumption of 10 cubic feet
to the data logger.
The electricity consumption of the seal quench furnace has been

metred by a Yew Clip-on AC Power Meter, shown in Figure 4. The
meter is connected into the three-phase power supply to the furnace.
The jaws of the meter surround one phase of the supply to measure
current while connections are made to all three phases to measure
voltage. The resulting power measured is displayed on the meter
and the energy consumption is relayed to the data logger by pulses
equivalent to a consumption of 0.604 kwhr.
-250-
3.2. Data Logger
The data logger used in this work, known as "Matilda", has been
developed by the Midlands Research Station (MRS) of British Gas.
The data logger, shown in Figure 5, is based on a TI 58C programmable
calculator with up to five inputs. The data logger has been pro-
grammed to accept and store data on three channels simultaneously
(salt bath, fluidised bed and seal quench furnace) and to compute and
print out times at which specifid quantities of energy are consumed
(lOO cu ft gas, 6.04 kWhr electricity) on each of the three channels.
The data recorder also prints out summaries of energy consumption
twice daily.
In the course of the project, modified programmes have been used to

allow a more detailed record of heat treatments to be made. The
data logger has been operated unattended for several months.
Periodic problems,such as power cuts and jamming of paper in the
printer, have caused a considerable amount of data to be lost.
3.3. Data Recorded by Furnace Operator
For each batch of heat treatment, the furnace operator records the
date and time (in and out), the type of material, the weight of
material and the temperaure(s) of heat treatment. The time records
are used to correlate the heat treatment batch with energy consumption
recorded by the data logger, and the weight of material and the temperature
are used to calculate the theoretical energy required for the heat treatment.
From these records, the efficiency of the furnaces can be calculated.
4. RESULTS
Energy consumption data and calculated efficiencies have been obtained

from the salt bath since the start of this project (early 1981), from
the fluidised bed since commissioning of the plant (November 1981)
and from the seal quench furnace since the electricity meter was
installed (July 1982). Some results for the three furnaces are shown
in Tables I, II and III.
The results shown in Table I refer to the salt bath operating either
with neutral salt or with cyanide salt (temperature 850 C ) . Apart
from the first line of the table (20-22 May 1981) the results are
for operating runs of one day or less, the bath being switched on
early in the morning and switched off after completion of heat
treatment. Daily efficiencies (theoretical divided by actual
energy consumption) are given in the final column of Table I and the
overall mean efficier.y is 2.6%.
-251 -
Table II shows the energy consumption during a week of heat treatment
in the seal quench furnace. This furnace operates continuously so
it is not possible to compute daily efficiencies. Also the energy
consumption figures cannot be directly related to the heat treatments
carried out> each day, since treatments frequently continue from one
day to the next. The mass of material given in column three is in
some cases an underestimate, since the furnace operator did not always
record the mass of the smaller batches. The overall efficiency is
computed to be 15.6%.
The energy consumption during several days of operation of the

fluidised bed furnace is shown in Table III. As with the salt bath,
the furnace operated on a daily basis, being switched on early in the
day and switched off after the last heat treatment. Daily computed
efficiencies are shown in the final column of Table III and the over-
all mean efficiency is 3.2%.
A typical pattern of gas consumption during one day of operation of

the salt bath is shown in Figure 6. The time taken to consume a
given quantity of gas (50 cu.ft.) is plotted against the gas con-
sumption (cumulative total for the day). Thus a low bar in Figure 6
indicates a high rate of gas consumption. The times at which batches
of material were put into and taken out of the furnace are arrowed.
The behaviour is as expected, the rate of gas consumption increasing
when material is loaded into the furnace. It is interesting to note
however that the rate of gas consumption remains fairly high even
when there is no load in the furnace or when material is being held
at a constant temperature in the furnace. This indicates that a
large proportion of the energy consumed is used to keep the furnace
at temperture and that much energy is wasted. This is borne out by
the low efficiency figure of 2.6%.
5. DISCUSSION
It is generally accepted that the energy efficiency of conventional

salt baths, electric seal quench furnaces and other radiant furnaces
in hardening steels is of the order of 20%, while manufacturers of
fluidised bed furnaces estimate that efficiencies of up to 54% can
be achieved. The results of this work show that the salt bath and
fluidised bed furnace are remarkable inefficient (around 3 % ) , while
the efficiency of the seal quench furnace (16%) is closer to the
value expected. The results indicate that the treatment cycle
(continuous or stop/start) has an overriding effect on furnace
efficiency.
The inefficiency of the salt bath can be attributed to the poor

insulation of the furnace and the stop-start mode in which it is
operated. The salt bath at Northern Heat Treatments Ltd. is
several years old and during operation of the furnace heat losses
through the sides and top of the furnace can be clearly discerned.
The small size of the furnace limits the size of the heat treatment
batches, and the daily throughput of material is of the order of
100kg. The low throughput rate coupled with daily shutting down of
the furnace and high heat losses lead to very low energy efficiencies,
generally below 3%.
252-
In contrast, the capacity and throughput rate of the seal quench
furnace are much larger (dally throughputs of 1000kg are not uncommon)
and the average efficiency is correspondingly higher at about 16%.
The seal quench furnaces at Northern Heat Treatments Ltd are much
better insulated than the salt bath, and heat losses are lower. In
addition, the operating practice at Northern Heat Treatments Ltd of
running the seal quench furnace continuously and attempting to keep it
loaded as much of the time as possible allow energy to be consumed in
the most efficient manner possible. Since the computed efficiency
figure does not take account of small unweighed parts of batches, the
efficiency value of 16% is anyway an underestimate and the true
efficiency is likely to be closer to the expected figure of 20%.
The low efficiency of the fluidised bed furnace is somewhat surprising

in view of the ideal efficiencies of over 50% which have been claimed
by manufacturers. However, the fluidised bed at Northern Heat
Treatments Ltd is operating in a far from ideal manner.
First, due to supply difficulties as mentioned above, the furnace is

operating without a recuperative hood which would recover heat from
flue gases and provide heat to a tempering furnace, which in this
case must be fuelled by gas. The gas consumption figures recorded
are therefore higher than they would be if a recuperative hood were
used.
Second, the furnace has not been operated continuously (as it was
designed to do) because of a number of operating difficulties. It
has proved difficult to achieve a reliable fast heating rate, although
this is likely to be a matter of experience and the situation will
probably improve with time or as minor modifications are made. It
has also been necessary to carefully design baskets to hold the
components in the furnace to ensure even heating throughout. This
is a particular problem with large components, which also suffer
pick-up of sand when the components are taken out of the furnace.
This can lead to quenching difficulties.
In a commercial heat treatment business, it is vitally important to

ensure the quality of the product and thus the reluctance of the
operators to commit any but the simplest tried and tested batches of
material to the fluidised bed is quite understandable. Daily material
throughput is currently of the order of 200kg, but the capacity of the
furnace would allow a much higher throughput. If reliable quality
were achieved and the furnace were operated continuously, it seems
likely that the efficiency of the furnace would be markedly improved.
-253-
6. CONCLUSIONS
1. This energy demonstration project has not been able to demonstrate

that energy savings can be achieved by use of fluidised bed
furnaces in steel heat treatment.
2. Measured energy efficiencies of a salt bath, a fluidised bed

furnace and a seal quench furnace are 2.6%, 3.2% and 15.6%
respectively. The large difference in efficiency between the
first two and the latter can be attributed to the difference in
operation mode, indicating that the major factors affecting
furnace efficiency are material throughput and continuity of
operation.
3. The fluidised bed shows a small but significant improvement in

energy efficiency over the salt bath operating in similar heat
treatment cycles, showing that energy losses in the fluidised
bed are lower. Further improvements would be possible with
the installation of a recuperative hood on the fluidised bed
furnace.
4. Large improvements in overall efficiency of the fluidised bed

furnace would be possible if it were operated continuously with
a higher daily material throughput. To operate the fluidised
bed in this manner on a commercial basis, improvements in
reliability of operation and product quality are required.
5. Problems with reliability of the fluidised bed arise from:
* difficulty in obtaining even heating throughout the

components in a batch
* difficulty in obtaining reliable high heating rates
* pick-up of sand by flat surfaces of components and

transfer of sand to quench baths.
It is likely that these problems could be solved by modifications

to the furnace, further experience withoperating the furnace and
development of improved baskets to hold the components in the
furnace and transfer them to quench baths.
-254-
Table I. Energy Consumption of Salt B aths.
Date Mass Temperature Gas Energy Theoretical Efficiency

(c> Consumption Consumption energy consumption
(kg)
(cu.ft) <kJ/kq) (kJAg)
%
2022 May 1981 304 823 9650 34200 501 1.5
27 May 1981 249 850 5530 23800 527 2.2
29 May 1981 15 BOO 850 60800 489 0.8
39 10OO 1660 4570O 640 1.4
20 June 1981 248 850 2490 1O8O0 527 4.9
26 Nov 1981 105 800 2160 221O0 489 2.2
Weighted Mean Efficiency 2.6

Total Mass Heat Treated 960 kg
No. of days 7
Table II. Energy Consumption of the Electric Seal Quench Furnace

Energy Consumed Weight of Theoretical
Temperature
Data 5am 5am Material Energy Consumption
(kWhr) (kg)
(C) (MJ)
2 Sept 1982 1212 1488 850 783.2
3 Sept 1982 865 312 825 158.3
186 9O0 104.9
339 8S0 178.4
4 Sept 1982 1041 523 850 275.3
832 9O0 469.3
5 Sept 1982 991 150 850 78.9
236 925 137.6
6 Sept 1982 1010 346 850 182.1
261 925 152.2
7 Sepf 1982 410 551 850 290.0
(to 5pm) 519 900 292.8
Total 5529 5743 3103

(19904MJ)
Efficiency 3103/19904 0.156 15.6*

Total Mass B eat Treated 5743Kg
No. of days 7
Table III. Energy Consumption of Fluidised B ed Furnace

Data Mass Temperature Gas Energy Theoretical Efficiency
(kg) (C> Consumption Consumption energy consumption
%
(cu.ft) (kJ/kg) (kJ/kg)
7 Feb 1982 128 9 SO

168 1000 5190 18800 624 3.3
8 Feb 1982 111 950 3970 38300 602 1.6
17 Fab 1982 122 950 4160 36600 602 1.6
4 Mar 1982 105 1 950 2780 28400 602 2.1
7 Mar 1982 89 950 3060 3690O 602 1.6
12 Mar 1982 261 950 3560 146O0 602 4.1
13 Mar 1982 261 950 3590 147O0 602 4.1
14 Mar 1982 302 . 9SO 4420 15700 602 3.8
15 Mar 1982 302 ' 950 4560 16200 602 3.7
16 Mar 1982 179 ' 950 4430 265O0 602 2.3
Weighted mean Efficiency 3.2\

Total Mass Heat Treated 2028 kg
No. of days 10
255
Fig. 1 The Salt Bath. The gas meter can be seen at the rear on the right
Fig. 2 The Seal Quench Furnace
-256-
LlH
I r)
*
Fig. 3a Apollo Heat Ltd., G76m Fireball Mark 111 Fluidised

Bed Furnace
Fig. 3b The Fluidised Bed (left) at Northern Heat Treatments Ltd.,

showing control equipment and baskets for loading components
-257-
Fig. 4 Yew Clip-on AC Power Meter
Fig. 5 "Matilda" Data

Recorder
-258-
30
O
o
in
800"C 1000 C
E
a 20
In Out In Out In Out In Out
I II I 11
E 10
0 200 400 800 800 1000 1200 1400 1600 1800 2000 2200 2400
Gat volume (cu. ft.)
Fig. 6 Neutral salt pot 29.5.81
DISCUSSION
I quite often say to people who are engaged in demonstration
projects that it is as important often to ident ify things
that don't work and why as to identify why they are successes,
Hiss Lenel has told us of what I hope is an intermediate stage
where the results are disappointing for fluidiz ed bed applica-
tion. I think it sounds to me as it is a repeti tion of the
problem what sig . GriI Li told us a little earl ier. There is
a problem of adapting the know-how which exists in fluidised
bed technology to these metallurgical applicati ons. I would
like to ask Miss Lenel if the people who are responsible for
designing and now presumably correcting fluidiz ed bed who are
presumaly not you. Do you think that they have got the measure
of the problem. Do they know how to tackle it?
Answer The people who supply the equipment have unfortunately now
gone out of business. The operators of the furnaces are trying
very hard to correct the things which are wrong with the fur-
naces with the help of some other manufacturers in the UK and
improvements certainly seem likely. You ask whether they have
the measure of the problem. I think they have a very good
-259-
idea of what is wrong, it is mostly due to fluidisation and
bad loading of components. Whether they are actually able to
modify this furnace to operate as it should do we don't really
know.There are some things which might need changing for in-
stance the burning system, but all this is very valuable ex-
perience in learning how to design future furnaces for steel
heat treatment.
Comment - Mr. Bannard, National Institute for Higher Education Limerick:

I unfortunately am not associated with the furnaces side of
our business which is regarded as a slightly separate art but
I feel I ought to say that we do have furnaces, we are able to
do tests and demonstrations on specific components which the
customer wants to heat treat and we do help methodologists
who are able to discuss these subjects and develop the pro-
cesses .
The boundary between a. research and development project and a
demonstration project is sometimes difficult to establish.
Particularly when one takes into account that sometimes the
endpoint of research and development is what you might call a
pilot of prototype which is a sizeble example which is usual-
ly a quite important scale but it realLy is supposed to do
technological proving and not necessarily an economic proving
because the demonstration is really supposed to be to show
whether the thing is economic. In this last case the project
which has been supported as a demonstration project for what
I have heard sofar sounds more like the endpoint of research
and development than the demonstration but of course I have
been told that this is not the endpoint it is intended that as
a result of hopefully a favourable result on energy saving one
would proceed to a larger demonstration of fluidized bed. Is
that right Hiss Lenel?
Answer If we can demonstrate that there are energy savings obtainable

then I think we would publicize these as widely as we can
throughout the Community and we would hope that it would help
fluidised beds to be used on a much larger scale.
Question - Chairman : .
Do you remain convinced that for the type of heat treatment
which the Company is doing theoretically you will expect a
higher efficiency with the fluidised bed than with the other
processes?
Answer It has advantages over both the other types of furnaces which
I think should be reflected in energy savings.
-260-
METALLURGICAL TREATMENT (QUENCHING AND TEMPERING) IN FLUIDIZED
BED OF TRACK CHAIN ELEMENTS RECOVERING THE RESIDUAL FORGING HEAT
Author : W. GRILLI
Head of project : W. GRILLI
Contractor : rTALTRACTOR
Address : Via per Modena, 152

I - 41014 CASTELVETRO
SUMMARY :
The standard method to forge and to heat treat links of undercarriage

of crawler vehicles provides for the following phases :
heating at high temperature (about 1200 C)

forging of links
cooling in air
heating at hardening temperature (800-850 C)
quick cooling - hardening
heating - tempering
quick cooling
In this thermal cycle three heatings are provided (at 1200, 850 and
550 * C) as well as three coolings at room temperature, making dis-
continous and expensive the manufacturing process.
According to the new suggested cycle, the steel bars, conceived for
the manufacturing of links are subjected to what follows :
. They are quickly heated into induction furnace at a temperature

of 1200 * C
. The are forged in subsequent deformations with a total forging time
not higher than 45 - 60 seconds and an end forging temperature
checked between 950 and 1050 *C.
. Forged links are cooled 2 - 4 * C/sec. down to a temperature of
800-620 -C.
-261-
. Thai links are stabilized or reheated by fluidized bed process
to the hardening temperature (800 - 850 C).
. Quick cool within the temperature range of 180- 350 C.
. Heat or stabilize in fluidized bed for 10 - 12 minutes at a
temperature of 400-550 C.
. Cool in water or air.
The new cycle allows to obtain remarkable energy savings, reducing

heatings and disconnecting at adequate temperatures the various
coolings. The fluidized bed process is used to heat the pieces
or to stabilize their temperature, since it shows greater advantages
than standard furnaces which will be here below reported.
Sommario
Nell'articolo che segue sono illustrate le linee essenziali del

progetto di recupero del calore residuo di stampaggio per ese-
guire il trattamento di bonifica per maglie per cingoli per trat-
tori.Tale procedimento consentira di ottenere nel trattamento di
17000 t/a di maglie, un risparmio di circa 930 tep/a, garantendo
nel contempo.il mantenimento o il miglioramento del la qualita ot-
tenuta coi sistemi tradizionali di bonifica.
1. PREMESSA
II progetto dimostrativo che verra illustrato nelle pagine seguenti,

e opera della I taltractor, che non e un'azienda costruttrice di
form', ma uno dei piu importanti produttori di componenti del
carro dei trattori ed escavatori quali:rulli, ruote tendicingolo,
ruote motrici, catene.
ANALISI DEI MATERI ALI UTI LI ZZATI DALLA I TALTRACTOR PER LA REALIZZA-
ZIONE DEI PROPRI PR0D0TTI
I materiali u t i l i z z a t i sono quelli t i p i c i della t r a t t o r i s t i c a , cioe-:
- acciai bassolegati da bonifica

- acciai bassolegati da cementazione
- ghisa meccanica e sferoidale
- bronzo
- gomma
Le analisi degli acciai sono sostanzialmente a base di: C,Mn, Cr,
Si, Mo.B.
I fabbisogni annui in varie sezioni e misure.oltre che analisi, rag-
giungono le 100.000 tonnellate ( 70% bonifica, 20% cementazione, 10%
altri tipi).
II fabbisogno in ghisa e bronzo raggiunge le 10.000 ton/anno.
-262-
3
- I^TI_TERMICI_ESEGUITI
Sono prevalentemente:
* la bonifica seguita da tempra superficiale a induzione;
* la tempra differenziale;
* la cementazione seguita da tempra;
* la globulizzazione e la ricottura subcritica.
I pezzi trattati sono alio stato grezzo di stampaggio o semilavorati,
ricavati da grezzi o da barre.
4. INCIDENZA DEL COSTO DI RISCALDAMENTO SUL COSTO DI TRATTAMENTO

E_yt_99I2_I9I*!il_9L_P?992II2z
Determinazioni ripetutamente eseguite hanno permesso di stabilire
che l'incidenza del costo di trattamento di bonifica sul costo del
prodotto,varia 6*10%.
Le spese energetiche sono valutate pari al 40-50% del costo della
bonifica. Conseguentemente il costo energetico per il trattamento
di bonifica incide 2,4t5%, sul costo del prodotto.
Detto valore tende ad aumentare in relazione agli aumenti crescenti
del costo del metano che sfiora ormai le 300 /mc.
Per gli altri trattamenti eseguiti, l'incidenza non e molto diversa.
Considerata l'importanza del trattamento termico sulla efficienza
del prodotto, sulla durata e sulla qualita dello stesso, l'ltaltrac-
tor si e posta il problema del contenimento del suo costo, sia diret-
to ( gas metano, personale e c c ) , che indiretto ( riduzione delle gia-
cenze di pezzi in lavoro, maggiore celerita del ciclo di lavoro e c c ) .
5
??95ni9.9I^9I?*Iiy9.II^5I9?
Trattamento metallurgico ( bonifica) in letto fluido di elementi della
catena ( maglie ) del trattore con recupero del calore residuo di stam-
paggio.
La maglia e un particolare molto complesso, come forma, e con sezioni
abbastanza diverse tra loro.destinato a lavorare in condizioni assai
difficili Fig.1.
5.1 Trattamento classico
II metodo classico di stampaggio e trattamento termico delle maglie

per catenarie di mezzi cingolati.prevede i seguenti passaggi:
- riscaldo ad elevata temperatura ( circa 1200C ) della billetta o
dello spezzone di billetta;
- stampaggio delle maglie;
-263-
- raffreddamento in aria;
- riscaldo alia temperatura di circa 850C;
- tempra in emulsione ( acqua olio ) ;
- riscaldo alia temperatura di rinvenimento di circa 550C;
- raffreddamento in emulsione.
In tale ciclo termico ci sono 3 riscaldamenti ( 1200-850-550C ) ,
e 3 raffreddamenti a temperatura ambiente , che rendono discontinuo
e costoso il procedimento di fabbricazione.
II ciclo termico descritto e rappresentato in Fig. 2
5.2 Trattamento proposto
Secondo il nuovo ciclo proposto i blocchetti di acciaio destinati
alia fabbricazione delle maglie vengono:
riscaldati rapidamente in forno ad induzione ad una temperatura di
1150t1200C, quindi sottoposti a stampaggio per deformazioni successi-
ve con un tempo totale di stampaggio non superiore a 45t60 secondi,
ed una temperatura di fine stampaggio control lata tra i 950 e i
1050C.
Le maglie vengono raffreddate con velocita 2*4 C al secondo fino
ad una temperatura di 830*620C, con eventuale ripristino.in oppor-
tuno forno di riscaldo e/o stabilizzazione della temperatura di tem-
pra (800*850C), in funzione delle caratteristiche del materiale da
trattare e dei risultati da ottenere a fine del ciclo termico.
Quindi il pezzo sara raffreddato rapidamente con velocita di 10-r15C
al secondo entro l'intervallo di temperatura compreso tra i valori
Ms e Mf del materiale, ( ossia tra circa 180 e circa 380C ) e suc-
cessivamente sara riscaldato e/o stabilizzato, per 10*12 minuti al-
ia temperatura che fornisce i valori di resistenza e di tenacita ri-
chiesti.
Seguira il raffreddamento in aria o in acqua al di sotto della tempe-
ratura Mf.
E' possibile nella seconda sosta del raffreddamento effettuare del-
le piccole varianti a seconda del tipo di struttura che si desidera
ottenere.
II ciclo termico descritto e rappresentato in Fig.3 ed e coperto, per
quanto riguarda le maglie, da un brevetto di utilitaltaltractor.
E' facile rendersi tonto del sensibile risparmio di energia, che di se-
guito viene calcolato.
5.3 Quantita di energia risparmiata sui due impianti Italtractor

II consumo medio in metano per bonificare maglie su un forno tradizio-
nale e:
8mc/ql = 80 mc/t
II consumo sul nuovo impianto viene stimato, col ciclo piu economico in:
1,5 mc/ql =15 mc/t.
II risparmio con il nuovo impianto e quindi : 6,5 mc/ql= 65mc/t.
1 Kg olio combustibile = 1,182 mc/metano (SNAM).
-264-
Risparmi annui energia per le due linee di stampaggio e bonifica ma-
glie Italtractor ( riferiti a 230 giorni lavorativi/anno; 20h/giorno)
1 Impianto : capacita produttiva 9700 t/a (circa 4 t/w teoriche;

R = 0,5*0,55 )
R = 65 x 9700 = 630.000 mc/a metano = 6 3 0 - 0 0 = 530 tep/anno
1 1.182.000
2 Impianto : capacita produttiva 7300 t/a (circa 2.5 t/h teoriche;
R = 0,6*0,65 )
R 2 = 65 x 7300 = 475.000 mc/a metano = 475 - 000 = 400 tep/a
1.182.000
Rt= 65 x 17000= 110.000 mc/a metano = 1.100.000 = 93u tep/a
1.182.000
II conseguente risparmio in lire sara percib ( metano a 300 /mc ) :

R = 1.100.000 x 300 = 330.000.000 /a.
e
In pratica, tenendo conto anche del risparmio di personale.di energia
elettrica, di materiali di consumo, di manutenzione e c c , nonche della
riduzione delle scorte.dovute alia riduzione dei tempi del ciclo di
lavoro, a seconda del ciclo termico seguito a parita di produzione,
i minori costi sono valutabili dell'ordine dei 460.000.000*520.000.000
/anno, pari a circa 30 per ogni Kg. di materiale trattato.
Considerando che 1'investimento globale e di 2.5*3 miliardi di lire,
il ritorno dell'investimento si ha in 5*6 anni con un tasso di redditi-
vita del 25-30%.
6
- 5t!Âj;ENERALE_DELLMMPIANT0
E1 rappresentato in Fig.4.
7
- CELTA_DEL_LETTO_FLUIDO
La scela dei forni a letto fluido per la stabilizzazione e/o il riscaldamen-

to alia temperatura di tempra e rinvenimento, e dovuta principalmente ai noti
vantaggi che essi presentano e cioe:
- Talta potenzialita richiesta per i due forni, rispettivamente 4t/h e 2,5t/h
- uniformita di temperatura mantenuta entro + 5C con conseguente possibility
di contenimento della deformazione delle maglie;
- elevata massa termica, in contatto con il materiale da trattare (3000*4000
volte rispetto a quello di un forno tradizionale);
- elevato scambio termico ( 5*10 volte rispetto a quello di un forno tradizio-
nale), con conseguente risparmio energetico del 35*50%;
- materiale trattato senza ossidazioni;
- grande flessibilita del forno che pub essere acceso o spento in qualsiasi mo
mento e la cui potenzialita pub essere facilmente adeguata a quella delle pr<
se, avendo una bassissima inerzia termica;
- assenza di fumi tossici.
-265-
8. ALCUNE CARATTERISTICHE MECCANICHE FONDAMENTALI DEL FORNO (COPERTO DA BREVETTO DI
UTILITA'_ITM) ; _
La Fig.5 rappresenta in pianta lo schema generale del forno.

Ciascun impianto e costituito da due letti fluidi funzionanti a metano.il primo
per il riscaldo o stabilizzazione alia temperatura di rinvenimento ( circa 550C),
E1 prevista la possibility di recuperare il calore residuo dei fumi che escono
dal forno di tempra per il riscaldo (Darziale) del forno di rinvenimento.
La Fig.6 rappresenta la zona di caricamento.
I pezzi prelevati da un robot dal nastro di uscita del la pressa, vengono carica-
ti ai ganci delle traverse di trasporto.quindi raffreddati fino a 620t830C,poi
inseriti nel forno di stabilizzazione.
La Fig.7 mostra i sistemi di estrazione delle traverse che portano appese ai ganc-
le maglie del forno,e il loro inserimento nei sistemi di raffreddamento.
Le traverse, estratte da un sistema di raffreddamento vengono inserite rispettiva-
mente nel forno di rinvenimento e sulla linea di ritorno.
La Fig.8 mostra una sezione trasversale del forno con la traversa porta ganci.
La Fig. 9 mostra il particolare del gancio e dei sistemi di blocco e rotazione di
questi.
9. ALCUNE CONSIDERAZIONI SULLE__PROVE_ESEGUITE
Sono state eseguite su un prototipo simulato in scala ridotta e in laboratorio.cor

lo scopo della definizione completa della specifica tecnica del forno e del pro-
cesso industriale.
9.1 Determinazione della legge di variazione della temperatura dopo lo
stampaggio ( necessaria per una corretta progettazione del forno).
In Fig.10 e rappresentata una maglia strumentata con coppie termoe-
lettriche per la misura della variazione della temperatura nelle va-
rie sezioni.In basso e visibile lo schema di parte dell'impianto.
In Fig.11 sono rappresentate le linee di raffreddamento dei pezzi pri-
ma e durante lo stampaggio.
In Fig.12 sono rappresentate le curve di raffreddamento in varie se-
zioni di maglie di diverse dimensioni.Si vede che in aria le tempera-
ture, nelle varie sezioni, tendono a differenziarsi sempre piCi.
In Fig. 14 sono rappresentate varie curve di raffreddamento,con di-
versi sistemi di 'raffreddamento del pezzo.
9.2 Diagrammi CCT e TTT dell'acciaio utilizzato ITM7 ( necessari per definire,
nel modo pip esatto, il ciclo termico da seguire.
In Fig. 14 e 15 sono rappresentati rispettivamente i diagrammi CCT e TTT
dell'acciaio ITM7 a grano fine, e in Fig. 16 il ciclo termico di affina-
mento del grano austenitico con pendolazione :
-266-
10
-?IH!:TATI_OTTENUTI
I risultati sino ad ora ottenuti hanno confermato la validita del progetto

proposto:
resistenza meccanica 95*105 Kg/mmq
resilienza a + 20C 6*6,5 Kgm
- 20C 4*4,4 Kgm
dimensioni del grano 8*8,5 ASTM
struttura M+20SB
Pezzi privi di cricche od altri difetti di tempera.
Alcune catene prodotte sono in esercizio e sino ad ora si stanno comportando
normalmente.
II ciclo prevelentemente utilizzato, per ragioni di sicurezza, e quello che
prevede il raffreddamento delle maglie stampate sino a circa 600C, e il suc-
cessive riscaldamento a 800*850CC ( pendolamento ),che non e
quello ottimale ai fini del risparmio energetico.
11.CONCLUSION
Si stanno attualmente eseguendo trattamenti termici con acciaio a grano con-

trollato ( al Nb-T ) e cid al fine di evitare il pendolamento di cui sopra.
Questo permettera di realizzare il massimo del risparmio energetico previsto
a parita di caratteristiche del prodotto ottenuto e con una grande affidabilil
per quanto riguarda la costanza delle caratteristiche, e la dimensione del grj
no e l'integrita del prodotto stesso.
II forno e in fase di avanzato studio di progettazione.
Fig. 1 - Maglia per Cingolo

Trattore
-267-
2_ "AUSTEMPERING"DAA.T.
ACCIAIO ATTUALE
ITH 7
KcT
(1000 i 1050C )
DA B.T. (710 *840C)

-
0,36C1,25Mn0,3Cr
i ~ 2" TEMPRA DIRETTA DA A.T.
RISCALDO 1235C STAMPAGGIO.1200C ESPULSIONE MAGLIA (1000 * 1050C) RINVEMIMENTO
A- (Tradizionale) dj'finale^ 4 ASTH 1100C v 550C
^ I D A B .T. (710 f 840C)
MW, AFFINAMENTO GRANO CON
A- RISCALDO 1200C a
STAMPAGGIO 1165C
dT finale= 5 ASTH
ESPULSIONE MAGLIA
> 1060C RAFFP. LENTO IN ARIAFINOA
, 635C0 RAPIDO IN LETTO
AUSTEMPERINGDA
;
BH 1 T -
TEMPRA DA RINVEMIMENTO
t&B
S " FLUIDO 0 SALI FUSI 810C " 550C
AFFINAMENTO GRANO CON

*Wm TEMPRA DA 840 C RINVEMIMENTO
RAFFR. FINO A 20C
(Tradizionale)
(Tradizionale) ~ 550C <D
ACCIAIO FUTURO
MICROLEGATQ NbTi
I
RISCALDO 1200C STAMPAGGIO 1165C ESPULSIONE MAGLIA
af TEMPRA DIRETTA DA A.T.
RIHVENIMENTO
(950 i1000C)
B dr = 6*8 ASTM ~1060C Atp"
^
DA B.T. (710 840C) ~ 550C -@
Table 1
I BOMIFICA
OocT
M'
N '
w
11
11
&'
1 X *i' AcJ
1 x. y
K 1
1
X^\. /
/ 4c 1
~'
fa."
\ /
w
1 >v /
w
1 Nv /
kg'
1 N. /
W
\ / te
01 n n ftt f w J*'
f*l ah
Fig. 2
| C tf I I TWM>C |
jtf_
4 * #to*
Fig. 3
269
2 R>C MO A/s=jMJfc> SPrzzou/.
3 . f a c t s * ^noMAriM * ' S M f t 4 i o
A MASneo -reASftxcmroee c o u . e < 4 n e u ) o PMOf.
5 . peesc* AtnowAncA ji C * J . I C 4 T V A , -
T O C H I TiuiUJnvtao: 6.STMtfHe i C A B C O E SC ARIC O P B Z I ALÂLC SHVO.

qUATOM AC W o . '
? -TOOHO SGfcVIUZCAElOHff TC MrKKATUtZA A00%MoC .
8 . iMromvo B A I T r t i c w iw nt> r e a o f-Ev.iefnreA,

? YcCMO ^TABIU2Z*2iC H*e"Te*lfeKMUfEA RMVB4((*l7b
CtfLCA i o o ' c .
Fig. 4
P'&UTA -PEU.O SC Ut MA OBMg/kLK K L f a H o
Fig. 5
270-
ZoulA tt CAglCAMtWO t L W > W
Fig. 6
SISTBMA M a S A V O M S b E L U TOW**** PbftV4C*Mci AAl_R>A*JO V (MS*iMeNTD ftJCLKJcrKXAM I I H M I N I I M V I I U
4? u IT y
BrTllI
Fig. 7
-271-
?
A
n_ JEL ir-n-
=H
M-
u
J17
ÂAAAJb 7^
minnrini
r TTTTTTTT
H i i
QflDflQQQfl AJ.
/
Fig. 8
Fig. 9
-272-
Mqglia. gfrtH*a1grtiv LtpKr,
I
*1h Peio 5p<Uor*
A B c V ' Stof b. ilsna
-1o} n 4 4 W 2 > *r2.

toSc 4 2 w I f 19 IT Vlo SF-1U
1ofc<i 4 1 1 ) w 1") If S.fF .iu
5204 f V k*k' 1* J, 4e2ir
1cUM M
?tf 1* 2.TS 46x310
pSSLpfoSrft
T
UW*0UBI3,0ÂfÛn e > . OC
[itt.>l.)
84oo
Fig. 10
Ant>AEnra DELIA TSMrr_ -ncxf-

RATUeA U K O S M I J t PRIMA,
bUHAWTK m DOPO l-O g7i>T>l>q-.
sunwtcjiu).,
5ASTMJ
5ox COAiao .* a*.V-
IH Am*,
Fig. 11
273-
IHvAfciE. S G Z I O H I P E a
F i g . 12
COM VABI MC221 b l

maalil V bJ C ) IIM.HIO
9oo /, ii ( ) daccettaO
i. " ( ) JeeeeHjo+1q?t olto
II it ( ) I'IUM. H z 0t1O/i olio
-w\aglia 1ofi>c (<w) 4occe Hio pareMmeutoe
C+VKK qiA4McU<'iM4t;')
SI no 91 l
Si no SI SI
se*t>io eli "-ti a ou*i.

ACClilO ITM
c Mil i ? P cr
cWmita % o.f !<.zt 0.24 e.oar e,et .4
rt
>c 4^*^l*c^7^st Auir,ôo C K = rnj'ii.cM pnMv o<ur M. 6 MSTM DIA60AMMA C C T
-Wo ACCIAIO lTM^f A
<*C/S<>-T3)
c/mi'K,
HRc.
t+^MU^^+A*^.^^**Mia^^^^
te4 4oS
-SOCT
to*
F i g . 14
G0*UtBZ2& C L GfUUO 64STM
URc
-I
-
P 1 AC
Q AMKA T T T
_2Q AC C IAI0 iTM^f
-22
--.
--28
1o iP 1o*
TEMPO stcouw
F i g . 15
275
ACCIAIO ITM7 caeAwoAt)cr-AASTMr
aliayaxutia CCT riafnita. (Mrte-tratte99K(ta)
Jet 7g6*C/Afc*7*9VA*f.'tgaCWr-<oWil.c*>yttm> guT.KI.4 -ASTM
1ooo
\ f ' .<25tymm.Ca*miM..see.J
800
*~
5oo
**>
3oo
2oo
/loo
MAC
6 Illl J I
1o 1o
AD* 1o
See.
ClCLoTCttMcoi>l APClHA-
*E MVO P E L QRAWO AOSTE-- Fig. 16
fJ/TlCO C*- P & i D < X A 2 l O f i e .
DISCUSSION
What we have heard suggests to me that when one has complex
pieces of metal which have, shall we say a difficult shape,
it would appear that the fluidised bed offers a very good way
of effecting the heat treatments required to get the necessary
properties and this must be happening. I mean this requirement
must be very extensive indeed in industry. All kind of steel
and steel components have to be heat treated. It seems to me
that if this is successful as initial results show that it
should have very wide application so that is really one
question I would like to ask Mr. GrilLi :
Does he anticipate that this could have very wide application
in other industries and the second question I'd like to put
to hi m is :
Could he perhaps indicate what is being the biggest difficul-
ty he had to overcome so far?
Answer Per quello che riguarda le responsabilita di applicazione, io

direi che in tutti gli impianti di stampaggio sufficientemen-
te automatizzati dove vengono prodotte delle serie di pezzi
abbastanza consistenti dovrebbero esserci delle possibility
di applicazione. Quindi nel settore della trattoristica, nel
-276-
settore del I'automobile dove si tende ad automatizzare il fun
zionamento dello stampaggio e nel medesimo tempo i lotti di
pezzi che vengono stampati sono sempre lotti molto consistenti
dovrebbero esserci delle possibility di applicazione. Infatti,
qualche azienda ha gia preso anche contatto con noi e ha
dimostrato un certo interesse per questo tipo di discorso.
Le difficolta sono state abbastanza grandi perche direi che
da parte dei costruttori a letto fluido non si trovano molte
persone che abbiano delle conoscenze metallurgiche e che con
sentano di applicare questi impianti al settore dei tratta
menti termici, cioe si trova la persona che magari ha una
buona conoscenza del I'impianto perd manca la conoscenza metal
lurgica perche forse, ripeto forse, non so se queste notizie
sono completamente esatte, non c'e ancora stato un sufficiente
numero di applicazioni del letto fluido nel campo del tratta
mento termico. Di solito i costruttori dei forni per tratta
menti termici sono anche degli ottimi metallurgist! nel caso
specifico invece questa esperienza non I'ho trovata salvo in
quelche rara eccezione. Quindi diventa difficile parlare lo
stesso linguaggio, usare lo stesso linguaggio, ecco, a noi
sono interessati si gli aspetti impiantistici del forno ecc.
per6 ad un certo punto abbiamo dovuto dare per scontato che
chi ci forniva il forno avesse queste conoscenze cioe sapesse
come doveva fare il forno anche se poi in pratica abbiamo do
vuto lavorare insieme per definire una specifica tecnica, ab
biamo dovuto fare delle prove di simulazione, abbiamo dovuto
fare una serie di esperienze in comune con questo possibile
fornitore. Per6 non tutte le persone cui ci siamo rivolti ave
vano delle esperienze del trattamento termico.
Comment Chairman :
There are two very interesting replies and it is also quite a
challenge it would appear that those who have got expertise in
designing fluidised beds for this type of application it often
happens that they have never thought how it could apply the
metallurgical knowhow to the knowhow about how to manage a
fluidised bed at a constant temperature and variant tempera
ture. This, I think, is a point we should certainly record,
it is an important point.
277
CLOSING ADDRESS
by
Dr. Michael DAVIS
The objective of this meeting was to give the opportunity to contractors

to present progress achieved and this is particularly valuable because
it reinforces the link between the contractors and individual project
leaders. We were pleased to have experts working for the Commission,
Mr. Knobbout and Mr. Carrasse to participate in our meeting. We also
intended that this meeting would provide an opportunity for different
contractors to get to know each other and exchange experiences. I
think that this objective has been achieved. Moreover, we have attemp-
ted for the first time to provide a bridge between research and develop-
ment on one hand and demonstration on another. This kind of interaction
is important. I was struck by the fact that several presentations
recognised the importance of economic aspects. Economic viability is
of prime importance in demonstration projects, but it is also very
important that the endpoint of research and development should lead
to economic results.
Certain research and development projects, for example that of the

National Coal Board on hot gas drying, seem to be sufficiently
advanced to merit demonstration. Other research projects, such as
the Creusot Loire work, enabled a pilot plant to resolve quite a number
of technological problems. Some scaling up remains now to be done on
projects like this.
Another group of research projects studied by Universities are at the

stage of trying to understand the process involved because they are not
very well understood. We think that in this type of projects those
responsbile should as soon as possible collaborate with industrial
partners so that they can better orientate their researches towards
possible applications.
-278-
I have attempted briefly to define what we mean by demonstration.
Those who are carrying out this type of project are attempting to
optimize known technology with specific and welldefined objectives
in fairly short term. In consequence their preoccupation is necessa
rily linked to economic viability and to the possibility of multipli
cation of such projects throughout the Community.
Fluidised beds which have attained the stage of demonstration seem

often to be rather complex and require specialized personnel to make
sure that they work. We must not overlook this aspect because it would
be regrettable if the potential for energy savings were frustrated for
lack of suitably trained staff.
We have heard some very interesting examples of work where automatic

processes and control processes have been applied.
It is very important that the results of demonstration projects should

be widely diffused and disseminated. I hope that contractors respon
sible for these projects will do their utmost to publicize their re
sults. For its part the Commission intends to disseminate results
very widely. We have just started producing easily readable brochures
about completed projects. The proceedings of this meeting will, of
course, be published.
I could have wished that time had allowed our discussion to be taken
further this afternoon, but this meeting is breaking new ground. How
ever, vritten contributions to the discussion would be welcome.
In concluding this meeting I would like to thank all of you for coming
here and particularly the authors of papers for giving their time and
for making some very clear presentations. I would like also to thank
Mr. Knobbout and Mr. Carrasse for chairing yesterday's meeting.
Our interpreters have helped greatly with our work in a subject for
which it is not always easy to give precise interpretation.
279
F L U I D I Z E D BED SYSTEMS
Proceedings of the Contractors' Meetings held in Brussels
on 12-13 October 1982
In order t o enhance the rational use o f energy and energy conser-

vation in t h e European C o m m u n i t y , it is necessary, among o t h e r
things, t o p r o m o t e the development and demonstration of f l u i d -
ized bed combustors and f l u i d i z e d bed heat exchangers. The C o m -
mission has t h e r e f o r e , concluded w i t h industries, public organiza-
t i o n s and universities o f t h e EC Member States a n u m b e r o f R & D
contracts dealing w i t h these subject areas. The progress o f w o r k
achieved in these contracts was reported at t h e c o n t r a c t o r s ' meet-
ing, t h e proceedings o f w h i c h c o n s t i t u t e this v o l u m e , aimed at
i n f o r m i n g Member States, research bodies and other interested
parties a b o u t existing development w o r k on f l u i d i z e d bed c o m -
bustors and heat exchangers. A l t h o u g h f l u i d i z e d bed c o m b u s t i o n
is n o t expected t o decrease overall energy c o n s u m p t i o n , it allows
t h e user t o reduce his dependence o n i m p o r t e d oil and gas, t o pro-
vide f o r a better diversification o f fuels and t o c o m p l y w i t h en-
v i r o n m e n t a l requirements. Fluidized bed techniques may also be
used in heat exchangers f o r waste heat recovery, in particular f o r
a p p l i c a t i o n in d i r t y and corrosive industrial fumes.


Fluidized Bed Systems

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

Fluidized Bed Systems

Încărcat de

Drepturi de autor:

Formate disponibile

Energy R&D and Demonstration Programmes

Fluidized Bed Systems

D. Reidel Publishing Company

Fluidized Bed Systems

PARL EUROP. Biblioth.

D. Reidel Publishing Company

Dordrecht / Boston / Lancaster

Organization of the Contractors' Meetings by

Published by D. Reidel Publishing Company

Sold and distributed in the U.S.A. and Canada

In all other countries, sold and distributed

All Rights Reserved

The main objectives of the Community's energy policy are to secure an

In the framework of its strategy for scientific and technological re-

In the European Community, 35% of the final energy is consumed by indus-

A number of classical fluidized bed combustors working at atmospheric

Although fluidized bed combustion is not expected to decrease the overall

In order to enhance the rational use of energy and energy conservation in

Before ending these short introductory remarks, I would like to express my

Dr. Albert STRUB

SESSION I - FLUIDIZED BED COMBUSTION

J. BANNARD The National Institute

SESSION II FLUIDIZED BED HEAT EXCHANGER

SESSION III : FLUIDIZED BED HEAT EXCHANGERS

R. DENOISEUX Metallurgie Hoboken-Overpelt

Dr. M. DAVIS, Director DG XVII

Dr. J. CARRASSE SYNTEC, France

Attendance List VII

Summary of the proceedinqs 1

PART I - R & D PROJECTS

SESSION I - FLUIDIZED BED COMBUSTION

Contractors' meeting held on 12 October 1982

The desiqn and development of a small-scale fluidised bed

Production of hot qas for drying processes using fluidised

Further exnerimental development of circulating fluidised

The burninq of RDF and industrial wastes in fluidized bed

The development of a demonstration furnace for the

Contractors' meeting held on 12 October 1982

Amelioration des echanges thermiques dans les lits

Etude de la circulation autoentretenue

Fluid bed high temperature gas/gas heat exchangers

Fluidized bed heat exchanger filter for waste heat recovery

PART II - DEMONSTRATION PROJECTS

SESSION III - FLUIDIZED BED HEAT EXCHANGERS

Contractors' meeting held on 13 October 1982

The M.H.O. process for the combustion of shales in fluidized

Combustion of carbonaceous washery shales in fluidized bed

Total precalcination by circulating fluid bed

Fluidized bed boiler used in the production of bricks

Fluidized bed combustion of waste acid tars

Atmospheric fluidized bed combustion boiler for electric

The application of fluidized combustion of coal to a low

Use of fluidised bed furnaces in steel heat treatment

Metallurgical treatment (quenching and tempering) in fluidized

P. Zegers , J. Carrasse and J.A. Knobbout

In the European Community, R and D on fluid bed combustion is carried out

STATE OF THE ART

Before discussing the progress of these projects a short description of

A fluid bed is basically a container with fine inert solid particles. In

* Commission of the European Communities

. Combustion takes place at relatively low temperatures (700-900 C ) .

Fluid bed combustors

Production of hot gas for drying purposes represents a significant use of

a) In a test installation of 400 kW the concept design for a fluid bed