07 revamps

14/8/07

12:42

Page 1

revamps
2007

ptq

Supplement to PTQ

Improving
Perfection
Enhanced Reactor Internals
from Topse
Improved Performance

Compressed Design

Stability over broader operating ranges - especially

for high gas rates as applicable in hydrocrackers

Both trays and mixers enable loading of

more catalysts

Easy to Install

Massive References

Installation can be done in as little as 4 hours per tray.

Fewer parts to handle and wedge fastening instead of
bolting, resulting in shorter downtime during shut-downs

More than 350 reactor internals installed in

130 hydrotreaters

The Catalyst and Technology Company

www.topsoe.com
HALDOR TOPSE A/S Denmark Phone +45 45 27 20 00 Telefax +45 45 27 29 99
HALDOR TOPSE INT. A/S Japan Phone +81 3 5511 8115 Telefax +81 3 5511 9115
HALDOR TOPSE INT. A/S India Phone +91 11 4175 0081 Telefax +91 11 4175 0252

HALDOR TOPSE A/S Russia Phone +7 495 629 6350 Telefax +7 495 956 3275
HALDOR TOPSOE, INC. Houston, TX, USA Phone +1 281 228 5000 Telefax +1 281 228 5109
HALDOR TOPSE INT. A/S Peoples Republic of China Phone +86 10 6515 8886 Telefax +86 10 6512 7381

ptq revamps
PETROLEUM TECHNOLOGY QUARTERLY

ptq

Editor
Ren G Gonzalez
editor@petroleumtechnology.com

Production Editor
Rachel Zamorski
production@petroleumtechnology.com

Graphics Editor
Mohammed Samiuddin
graphics@petroleumtechnology.com

Editorial
PO Box 11283
Spring TX 77391, USA
tel +1 281 374 8240
fax +1 281 257 0582
Advertising Sales Manager
Paul Mason
sales@petroleumtechnology.com

3 Identifying poor heat exchanger performance

Mark Fernsby and Abe DuPont Natref
Tony Barletta and Steve White Process Consulting Services Inc

9 FCC gas concentration unit stripper revamp

Dave Langdon Total Lindsey Oil Refinery
Scott Golden and Edward Hartman Process Consulting Services Inc

15 Lube vacuum column revamp

Kevin Basham Marathon Petroleum Company LLC
Edward Hartman Process Consulting Services Inc

21 Reducing a plants turnaround schedule

Mike Lippold BJ Process & Pipeline Services

Advertising Sales
Bob Aldridge
sales@petroleumtechnology.com

Advertising Sales Office

tel +44 870 90 303 90
fax +44 870 90 246 90

27 Revamping atmospheric crude heaters

Michael Whatley Navajo Refining Company
Scott Golden and Jason Nigg Process Consulting Services Inc

31 Optimising turnaround maintenance performance

Publisher
Nic Allen
publisher@petroleumtechnology.com

Circulation
Jacki Watts
circulation@petroleumtechnology.com

Salah Massoud Elemnifi Arabian Gulf Oil Company

Farag E Elfeituri University of Garyounis

37 OSBL considerations for expansion projects

Dominic M Varraveto Burns & McDonnell

Crambeth Allen Publishing Ltd

Hopesay, Craven Arms SY7 8HD, UK
tel +44 870 90 600 20
fax +44 870 90 600 40

Crambeth Allen Publishing gratefully acknowledges the contribution of Process

Consulting Services Inc to the publication of Revamps.

Cover: The lube vacuum column revamp executed at Marathons Catlettsburg

(Kentucky, USA) refinery led to several benefits: sharper lube products
fractionation, lower gas oil contaminants, improved asphalt properties and a
higher unit charge rate. Details are given in the article starting on p15.
Photo courtesy of Marathon Petroleum Company LLC.

2007. The entire content of this publication is protected by copyright full details of which are available from the publishers. All
rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any
means electronic, mechanical, photocopying, recording or otherwise without the prior permission of the copyright owner.
The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while
every care has been taken in the preparation of all material included in Petroleum Technology Quarterly the publisher cannot be held
responsible for any statements, opinions or views or for any inaccuracies.

SENTRY
GUARD CATALYSTS & SUPPORT

Protect your
catalyst.
Protect your
bottom line.
Introducing the most
complete guard catalyst and
support portfolio available
Now theres Sentry Guard Catalysts &
Support -- an entire family of products from
Criterion Catalysts & Technologies designed
to protect your hydroprocessing operations
from the contaminants found in opportunity
feedstocks. Fouling and contamination can
negatively affect your return on investment
thats why Criterion has committed R&D
resources to address these problems you
face every day.
The Sentry platform offers solutions for the
prevention, diagnosis and resolution of
contamination issues -- protecting both your
catalyst investment and your bottom line.
Contact us to learn more.

www.criterioncatalysts.com
2007 Criterion Catalysts & Technologies L.P. Sentry is a trademark of
Criterion Catalysts & Technologies. (cri703)

Identifying poor heat

exchanger performance
Identifying root causes of poor crude unit debutaniser reboiler performance and low-cost
solutions that help meet exchanger original design conditions. Of primary importance is
the impact factors such as fouling have on the service heat-transfer coefficient
Mark Fernsby and Abe DuPont Natref
Tony Barletta and Steve White Process Consulting Services Inc

eat exchangers are designed

with engineering tools that
allow the designer to evaluate
the influence of various parameters
on projected performance and select
the optimum combination. However,
computer model results are just
calculations and only reflect actual
performance if the model assumptions
correctly predict what actually happens
inside the exchanger. Consequently,
when an exchanger does not perform
per design, the model assumptions
are not correct. Heat exchanger
performance, just like that of any other
refinery process equipment, depends on
specific equipment design and not on
the ideals of the computer model.

Unstabilised
naphtha

Gas oil
product

42 kg/cm2
steam

Stabilised
naphtha

Figure 1 Crude debutaniser reboiler system

Reboiler system
In 2000, National Petroleum Refiners of
South Africa (PTY) Ltd (Natref, a joint
venture of Sasol and Total South Africa)
commissioned a major crude/vacuum
unit expansion. As part of the revamp, a
new reboiler was installed on the crude
debutaniser to provide increased duty to
stabilise the increased production from
the preflash and atmospheric crude
columns. Total reboiler heat input
needed to increase to remove enough of
the C4s to meet the light straight-run
(LSR) Rvp specification for gasoline pool
blending. However, the debutaniser
reboiler heat input was not sufficient to
meet the light naphtha Rvp specifications
after startup, particularly when processing
light crude blends. The debutanisers
maximum throughput was limited to
90% of design when the reboilers were
clean. While in operation, fouling
resulted in a 3.5% reduction in reboiler
duty per month (equivalent to a reduced
maximum throughput of about 750 bpd
every month). At times, this began to
limit the overall crude unit throughput.
The revamped debutaniser column
was designed to operate with two
reboilers. One used gas oil product. Its
heat input was therefore set by product
yield. This reboiler, which had not been
mechanically altered during the revamp,

www.eptq.com

was operating satisfactorily. The new

steam reboiler was designed to supply
the balance of the debutaniser heat
requirements during normal operation
and the total column heat requirements,
so the gas oil reboiler could be taken
offline for maintenance. This larger
reboiler was designed for 42 kg/cm2
pressure steam (Figure 1) to replace one
using 16 kg/cm2 pressure steam. After
commissioning, the exchanger had a
calculated service overall heat-transfer
coefficient (U) as low as 25% of design.
Various field tests were done to evaluate
potential root causes. The exchanger
was eventually taken out of service and
inspected to ensure the actual design
conformed to the drawings and assess
whether the shell or tube sides were
badly fouled. However, only a marginal
performance improvement was noted
when the exchanger was put back into
service. Even when the reboiler was
clean, its U value was less than 25% of
design clean performance. Although the
reboiler was never significantly fouled,
its performance was sensitive even to
small amounts of fouling compared to
the other reboiler.
It is common to speculate on
potential causes when equipment does
not meet expected performance. In this

case, theories included inherently poor

heat transfer when using U-tubes instead
of straight tubes, insufficient shell-side
fluid circulation resulting in excessive
vapourisation and two-phase flow
irregularities in the reboiler return
piping. Another was that the exchanger
simply was not big enough and a new
larger exchanger was needed. Many tests
were conducted to investigate these and
various other hypotheses that could
explain the poor performance.
Determination of root causes must
begin with a review of the specific
system and equipment design, and the
application of fundamental engineering
principles. As engineering tools become
increasingly complex, sophisticated and
easier to use, such as computational
fluid dynamics (CFD) models, engineers
have a tendency to rely on these models
and search for complex causes and
solutions. This should not be a surprise
given the increasing focus on
engineering tools rather than on
understanding fundamental equipment
operating principles. Yet, understanding
these principles remains the single most
important factor in the design of process
equipment and troubleshooting when
the equipment does not perform as
expected.

PTQ REVAMPS 3

Independent Advice.

Global
Reach.
;I@M@E><O:<CC<E:<

Proven methodologies. Unique plant-wide software

applications. Unparalleled technical expertise. Thorough
understanding of key drivers in the global business environment
and within oil and gas, refining, or petrochemical facilities.
Independent consulting. Experienced people. Expert strategic
advice. Combine all of the above attributes, and youll see
what makes KBC strong.

KBC is the leading independent global consulting organisation with over 300 clients
in 50 countries. After nearly three decades of experience, we understand the complex
issues facing processing facilities. KBC provides a comprehensive range of service
offerings, which are custom-designed to reflect your needs while focusing on successful
implementation and sustainable competitive advantage.

For more information on how KBC can help you achieve Operational or Capital Excellence,
contact us at:
AMERICAS +1 281 293 8200 EMEA +44 (0)1932 242424 ASIA +65 6735 5488
salesinfo@kbcat.com www.kbcat.com

Outlet

Steam
in
Steam

hl

Condensate
out
Inlet

Condensate

Figure 2 H-Shell exchanger

Reboiler basics

Figure 3 Reboiler liquid head

Potential reboiler problems can generally

be separated into three areas: steam-side,
process-side and exchanger design.
Generally, there are only a couple of
steam-side problems, the most common
of which is flooding the exchanger with
condensate,
which
reduces
the
condensing surface area. Potential
process-side problems include a high
system pressure drop, excessively low
and high circulation rates, and slug flow
in
the
reboiler
return
piping.
Thermosiphon reboilers circulate tower
bottoms fluid through the shell side of
the exchanger based on system
hydraulics. The liquid level in the
bottom of the column and the globe
valve at the inlet of the shell side (if
present) is used to control circulation.
Too low a circulation rate increases the
percentage of vapourisation, which
reduces the heat-transfer coefficient, and
too
high
circulation
reduces
vapourisation, which can lead to slug
flow in the return piping. As for potential
exchanger design errors, many have
caused low duty.
Exchanger design problems can
reduce
the
service
heat-transfer
coefficient by affecting inside and
outside film coefficients. Thermal
conductivity of the tube is small
compared to the film coefficients.
Therefore, this term can be ignored.
Thus,
the
clean
heat-transfer
coefficient (UC) is calculated from
Equation 1 (Eq 1):
1 = 1 + 1
Uc
hi
ho

Eq 1

where:
UC = heat-transfer coefficient (clean)
hi = inside film coefficient
ho = outside film coefficient

Computer models calculate film

coefficients based on the specific
exchanger design, fluid properties, model
assumptions and equations. For example,

www.eptq.com

heat exchanger models assume the shellside fluid flow is distributed so that the
total exchanger surface area is utilised for
heat transfer. In reality, the specific
exchanger design will determine whether
the fluid entering the exchanger is
uniformly distributed or not. In the case
under consideration, the new steam
reboiler design UC was 1355 kcal/h
m2 C, but actually achieved only 300
kcal/h m2 C.
Clean exchanger coefficients do not
include
fouling
resistance.
The
exchanger
service
heat-transfer
coefficient includes fouling resistance
on the inside and outside of the tubes.
Total resistance to heat transfer is the
sum of the shell-side film coefficient,
tube-side film coefficient and the
fouling resistance. For a crude unit
debutaniser steam reboiler, fouling
resistance is generally 2535% of the
total resistance, with shell-side fouling
more common. It is rare to have any
significant fouling inside the tube when
using steam.
In practice, inside and outside fouling
resistances are lumped together in an
overall fouling resistance commonly
referred to as a fouling factor. The dirty
(or service) heat-transfer coefficient can
be calculated as shown in Eq 2:
1 = 1 + R
F
UD UC

Eq 2

where:
UD = heat-transfer coefficient, dirty
UC = heat-transfer coefficient, clean
RF = overall fouling factor

Fouling begins as soon as an

exchanger is put into service. Crude
unit debutanisers can foul badly on the
shell side from corrosion products and
water in the feed from the crude unit
overhead receiver. At one point in the
case study, the reboiler UD was only 110
kcal/h m2 C.

Reboiler design
The new steam reboiler was a TEMA Hshell type designed for a low pressure
drop with no vertical baffles (Figure 2).
H-shell exchangers have two inlet and
two outlet nozzles, with a horizontal
baffle separating the inlet and outlet
nozzles. A partition baffle in the middle
of the exchanger essentially splits the
shell side into two separate sections,
each taking flow from its own inlet
nozzle.
H-shells can be designed with or
without vertical baffles. The maximum
allowable exchanger pressure drop
depends on the overall system design.
When the column liquid level above the
centreline of the exchanger is low, the
maximum allowable pressure must be
low, as the available head is limited.
However, the maximum pressure drop is
sometimes intentionally specified very
low because of conservatism or design
guidelines. The consequences are a low
heat-transfer coefficient, large surface
area and increased likelihood of poor
shell-side flow distribution. The H-shell
exchanger pressure drop must be
balanced against the resultant heattransfer coefficient.
The reboiler maximum allowable
pressure drop depends on the overall
system design. Process-side hydraulics
must be carefully evaluated at the design
stage so the exchanger design is not
compromised. Shell-side fluid circulation
depends on the available liquid level,
exchanger pressure drop, density
difference between the liquid in and
mixed phase outlet, and the piping
system pressure drop. Often, revamps
are constrained by the existing vessel
skirt height, liquid level and reboiler
return nozzle location (Figure 3). Ideally,
shell-side circulation rates should result
in reboiler outlet conditions with 25
35% vapour in the mixed phase.
Balancing the exchanger pressure
drop and heat-transfer coefficient is

PTQ REVAMPS 5

FC

42 kg/cm2 steam

BFW
TC

PC

LP
steam
LC

LC

HP
condensate

Figure 4 Reboiler control system

Skid bars
on bottom

Inlet nozzle
on bottom

Plan view
(top)

Impingement
baffle

Skid bars

Figure 5 Exchanger bundle

critical. Calculating the shell-side
pressure drop accurately requires a
model that predicts the phase change as
the shell-side fluid flows upward
through the reboiler and vapourises.
Since this calculation is subject to a
number of inputs and uncertainty
inherent in the correlations, it is not
unusual for the process engineer to
specify a conservatively low maximum
allowable pressure drop. Moreover, the
exchanger design engineer will also

6 PTQ REVAMPS

provide some margin. Often, not

surprisingly, the result is an exchanger
that is designed larger than necessary
without baffles. All this increases the
likelihood of shell-side flow irregularities
and poor performance. H-shell reboilers
or condensers designed without baffles
are prone to lower-than-expected heattransfer coefficients.
Following the initial startup of the
unit, the maximum observed duty was
much lower than the design. The steam
flow-control valve was nearly wide open,
pushing the condensing temperature to
maximum. The original exchanger data
sheet clean heat-transfer coefficient (UC)
was calculated at 1355 kcal/h m2 C, with
a design service heat-transfer coefficient
UD of 627 kcal/h m2 C. The highest
average
monthly
heat-transfer
coefficient, UD, achieved was only 391
kcal/h m2 C, with the exchanger
operating at or below 250 kcal/h m2 C
much of the time.

Identifying the problem

When something does not perform
according to design, the most likely
causes should be investigated first. Eq 1
shows that exchanger performance
depends on both shell- and tube-side
heat-transfer coefficients. Since steam
reboilers generally have a high inside
heat-transfer coefficient except when
there is a large amount of superheat or
condensate floods the tubes, effectively
reducing surface area tube-side

problems are not the cause of poor

performance. Moreover, the Natref
system was designed with a steam
desuperheater, and field testing
confirmed the desuperheater was
performing per design. (Figure 4).
Tube-side operation depends on the
specific control system and condensate
drain system. In this example, steam
flow was controlled to the reboiler based
on required duty. As duty goes up, the
steam rate increases, with the flowcontrol valve opening to allow more
flow. As the valve opens, the pressure
downstream of the control valve
increases. Once the valve is fully open
and the downstream pressure is at
maximum, no more heat can be added.
As pressure downstream of the flowcontrol valve increases, the condensing
temperature increases, raising the
exchanger LMTD. The tube-side
condensing pressure can change from a
minimum of approximately 15 kg/cm2
condensate header pressure to a
maximum of 39 kg/cm2 steam pressure.
Pressure downstream of the flow-control
valve is a good indicator of exchanger
performance, because the higher the
pressure needed for a given duty, the
lower the heat-transfer coefficient. Even
with the flow-control valve wide open,
the steam flow rate was low and so was
the exchanger duty.
Condensate flooding is the other
common problem. Condensate flows
from the exchanger into the condensate
drum. As long as the condensate level is
in the external drum, the exchanger is
not flooded with condensate. But once
the external drum is full, the condensate
level is inside the exchanger, reducing
the surface area. Field tests showed the
condensate level was in the external
drum and not flooding the exchanger.
This was also confirmed through
neutron baskscatter tests that were
conducted on the reboiler.
The exchanger bundle was removed
from service and visually inspected to
determine if fouling was the problem.
Visual inspection and poor performance
after cleaning showed that fouling was
not the root cause of poor performance.

Exchanger design
Poor shell-side performance was
therefore postulated. The exchanger
was a six-metre long U-tube design
with two horizontal baffles and no
vertical baffles. The exchanger design
was evaluated using Heat Transfer
Research Institutes (HTRI) proprietary
IST model. Using reasonable fouling
factors, the calculated performance was
much better than actual duty.
Review of the exchanger design
showed mechanical features that caused
poor shell-side fluid flow distribution
through the bundle. Poor flow patterns
can cause shell-side fluid to bypass

www.eptq.com

portions of tube surface. The inlet nozzle

had an impingement plate, which
restricted flow into the tube bundle.
Furthermore, Figures 5 and 6 show the
exchanger was designed with skid bars
running its complete length. The reduced
inlet area and skid bar design were
causing a significant portion of the feed
to actually bypass much of the exchanger
surface area (Figure 6). The low calculated
coefficient was a result of a significant
bypass exacerbated by a low bundle
pressure drop design (no vertical baffles).

Solution

Abe du Pont is the manager for process

engineering at Natref. Du Pont obtained
a BSc Eng (Chem) at the University of
Pretoria in 1984.
Mark Fernsby is a process engineer
at the Natref oil refinery in Sasolburg,
South Africa. Fernsby has BSc and MSc
degrees in chemical engineering from the
University of Cape Town.
Tony Barletta is a chemical engineer with
Process Consulting Services in Houston,
Texas, USA.
Email: tbarletta@revamps.com
Steve White is a chemical engineer with
Process Consulting Services in Houston,
Texas, USA. Email: swhite@revamps.com

www.eptq.com

Double
segmental
baffle

Full-diameter
baffle

Double
segmental
baffle

Full-diameter
baffle

Plan view
(top)
Figure 7 Bundle modification

Figure 8 New bundle

30
25

Installation
of new
tube bundle

GJ/hr C11068 duty

kcal/h m2 CU

1400
1200

20

1000

15

800

10

600
400

5
0

28

g
Au

200

Heat transfer coefficent U,

Kcal/h. m2. C

The authors would like to acknowledge the

support of the Natref personnel who were
involved in this project, particularly the
CDU operations division.

Figure 6 Original bundle

Duty, GJ/hr

Many solutions were considered

including a new larger exchanger shell
and bundle, and new reboiler return
piping, both of which were expensive
options. Furthermore, these options did
not address the root cause and may not
have worked. Instead, a new exchanger
bundle was designed and installed in
the same shell.
A new bundle was designed with the
intent to improve the flow distribution
to fully utilise the tube surface area
(Figures 7 and 8) and maximise the tube
outside coefficient ho. Vertical double
segmental baffles were installed to
improve the flow distribution. The baffle
design was optimised to ensure shellside fluid was forced through the entire
bundle. Thermosiphon hydraulics were
checked to ensure the higher pressure
drop would not reduce circulation and
cause problems with high percentage
vapourisation. The skid bars were
redesigned and impingement plates
eliminated. In addition, full-diameter
vertical baffles were installed between
the inlet nozzles to help with flow
distribution. After the new bundle was
installed, the service heat-transfer
coefficient improved from 250 kcal/h m2
C or less to 950 kcal/h m2 C (Figure 9).
The debutaniser was now able to process
light crudes while meeting LSR product
RVP specification.

0
06 006 006 006 006
06 006 006 006 006 006 006
06
2
2
2
20
20
2
2
20
2
2
2
2
2
g
p
p
p
p
p
p
p
p
p
p
p
e
e
e
u
e
e
S
S
S
S
S
Se
Se
Se
Se
Se Se
A
6
4
2
9
11
13
18
15
21
23 26
31

Figure 9 Improvement in service heat-transfer coefficient after new bundle installation

PTQ REVAMPS 7

The Ultimate De-bottlenecker

AMISTCOs MistMaster III
Mist Eliminator
In testing, AMISTCO Separation Products
patented MistMaster III Mist Eliminator has
realized a 18% gain in gas flow capacity with no
loss in efficiency. For de-bottlenecking or maybe
reducing the incidence of fouling, the dedicated
gas flow channels on the bottom side of the
MistMaster III Mist Eliminator have proven to
avoid liquid build-up, increase capacity, lower
pressure drop all without a loss in efficiency.

The 18% More

Liquid Gas Flow
NO LOSS IN EFFICIENCY

Mist Eliminator

Available in a wide range of materials, the

MistMaster III Mist Eliminator installs just like a
conventional mist eliminator. The only difference
is the high flow will allow you to reduce
vessel size requirements in new applications
and provide greater throughput in existing processes.
For more details call or visit our website.

www.amistco.com 24hr EMERGENCY SERVICE 1-800-839-6374

FCC gas concentration unit

stripper revamp
Unit charge rate and reactor conversion were increased by the elimination of stripper
column flooding. Chemical-grade propylene specifications were met and C2 excursions
in mixed C2/C3 streams were eliminated
Dave Langdon Total Lindsey Oil Refinery
Scott Golden and Edward Hartman Process Consulting Services Inc

Mixed C4

C3/C4 splitter

C2s/C3s

Debutaniser
feed

C3 = product

C3 splitter

Deethaniser

Debutaniser

Treating

C3 product

Primary
absorber

Gasoline product

Figure 1 Generic process flow scheme of GCU fractionation system (note: the Lindsey refinery GCU operates with two C3 splitters)

he FCC gas concentration unit

(GCU) stripper column at the
Total Lindsey Oil Refinery (LOR)
was revamped in 2005 to eliminate
flooding, consistently maintain C2s
in GCU mixed C3s at 5000 ppmwt
or less (ie, <5000 ppmwt C2s in C3
mixture) and maximise C3 recovery
(Figure 1). Flooding in the stripper
column limited the FCC units (FCCU)
charge rate and reactor conversion
prior to the revamp, requiring a bypass
from the main columns overhead
receiver to the bottom of the stripper
to manage flooding. C2s excursions
in the GCU mixed C3s exceeding
20 000 ppmwt were also noted. Since
LOR produces chemical-grade propylene

www.eptq.com

(C3=) requiring C3= that exceeds 95%/

purity with total C2s and ethylene
specifications of 400 and 50 ppmwt
respectively, these 20 000+ ppmwt C2
excursions were unacceptable, as they
overloaded the C2 stripping capacity of
the two downstream propane/propylene
(PP) splitting units. FCCU conversion
had to be reduced to the stripper
columns hydraulic limit to meet the
C3= product specifications. Following
the revamp, stripper column flooding
was eliminated, total C2s in the GCU
mixed C3s were reduced to an average
of 3000 ppmwt, and the FCCUs reactor
charge rate and conversion has increased
to record levels. The stripper columns
feed rate is 33% higher than before

the revamp, with no further flooding,

and is no longer the FCCUs constraint.

Pre-revamp operation
Prior to the revamp, the stripper column
limited the FCCUs charge and
conversion for several years. During the
previous turnaround, the column
internals were revamped to increase
capacity, but the stripper did not meet
its design objectives. Following the 2000
revamp, the reactors outlet temperature
(ROT) had to be reduced 15F (8C) to
lower the gasoline and LPG yield.
Additionally, a wild naphtha product
bypass line was installed from the main
columns overhead receiver to the base
of the stripper column to reduce the

PTQ REVAMPS 9

Technology developer, owner-operator

real-world experience translates into
real-world success

$
Higher unit
throughput plus
longer run lengths

$
Superior liquid yields
at favorable costs

$
$
Only licensor that designs,
owns and operates
delayed cokers: 50+ years

Leading marketer of all grades

of specialty coke; better understanding of complex feedstocks

Lower emissions for

a cleaner environment

How do you convert heavy crude into high value? Pick the brains with hands-on experience.
As a leading technology developer and owner-operator, were the brains behind the most economical refining processes
on the market. With a history rich in proven technologies that deliver proven competitive advantages. At ConocoPhillips,
we know about real-world results from our own operations. Its why our ThruPlus Delayed Coking Technology is the
industrys leading economical choice for upgrading petroleum residues into high-value liquid products. Its bottom-line
value comes from the best qualified, most experienced delayed coking licensor. Higher liquid yields. Lower coke yields.
Maximized coking capacity. And feedstock flexibility. If youre ready to profit by making the most of your feedstocks, pick
the most experienced brains in the industry. ConocoPhillips. Performing

today. Preparing for tomorrow.

For help solving your delayed coking challenges, e-mail us at technologysolutions@conocophillips.com.

ThruPlus PTQ July07.indd 1

6/5/07 8:42:04 PM

Sponge absorber
Main
fractionator
vapour

Debut. gasoline

Primary
absorber
Comp.
driver

Overhead
receiver

Wet gas

High-pres.
receiver

Interstage
receiver

Stripper

Preheat
Sour
water
stripper

LCO PA

Debut.
feed

FI

By-pass
line

Figure 2 Stripper column bypass

strippers feed rate (Figure 2). Even with
these changes, the stripper column
limited the FCCUs charge rate and
conversion.
While the process flow scheme in
Figure 1 is a generic depiction (eg, only
one C3 splitter is depicted), the Lindsey
facility actually has two PP splitter units
(noted as C3 splitter in Figure 1), each
with a different arrangement. They both
serve the same purpose, but one utilises
three towers: one tower operates as a
separate deethaniser, followed by a twoshell PP splitter, essentially one for
stripping and one for rectification. The
other PP splitter unit achieves the same
results in just one tower, with a C2
stripping section provided at the top of
the tower and the C3= drawn as a
side product.

Absorber/stripper systems
Primary absorber/stripper systems are
the heart of the FCC GCU. These
columns and the high-pressure (HP)
receiver control C3 recovery and C2
rejection. Ideally, the stripper bottoms
products C2 content is controlled at the
maximum required to meet the GCU
mixed C3s total allowable C2s of the
previously noted 5000 ppmwt, because

www.eptq.com

this maximises C3 recovery. Both

columns are prone to flooding, with the
stripper being the most problematic
industry-wide.
Primary
absorbers
operate at extremely high internal
liquid rates, have moderate foaming
tendencies, almost always flood due to
inadequate downcomer top area or
downcomer backup, and often flood at
the intercooler draw or returns because
of inadvertent restrictions. The stripper
column also operates at extremely high
liquid rates, with flooding initiated by
either downcomer choke or backup
flood mechanisms.
Rarely is jet flooding a problem in
either the primary absorber or the
stripper column. Moreover, the stripper
can entrap water because the top
temperature is too cold for all the water
to go overhead, and the bottom
temperature is too hot to allow the
water to leave the bottom. This forces a
water phase in the top and middle
sections of the column, leading to
severe flooding. When the stripper
floods, it is not uncommon to rapidly
fill the HP receiver, because the column
feed rate is higher than the tray capacity.
When the stripper floods, the FCCs
charge rate must be lowered and the

reboilers heat input drastically reduced.

Unfortunately, large reductions in the
stripper heat input can cause major
excursions of residual C2s in the mixed
C3s product exiting the GCU. In this
case, total C2s increased to 20 000
ppmwt or higher.
The stripper experienced severe
flooding following the 2000 revamp. In
an effort to mitigate this, LOR installed
a bypass to route some of the wild
naphtha around the primary absorber/
stripper columns, thereby reducing the
strippers feed rate. The wild naphtha
maximum bypass was about 40 m3/hr
out of a total stripper feed of 280 m3/hr.
Therefore, the strippers feed rate
unloaded by up to 15%. However, as a
liquid proportion of the main liquid to
the primary absorber, the reduction was
25% or more, which reduced the
primary absorbers internal liquid/
vapour (L/V) ratio, thus reducing C3
recovery. Furthermore, the wild naphtha
contained more than 2000 ppmwt C2s
and high H2S, forcing the stripper
column to overstrip its feed to meet the
mixed C3s products C2s specifications.
While this improved the overall
performance of the FCC, the C2s
excursions still occurred.

PTQ REVAMPS 11

Primary absorber/stripper
system revamp
LORs 2005 revamp installed new
internals in the primary absorber/stripper
columns, as well as a riser on the stripper
feed draw nozzle in the HP receiver to
improve oil water separation. Primary
absorber modifications included new
trays with a larger downcomer top area
and modifications to the intercooler
draws to eliminate restrictions. These
changes have allowed the FCCs feed rate
and conversion to be increased without
having to operate the wild naphtha
bypass. Since startup, the unit has been
tested at a maximum charge rate and
conversion without any indications of
either columns flooding. However,
further charge rate increases have pushed
the absorber/stripper system to a cooling
limit. Thus, during summer operation,
the GCUs offgas C3+ content has
increased from 8.0 to 12 mol%. Since this
offgas is fed to a cryogenic gas plant
where most of the C3+ is recovered, the
impact of reduced GCU C3= recovery on
the overall C3= product yield is less
discernible. This said, the GCUs
operating objective is still to try to
minimise C3+ from the sponge absorber,
because C3= recovery efficiency on the
cryogenic unit is only 70% maximum. So
directionally more C3= leaving the
sponge absorber will result in more C3=
loss to fuel, with significant economic
penalty. Given the overall FCC
economics, the C3= losses do not impact
the feed rate/ROT policy despite the
GCUs efforts to drive the losses from
12% back down to 8.0%.
All the stripper columns internals
were replaced. The existing top
downcomer area was too small and the
active decks open area was too low,
resulting in downcomer backup flooding
and downcomer choke flood. The tray
open area was smaller in the top half of
the tower than in the bottom, in an
attempt to match the vapour load profile
and the tray open area. The flooding
profile was one of progressively increasing
liquid/foam heights from the middle of
the column upwards, until there was no
discernible vapour space above the top
four trays. Once the downcomer capacity
was reached, the column pressure drop
increased dramatically because the
accumulation of liquid above the tray
deck eventually filled the top of the
column with liquid. This caused a
massive liquid carryover into the HP
receiver, resulting in a high liquid level.
Stripper column flooding is common
due to extremely high liquid rates,
difficulties separating the vapour carried
into the downcomer and the presence of
a water phase. Liquid flowing into the
downcomers entrains vapour with it,
the amount depending on a number of
variables including tray design and
liquid/vapour physical properties. The

12 PTQ REVAMPS

Condenser outlet
three-phase

Gas to primary
absorber

High-pressure separator
Riser
Oil

Water
IC

Water

Ideally only
soluble water

Stripper feed
Sour water
stripper

Figure 3 HP receiver oil and water separation

top downcomer area must be large
enough to allow the entrained vapour to
flow out at the same time as the liquid
and vapour flow in. In an atmospheric
crude column, it is possible to have
liquid loadings based on the top area of
the downcomer higher than 200 gpm/
ft2, whereas stripper downcomer
loadings should not exceed 130 gpm/ft2.
Higher loadings choke the inlet,
preventing additional flow into the
downcomer and causing liquid to
accumulate on the tray deck.
Downcomers also flood via a backup
mechanism. Once the froth height in
the downcomer reaches tray spacing, the
column will flood. Since vapour is
entrained with the liquid entering the
downcomer, there is a density difference
from the bottom of the downcomer to
the top. Clear (or nearly clear) liquid is
present in the bottom and an aerated
froth in the top. In an atmospheric
crude column, it may be possible to fill
the downcomers by 5060%, assuming a
clear liquid level, whereas in the GCU
stripper column the downcomer backup,
assuming clear liquid, should not exceed
35% of tray spacing. The froth sitting on
top of the clear liquid must be
maintained below the tray level to
prevent downcomer backup flood.
The absorber/stripper column trays
must be designed with sufficient
downcomer top area to prevent choke
flood, and the downcomer backup must
be kept at 35% or less of tray spacing
based on clear liquid. Downcomer
backup is a function of the weir height,
downcomer clearance, weir liquid crest
and tray open area. The designer
needs to adjust these to maintain backup
at or below 35%; otherwise, the trays
will flood.

Stripper water phase

The water phase inside the stripper will
flood the column prematurely. The two
causes are free water in the feed from the

HP receiver, or low-temperature feed

forcing a water phase to form even
though there is only soluble water in
the feed. Poor oil/water separation in
the HP receiver results in free water in
the feed. The HP receiver may be too
small or there may be problems
controlling the boot interface. Prior to
the 2005 revamp, LORs HP receivers
hydrocarbon was withdrawn from the
bottom of the receiver because the draw
nozzle did not have a riser. Hence,
when the water level reached the
hydrocarbon draw, the stripper feed
contained free water. During the 2005
turnaround, a riser was added to the HP
receivers hydrocarbon draw nozzle to
prevent water from inadvertently being
withdrawn (Figure 3).
A water phase forms inside the
column when the feed is too cold.
When the stripper feed is cold, the
column needs a water draw in the top
section to remove the water phase.
LORs stripper feed is heated prior to
entering the column. Therefore, it will
not form a water phase. However, to
ensure any water in the feed could be
withdrawn, a properly designed water
draw was installed.

Revamp results
Since modifying the primary absorber/
stripper columns internals, and making
other changes in 2005, the FCCU has
achieved record throughput and
conversion. The stripper is no longer a
constraint on capacity or fractionation
efficiency. The wild naphtha bypass has
been operated closed for several months
without any flooding in the primary
absorber/stripper columns. Currently, it
is open to relieve a hydraulic constraint
from the HP receivers stripper feed
pumps to the stripper column. As a
consequence of this hydraulic limit, it
has been impossible to flood the
stripper column.
Prior to the 2005 revamp, the stripper

www.eptq.com

columns feed rate was maintained at

250 m3/hr with a maximum of 300 m3/
hr by adjusting the wild naphtha bypass.
Currently, the feed rate is approximately
400 m3/hr. Prior to the turnaround, the
residual C2s in mixed C3s were in the
region of 5000 ppmwt, with regular
intolerable excursions up to 20 000
ppmwt or more. Since the revamp, the
C2s have been reduced to an average of
3000 ppmwt with less variation. Since
the mixed C3s product is further stripped
in a deethaniser in front of one of the PP
splitters, and in the deethanising section
within the single PP splitter tower (and
5000 ppmwt C2s is acceptable), there is
probably some overstripping increasing
the hydraulic load on the stripper feed
system. But because of the inherent
difficulties in controlling GCU stripper
column bottoms product C2 content
precisely, and because of problems with
the overhead gas meter from the
column, further optimisation will have
to wait until the gas meter can be put
in service.
Following startup, there was some
evidence of water accumulation on the
stripper columns water draw tray, but
more recently the level control valve has
been in the closed position. LOR is
constructing a sight glass to be able to
crosscheck the oil/water interface
controller on the tray. Since a riser was
installed on the HP receivers stripper

Prior to the turnaround,

the residual C2s in mixed
C3s were 5000 ppmwt,
with regular excursions
up to 20 000 ppmwt or
more. Since the revamp,
the C2s have been
reduced to an average
of 3000 ppmwt with less
variation
feed draw nozzle, this potentially
prevents entrainment of free water into
the stripper. Although an improved
water draw tray was specified for the
stripper, there has been no clear
evidence of water collection and
decantation in the column.
The primary absorber has operated
without capacity constraints, with the
stripper debottlenecking permitting a
higher flow of debutanised gasoline to
the primary absorber. The C3+ content
of the sponge absorber off-gas is typically
maintained at about 8.0 mol%, which is
acceptable considering the cryogenic
unit recovers most of this material.
During the summer months, the

intercoolers and HP receivers condensers

operate at maximum heat removal, with
the primary absorbers vapour line
temperature increasing from its normal
target of about 104F (40C) to higher
than 113F (45C), which increases the
C3+ content to about 12 mol%. The tray
hydraulic capacity has been tested by
closing the wild naphtha bypass and
increasing the recycle gasoline to
maximum without any indication of
flooding. Ultimately, the intercoolers
and HP receivers heat-removal limits, as
well as the HP receiver-to-stripper
hydraulic bottleneck, restrict the
primary absorber and stripper columns
feed rate rather than the previous
column flooding.

Dave Langdon is a chemical engineer with

Total Lindsey oil refinery, N Killingholme,
Immingham, UK. Langdon graduated from
Loughborough University of Technology,
UK, with a BEng degree in chemical
engineering.
Scott Golden is a chemical engineer with
Process Consulting Services in Houston,
Texas, USA. Email: sgolden@revamps.com
Edward Hartman is a process engineer
with Process Consulting Services in
Houston, Texas, USA.
Email: ehartman@revamps.com

Give the New Crew

a Knowledge Boost
Refining Courses now filling for
2008 schedule
RF-31 Refining Technology Overview
RF-61 Refinery Gas Treating, Sour Water,
Sulfur and Tail Gas
RF-62 Refinery Process Simulation

Know-how for the New Crew.

Exclusive provider of PetroSkills facilities training.

www.eptq.com

John M. Campbell celebrates 40 years of excellence

in petroleum training. Launching with the now famous
Campbell Gas Course in 1968, Campbell was asked
to manage the PetroSkills Facilities training delivery in
2003. PetroSkills is the leading competency and skill
assessment system in the Oil and Gas Industry.
The worlds major Petro companies, (Shell, Halliburton,
Saudi Aramco, Oxy, ConocoPhillips, Chevron,
Marathon, RepsolYPF) have come together to assure
quality, course content and worldwide availability.

www. petroskills.com www.jmcampbell.com

PTQ REVAMPS 13

1
2

Better sealing saves money

Rings and packing are crucial elements in ensuring the reliability and availability of reciprocating
compressors. The key to the efficient operation of a compressor is a combination of component
materials and design which is ideally suited to the specific compressor and its operating conditions.

www.hoerbiger.com

In this respect, HOERBIGER demonstrates that it thoroughly deserves its place at the forefront
of engineered sealing solutions.
1

Consider, for example, BOT pressure balancing ring pairs. These reduce the

friction between the rings and the piston rod and thereby significantly reduce the wear
of both components. This ensures the longer service life of the wear parts, and thus
fewer shutdowns.
2

The result of HOERBIGERs OT Oil wiper packing is almost 100% oil wiping

efficiency and the prevention of oil leaks from reciprocating compressors.

Considerable financial savings may be made due to subsequent reduction in oil
purchase and disposal and recycling costs.
3

Thermosleeve packing facilitates the transfer of heat directly into

the cooling channel of the cylinder, so that a separate cooling system for
pressure packing is no longer required.
4 HOERBIGERs piston, rider and packing rings, made from reliable HY materials
(e. g. HY50, the material for bone-dry applications), offer a high level of resistance
to wear, even in critical applications.
These are only four of many reasons why our customers say:
...better with HOERBIGER.

Compressor Valves - Rings & Packings - Control & Monitoring Systems - Compressor Services - Compressor Conversion

B tt

li

07 02 i dd 1

20 02 2007 17 59 14

Lube vacuum
column revamp
The basis of a lube vacuum column revamp and its results are discussed. The primary
goals were to increase lube product fractionation, minimise HVGO product
contaminants and improve vacuum bottoms specifications for asphalt production
Kevin Basham Marathon Petroleum Company LLC
Edward Hartman Process Consulting Services Inc

he lube column at Marathon

Petroleum Companys (MPC)
Cattlesburg,
Kentucky,
USA,
refinery was revamped in 2006 (Figure 1)
to improve lube product fractionation,
reduce the heavy vacuum gas oil
(HVGO) products micro-carbon residue
(MCR) and metals, and improve the
vacuum bottoms specifications for
the production of asphalt. Prior to
the revamp, the HVGO product was
black and cylinder stock yield was
excessive. Cylinder stock was fed to
another vacuum unit so that the lube
column bottom product could meet
the asphalt product specifications, as
well as recover a portion of the HVGO
boiling range material in the cylinder
stock. However, reprocessing this stream
consumed some of the other units
capacity and increased its heater firing.
All the projects justification benefits
were met following startup. Moreover,
a lower column operating pressure and
improved stripping efficiency led to
a 4.0 Mbpd higher crude charge rate
due to the lower heaters cracked gas
production freeing up some compressor
capacity to process more crude.

Off-gas

Crude off-gas

Ejector
system
Oil
Sour water
LVGO

Atmospheric
crude
HGO
product

www.eptq.com

Lube vacuum
SS #1
SS #2
HVGO
Cylinder stock

Preflash
crude
Steam

Stripping
steam
Vacuum bottoms
Fuel
gas

Process flow scheme

The previously mentioned Figure 1
shows the simplified process flow
scheme for the atmospheric and vacuum
column prior to the revamp. Off-gas
from the atmospheric crude and lube
vacuum columns was handled with a
common compressor. Atmospheric
column heavy gas oil (HGO) product
was routed to the vacuum column to
recover some of the lube-quality
material. The vacuum column produced
light vacuum gas oil (LVGO), side
stream (SS) #1, side stream (SS) #2,
HVGO, cylinder stock and asphalt
products. It was necessary to yield
cylinder stock to meet the asphalt
specifications on the bottom product.
LVGO and HVGO were routed to cat
feed hydrotreating. SS#1 and SS#2 were
lube-quality base stocks targeted for
further processing.

Ejector off-gas

Figure 1 Lube vacuum unit

The vacuum unit is a wet design with
coil and residue stripping steam, plus a
precondenser prior to the first-stage
ejector. The top column operating
pressure on a wet column is set by the
ejector system load, which changes with
the seasonal cooling water temperature.
During winter, the first-stage ejector
load is lower due to the reduced cooling
water temperature; hence, the columns
top pressure is as low as 35 mmHg. This
pressure increases to 55 mmHg in the
summer when the cooling water
temperature is higher. The vacuum
column internals consisted of three
packed beds and 12 trays to remove heat
and fractionate the feed into five side-

cut products (Figure 2). The tray pressure

drop was approximately 3.5 mmHg per
tray, with the packing contributing only
a small amount, which resulted in a
flash zone pressure of 95105 mmHg,
depending on the ambient temperature.
The HVGO product was black from
entrained vacuum residue. Furthermore,
the trays were prone to damage
and
leaks,
resulting
in
poor
fractionation, excessive cylinder stock
yield and difficulty in meeting asphalt
specifications.
Typical unit charge was a blend of
Middle Eastern crudes and was limited
by off-gas compressor capacity. During
the summer, the atmospheric columns

PTQ REVAMPS 15

Ejector

LVGO
product

CW

SS #1

HGO

SS #2
HVGO
Feed

Cylinder stock

Stripping
steam
Vacuum bottoms

Figure 2 Lube column before revamp

Minimising HVGO
product contaminants

Higher heater
outlet
temperature
Very low
strip-out
S-1
S-2

Shed S-3
trays
S-4
Stripping
steam

Vacuum bottoms

Figure 3 Residue stripping section before revamp

overhead receiver temperature increased,
producing more off-gas. At the same
time, the vacuum ejectors off-gas rate
was higher because the heaters outlet
temperature was at maximum. The
crude rate had to be reduced once the
off-gas compressor reached its maximum
capacity.
In 2005, a study was completed to
determine the cost/benefit of revamping
the unit. Major economic incentives were
improving
the
lube
products
fractionation, minimising the HVGO
product MCR and metals contaminants,
producing on-specification asphalt and
improving internals reliability.

16 PTQ REVAMPS

drop and inherently poor efficiency in

lube column service. In 1984, the first
large-diameter lube vacuum column
was revamped from bubblecap trays to
structured packing, with several others
modified since then. Some were
successful, while others were not due to
poor-quality liquid distributors.
Trays have inherently low efficiency
in lube columns because the liquid rate
is low and the tray weir length is large
(because of large diameters). Conversely,
structured packings inherent efficiency
is good at a low liquid rate, assuming a
high-quality liquid distributor is used. A
structured packing beds efficiency is
largely controlled by the liquid
distributors performance. Since column
diameters are large and liquid rates are
generally low (~1 gpm/ft2 of column
area), distributing the liquid uniformly
is a challenge.
MPCs lube vacuum column flash
zone pressure was high because it had
12 trays above the flash zone. Trays
produce a high pressure drop per
theoretical stage because each tray
generates about 3.5 mmHg pressure
drop and three trays are needed to
achieve a theoretical stage. Thus, each
theoretical stage creates approximately
10 mmHg pressure drop. Conversely,
structured packing produces only 1.5
mmHg per theoretical stage.

Lube products fractionation

Fractionation between lube products
depends on reflux and the number of
theoretical stages. Increasing either one
improves fractionation. This requires
high-efficiency mass-transfer internals
and minimum column flash zone
pressure. Prior to the revamp, the
column pressure drop was high,
stripping efficiency was low, and
fractionation between lube cuts and
HVGO was poor. Before the mid-1980s,
most lube vacuum columns were
designed with bubblecap trays and
occasionally valve trays. Yet, trays have
the disadvantages of high pressure

HVGO product contaminants consist

of volatile MCR and metals in the
product boiling range as well as
entrained vacuum residue containing
high amounts of MCR and metals.
There will always be some volatile
contaminants present, irrespective of
the column design, but the amount of
volatile contaminants depends on the
efficiency of the wash zone and residue
stripping. Surprisingly, many vacuum
columns produce black HVGO product
because of poorly designed column
flash zone, wash zone and stripping
section internals. In this case study,
because the unit processed low metal
crudes and operated at low cutpoints,
eliminating entrainment would reduce
the HVGO products metals and MCR
to low levels.
Vacuum residue entrainment in the
flash zone depends on transfer line
critical-flow expansions, the flash zone
vapour horn and wash section internals.
Poorly designed transfer lines with high
pressure drop critical-flow expansions at
the column inlet nozzle generate fine
mists that are difficult to remove.
Improperly designed flash zone vapour
horns do not separate the feed stream
vapour and liquid or properly distribute
the vapour to the wash section. Poor
vapour distribution results in high
localised vapour velocity in parts of the

www.eptq.com

wash section, causing the HVGO

product to be black from entrainment.
Yet, entrainment can be nearly
eliminated through prudent column
internal designs, as long as the column
wash section capacity factor does not
exceed 0.40.45 ft/sec.

Off-gas

Crude off-gas

Residue stripping
The efficiency of the vacuum column
stripping section influences the HVGO
yield, asphalt quality and HVGO
products metals. Yet, it is often
overlooked as an important design
variable. Maximising stripping efficiency
will raise the HVGO products TBP
cutpoint by 2060F, reduce metals in
the HVGO product by 50% or more and
allow for the production of highervalued asphalt grades from the same
crude. Greater efficiency also reduces
the vacuum heater outlet temperature
by up to 30F in some instances.
Moreover, residue stripping vapourises a
lower boiling range hydrocarbon than
the heater produces for the same
amount of vapourisation. Therefore, it
reduces the HVGO products MCR and
nickel and vanadium content.
Residue stripping uses steam to reduce
the oil partial pressure on the trays.
Even a well-designed tray has only
25% efficiency, with a typical tray
achieving just 510% efficiency. Since
the stripping trays vapourise lower
boiling range material than the heater,
the metals content is lower than the
same volume of material produced in
the heater. In addition, improved
efficiency vapourises more flash zone
liquid, allowing the heater outlet
temperature to be reduced. Since this
unit produces asphalt, it is the asphalt
specifications that determine the HVGO
products cutpoint rather than a specific
target. Another benefit is that the
asphalt product viscosity is easier to
meet because the light material that
must be removed to meet the viscosity
specifications is stripped out. Before the
revamp, the stripping section was
designed with four baffle trays, so it had
very little efficiency, which resulted in a
low strip-out and made it difficult
to meet the asphalt specifications
(Figure 3).

Revamp process flow scheme

The revamp flow scheme is shown in
Figure 4. All the trays above the flash
zone were replaced with packing, and
other minor modifications were made to
the existing packed beds, plus a much
more efficient stripping section was
installed. New packed column internals
reduced the flash zone pressure by about
30 mmHg. A lower flash zone pressure
allowed for a lower heater outlet
temperature, which reduced cracked gas
production (Figure 5). Improved
stripping efficiency further reduced the

www.eptq.com

Ejector off-gas
Ejector
system
Oil
Sour water

Atmospheric
crude

LVGO

Lube vacuum
HGO
product

SS #1
SS #2
HVGO

Preflash
crude

Cylinder stock

Steam

Stripping
steam
Vacuum bottoms
Fuel
gas

Figure 4 Revamp process flow scheme

Lower
Increased heater
vacuum cracked
distillate
gas

Before revamp

After revamp

730
105

710

Flash zone
conditions

75

Feed

Stripping
steam
Temperature, F
Pressure, mmHg

Vacuum bottoms

Figure 5 Flash zone temperatures and pressures

heater outlet temperature, lowering
cracked gas production again.
The crude/vacuum unit capacity is
constrained by the off-gas compressors
capacity. By reducing the amount of
ejector off-gas, more atmospheric
column overhead receiver off-gas can be
handled, allowing for an increased crude
charge rate (Figure 6).

Column design
Prior to the revamp, the column was
designed with a combination of grid,
random packing, valve and bubble cap
trays, and baffle trays. The new internals
would:
Reduce the column pressure drop
Provide the required fractionation
for lube oil production

PTQ REVAMPS 17

Reduced
heater outlet
temperature

Off-gas to
treating

Off-gas
compressor

S-1
S-2
S-3
S-4
S-5

Stripping
steam

Crude off-gas
(higher rate)
Vacuum off-gas
(reduced rate)

Vacuum bottoms

Figure 6 Off-gas compressor

Figure 8 Stripping section after revamp

Ejector

LVGO
product

CW

HGO

SS #1
SS #2
HVGO
Wash oil
Feed

Higher
strip-out

Cylinder stock

Stripping
steam
Vacuum bottoms

Stripping section efficiency was

dramatically improved by changing
from baffle to sieve trays, increasing the
number of trays from four to five,
designing each tray with an optimised
hole area for higher tray efficiency and
installing a new collector to feed the top
tray (Figure 8).

Side-strippers
The side-strippers tray design was
modified to increase tray efficiency,
allowing for more strip-out of the
product distillation front-end. The
changes were low cost, with the tray
active panels replaced with an optimised
open area and a combination of light
and heavy valves.

Revamp results
The primary goals were to improve lube
product fractionation, minimise HVGO
product contaminants, meet vacuum
bottoms properties for asphalt production
and improve internals mechanical
reliability. All objectives were met. The
HVGO product MCR and metals have
been reduced. The wax colour from SS #2
is now clear because colour bodies from
residue
entrainment
have
been
eliminated. In addition, the crude charge
rate has been increased by 4.0 Mbpd as a
result of reducing the vacuum heaters
cracked gas production.

Figure 7 Lube column after revamp

Improve HVGO quality by eliminating
vacuum residue entrainment
Maximise stripping section efficiency
Provide adequate wash oil to keep
the wash bed wetted and prevent
coking while minimising cylinder stock
yield
Improve internals reliability.
The existing 12 valve and bubblecap
trays were replaced with structured
packing beds and associated internals. A
new vapour horn was installed to
minimise entrainment and improve
vapour distribution to the wash section.
The four existing baffle trays used for

18 PTQ REVAMPS

stripping were replaced with five

sieve trays.
The fractionation section used
structured packing and high-quality
narrow trough liquid distributors to
maximise efficiency within the available
height. The wash section was installed
in an internal shroud to allow for
maximum lube product fractionation.
Existing tray support rings were reused
to reduce the installation time. The new
internals have improved mechanical
design to maintain integrity during
normal operating conditions and upsets
(Figure 7).

The authors would like to thank all who

participated in the successful completion of
this revamp.

Kevin Basham is a technical services

co-ordinator at the Marathon Petroleum
Company LLC Cattlesburg, Kentucky, USA,
refinery. He has 15 years experience in
the optimisation and troubleshooting of
crude distillation units.
Edward Hartman is a process engineer
with Process Consulting Services Inc
in Houston, Texas, USA. His primary
responsibilities include process design
packages (PDPs) and equipment design
for refinery units.

www.eptq.com

What if your assets had appreciating value instead of depreciating value?

With IBM Maximo Asset Manager, you not only see the performance of all your operational and IT assets across your enterprise,
but also the untapped potential within them. By more efficiently managing the convergence of these assets, you can optimize
asset utilization and increase real-time visibility of production performance to minimize downtime and meet regulatory
compliance. To learn how our oil and gas asset and service management solutions can help you, call 800-326-5765,
or download our white paper, Convergence of Asset and Service Management, at www.maximo.com/rv.

COUNTED

CONTROLLED

MAXIMIZED

IBM, the IBM logo, and Maximo are registered trademarks or trademarks of International Business Machines Corporation in the United States, other countries, or both. Other company, product, and service names may
be trademarks or service marks of others. Copyright IBM Corporation 2007. ALL RIGHTS RESERVED.

EgVXi^XVahdaji^dchidhadl\adWValVgb^c\
L]^ahii]ZldgaYY^hXjhhZhi]ZXa^bViZX]Vc\ZVcY^ihXdchZfjZcXZh!
A^cYZ]VhWZZcYZkZade^c\XdcXZeihidh^\c^[^XVciangZYjXZZcZg\n
Xdchjbei^dcVcY\gZZc]djhZ\VhZb^hh^dc^cZm^hi^c\Zi]naZcZeaVcih#
DcZd[A^cYZh^ccdkVi^kZhdaji^dch^hidgZeaVXZi]Z[jgcVXZhZXi^dcd[V
XgVX`ZgYjg^c\VijgcVgdjcYVcYi]jh^chiVaahiViZ"d["i]Z"VgiiZX]cdad\n
^cVcZm^hi^c\eaVcil^i]djiVcnegdYjXi^dcadhh#A^cYZ^hXjggZcian^beaZ"
bZci^c\i]^hVeegdVX]^cV:jgdeZVcXgVX`Zg#
I]^hcZlA^cYZ[jgcVXZiZX]cdad\ngZYjXZh\gZZc]djhZ\VhZb^hh^dch
l]^aZ^cXgZVh^c\n^ZaYhVcYhjWhiVci^Vaan^begdk^c\gZa^VW^a^inVcYZXd"
cdb^XZ[[^X^ZcXn#
Di]Zg^ccdkVi^kZXdcXZeihXjhidb^oZYid^cY^k^YjVaeaVcigZfj^gZbZcih
VgZVkV^aVWaZ[gdbA^cYZ!i]ZaZVY^c\Zi]naZcZiZX]cdad\nXdcigVXidg#

9Zh^\c^c\EgdXZhhZh8dchigjXi^c\EaVcih#
A^cYZ6<
@]bXY9b[]bYYf]b[8]j]g]cb 8f"!7Uf`!jcb!@]bXY!GhfUggY*!%( ,&$(-Di``UW\ ;YfaUbm
D\cbY(-",-"+(()!$ :Ul(-",-"+(()!(-$, 9!AU]`.]bZc4`]bXY!`Y"Wca kkk"`]bXY"Wca

CO?A?INDD



Reducing a plants
turnaround schedule
The removal of asphaltenes and residual hydrocarbons from vessels during the
cool-down and steam-out phases can eliminate the source of VOC off-gassing and
reduce a plants turnaround schedule by several days
Mike Lippold
BJ Process & Pipeline Services

he current way to clean process

units is to cool them with a lightcycle oil (LCO) flush and then
steam them out to remove any hazardous
vapours before personnel are permitted
to enter inside. Chemical applications
and/or mechanical methods such as
hydroblasting are then used for cleaning
after the units have been blinded off,
locked out and tagged out. Depending
upon the degree of fouling, the process
takes three to ten days to complete.
Recent advances using soaps and/or
terpene solvents have shortened the
steam-out phase, but do not leave the
unit clean and ready for inspection or
hot work. Additional extensive cleaning
is still required for these to take place.
The proprietary BJ SmartTurn
chemical process allows most cleaning
to take place during the cool-down and
steam-out phases of the shutdown. The
process can be formulated to remove
heavy asphaltenic solids, heavy oils and
greases, pyrophoric iron, hydrogen
sulphide (H2S), benzene and other
volatile organic compounds. By moving
the cleaning portion of the turnaround
into the cool-down and steam-out
phases, many days can be shaved from a
plants turnaround schedule, potentially
saving millions of pounds/dollars.

Three phases of cleaning

Degassing is the removal of hazardous
gases from the vapour spaces of the
system or operating unit (Figure 1).
Steaming is the most common method
used in the industry to remove H2S,
benzene and other volatile organic
compounds
(VOC)
that
leave
hydrocarbon levels above the lower
explosive limit (LEL) in the vessel
vapour space.
This phase leaves the vapour space
clear for personnel entry and solids
removal. However, the process does not
remove oils, pyrophoric iron or heavy
organic matter. This residual material
continues to off-gas, which can cause
benzene levels and VOC to rise and
produce hydrocarbon levels above the
LEL, necessitating additional steaming.

www.eptq.com

Decontamination

Degassing

Removal of hazardous
materials, vapours,
some liquids

Removal of vapours
LEL, benzene, H2S
Ready for personnel
entry for solids
removal

Ready for personnel

entry for final cleaning
Not ready for hot
work/inspection

Typically equipment
remains with solids/
liquid present
Not ready for hot
work/inspection

Chemical cleaning
Removal of vapours, liquids, solids
including hazardous wastes
Ready for hot work/inspection

Figure 1 Three phases of cleaning

In many cases, the unit is not ready for
inspection or hot work either.
Decontamination is the removal of
some liquids (primarily light oils) from
the units surfaces in addition to
removing hazardous gases from the
vapour space. The current industry
practice is to add surfactant packages or
terpene/surfactant packages into the
degassing steam, but this process may
leave behind pyrophoric iron, tars,
greases, paraffinic and heavy asphaltenic
matter. This material must then be
removed later in the turnaround by
hydroblasting or chemical cleaning.
Most of the solid material is considered
RCRA (Resource Conservation and
Recovery Act) hazardous waste. The unit
may then be safe for entry, but is
unlikely to be ready for inspection or
hot work.
Chemical cleaning removes the
liquids (light and heavy oil), tars,
organic solids (paraffinic and heavy
asphaltenic material) and pyrophoric
iron. Following chemical cleaning, the
unit is typically safe for entry and ready
for inspection and hot work.
Current industry practice involves
multi-stage cleaning using hazardous

chemical products after the unit is shut

down and individual vessels blinded off.
The process requires extensive temporary
piping and special chemical cleaning
pumps. It generally creates RCRA
hazardous waste that often cannot be
readily processed by the plant
wastewater-treatment system.

Chemical cleaning process

Current practice involves extensive
hydroblasting or chemical cleaning
services following the shutdown/
blinding process. This work can add
many days to the critical path of the
turnaround. The SmartTurn process
moves the entire chemical cleaning
process to the de-inventory/shutdown
phase of the turnaround. The cleaning is
not performed by a single chemical
package, but by a series of compatible
products that remove the asphaltenes,
benzene, VOC, H2S and pyrophoric iron.
This allows the treatment to be
customised to address problems in
individual sections of the unit. When
this process is complete, the units are
normally ready for inspection and hot
work without adding to the critical
paths duration.

PTQ REVAMPS 21

eptq.com/
QandA
Do you have a
technical query about
a Revamp, or any
other aspect of
Refining, Gas or

Figure 2 BJ Parasol 370 asphaltene dispersant breaks up heavy asphaltenic material

Petrochemical
Processing?
Liquid
level

Liquid
level

Liquid
level

Go to
Product A

www.eptq.com/qanda
and ask it

Simple as that

www.eptq.com/qanda
The Online Q&A
22 PTQ REVAMPS

Product B

Pentol-S3

Figure 3 Comparison of surface cleaning ability of various products

Removal of asphaltenic
material
During the shutdown process, the unit
is normally de-inventoried and cooled
by displacing the crude oil with a cutter
stock such as diesel oil, LCO or gas oil.
For the sake of this explanation, cutter
stock will be referred to as LCO.
The primary functions of the LCO
are to remove the heavy residual oil
from the system and cool the metal
surfaces before the steaming process.
While the LCO effectively cuts the
residual oil, it has limited ability to
dissolve the heavy asphaltenic material
present in the tower bottoms, heat
exchangers and furnaces of the unit.
This material varies in consistency from
a tar-like to a rock-hard solid. The
solid material is generally high in
asphaltenes and so hard it is often
mistaken for coke.
The terpene solvents and surfactant

packages frequently used for degassing

or decontamination do not readily
dissolve these solid materials. They are
typically removed by hydroblasting or
some other mechanical means after
shutdown and blinding.
In the SmartTurn cleaning process,
after the residual oil has been displaced
by the LCO, circulation is established
through the zones where heavy
asphaltenic has been historically found.
This is normally the tower bottoms,
furnaces and heavy oil heat exchangers.
Once circulation has been established,
asphaltene dispersant is typically
metered into the circulating LCO
(Figure 2).
The asphaltene dispersant is a solventbased material designed to break the
binders holding the heavy asphaltenes
together. The product contains a
suspending agent that helps prevent the
asphaltenes from precipitating out of

www.eptq.com

the LCO blend. The circulation

continues until testing indicates the
process is complete.
At this point, the LCO/dispersant
blend is drained from the system to the
slop oil tank or some other storage. The
suspending agents properties are
generally long lasting and typically
prevent
the
asphaltenes
from
precipitating in the slop oil tank or
coming out of suspension when
processed in the crude unit.
The current industry practice is to
hydroblast the asphaltenic material
from the tower bottoms, heat exchangers
and piping. This process can take three
to ten days or longer once the unit has
been shut down and blinded off. This
process typically generates large
quantities of water waste and many tons
of solid waste. Inspection and hot work
cannot begin until the hydroblasting
contractor is out of the unit.
The SmartTurn process can eliminate
the need for much of this hydroblasting
and the disposal of tons of RCRA
hazardous solid waste from tower
bottoms and heat exchangers. The LCO/
dispersant blend and the dissolved
solids are added to the crude oil feed of
the crude unit.

To flare

To flare

Nitrogen
regulator
Nitrogen
regulator

Reflux
pump

Surface cleaning
After the system is free of heavy
asphaltenic material, a heavy oil surface
often remains. This oil is a problematic
source of benzene and other VOCs that
prolong the degassing process. A good
practice in many cases is to cascade
water from the top of the towers to float
residual oil from the trays and the
circulating loops. After the oil has been
decanted from the unit, it is easy to
establish circulation in the various zones
and meter in BJ Pentol-S3 agent and an
added surfactant, PSol-542.
The Pentol S3 degasser has been
designed as a hard-surface cleaner. The
surfactants penetrate the oil and
emulsify it, thus removing the oil,
benzene and VOC from the system.
Figure
3
demonstrates
the
effectiveness of the degassing agent
versus surfactant- and terpene-based
cleaners.1 For this test, carbon steel
coupons were coated with crude oil and
baked in an oven. The coupons
were then placed in a constant
temperature bath. Product concentration
contact time and temperature were
the same for all three products. Note
that Product A, a popular surfactant
package degasser, has had little effect on
the oily surface. Product B, a terpenebased solvent degassing package, was
more effective than the surfactant
package, but a significant amount of
oil remains on the surface. In this test,
the Pentol-S3 degasser effectively
removed the entire oil residue from the
coupon surface.

www.eptq.com

Steam line
Water
&
pentol

Reflux
pump

Steam line

Pentol

Figure 4 A typical steam-out process used to clean and degas a tower with BJ Pentol-S3

Steam out and degassing

Steaming has been used for years to clear
the hazardous vapours from units
following shutdown. Due to the high
concentration of liquids and asphaltenic
material left in the unit, the steam-out
traditionally took many days to
complete. In many cases, when the
vessels were opened up and work began,
frequent halts in the work were required
to again steam out the system due to
high benzene or organic vapours, with
hydrocarbon levels above the LEL.
The need to repeatedly steam out the
system was typically due to residual
hydrocarbon liquids and organic solids
left in the unit. Once the steam-out was
complete, the hydrocarbons, in many
cases, started to off-gas, rising back
above the LEL and then forcing the

evacuation of personnel from the unit.

Recently, surfactant packages and
terpene solvent packages have been
introduced to reduce the time required
to complete the steam-out phase of a
shutdown. In the SmartTurn process,
during the steam out, Pentol-S3 degasser
is injected into the steam and designed
to remove the oily residue from the
surfaces that were not exposed to the
wash water solutions (Figure 4). As the
steam moves through the system, it
condenses on the trays, walls, packing
and demister pads, and the condensate/
degassing agent solution cleans the
inside of the unit. This condensate flows
down the walls to the condensate drains
and is generally processed in the plants
wastewater-treatment plant.
In addition to cleaning the hard

PTQ REVAMPS 23

Benzene absorption
Sample
Benzene added, mg/l
1000
5000
10 000
50 000
100 000

Pentol-S3
mg/l
990
4981
9944
49 816
74 709

Product C
mg/l
650
1717
1688
1761
1636

Product B
mg/l
565
1590
1615
1511
1542

A 3% aqueous test solution of various products after varying amounts of benzene have been added.
Analysis by Gas Chromatography1

Table 1
surfaces of the unit, the degassing agent
can be used to remove the hazardous
vapours from the vapour space. The
Pentol-S3 agents encapsulation process
ties up the hazardous components from
the liquid and vapour phases, resulting
in a removal process that has been
shown to be more effective than other
methods. The VOCs typically remain
tied up in the condensate solution and
are not released from the solution until
the product biodegrades in the wastetreatment system.
Table 1 illustrates how much
additional benzene is removed when
using the Pentol-S3 agent rather than a
surfactant during degassing. It is
interesting to note that neither the
terpene package (product C) nor the
surfactant package (product B) held
significant amounts of benzene in
solution. In fact, the solubility of
benzene in water alone is 1800 mg/l.2
It is generally the oily residue coupled
with the heavy asphaltenic material left
in the unit that releases the benzene and
other VOCs, preventing entry into the
unit. The process has been shown to
effectively remove the source of the
VOCs and benzene so that levels fall
below the LEL. When the sources of the
VOCs are gone, the systems atmosphere
will be clear and safe for personnel to
enter and perform hot work without
additional steaming. Steaming time may
be reduced by up to 50%.

Hydrogen sulphide
H2S is a common problem in the refining
and petrochemical industries, especially
when dealing with sour crudes and
gases. In a typical degassing or cleaning
scenario, the Pentol-S3 agent can
remove up to 10 ppm of H2S in the
vessel vapour space. For higher
concentrations, BJ HS-5 is generally
used. HS-5 is a water-based sulphide
suppression agent. It is designed to tie
up the H2S and hold the H2S, even with
pH fluctuations.
The HS-5 agent is compatible with
the other water-based chemical cleaning
products used in the SmartTurn process,
so it can be used during the cleaning
and degassing process in conjunction

24 PTQ REVAMPS

with the other process products. This

compatibility may save time during a
busy turnaround because an additional
treatment phase is generally not needed
to complete the removal of high
concentrations of H2S.

Pyrophoric iron removal

Pyrophoric iron is frequently found in
the refining and petrochemical industries.
Iron sulphide is common in the upper
portions of towers and overheads in
crude units as well as in equipment
where H2S is present. When pyrophoric
iron sulphide is exposed to air, the
reaction is highly exothermic and, if
hydrocarbons are present in or near the
scale, it presents a risk of fire or
an explosion.
As part of the cleaning process, the
proprietary Ferrotrol 600 iron-control
agent is added to the steam during the
degassing phase of the treatment. The
product is water-based and, like HS-5, it
is compatible with the other water-based
SmartTurn process products. Therefore,
it may be used during the degassing
process in combination with the other
products. Instead of treating the entire
process unit with this chemistry, the
iron-control agent is typically added
only to the areas where pyrophoric iron
is anticipated.
It is normal industry practice to treat
this type of scale with potassium
permanganate or acids containing
formaldehyde or gyloxal, all of which
are hazardous materials. These processes
take place after the unit has been shut
down and the steam-out is complete,
the units isolated and significant
temporary piping and portable pumps
are installed. The cleaning process
generally takes place on the critical path
of the shutdown because little work can
proceed until the cleaning is complete.
The iron-control agent is a nonhazardous material that reduces time
and several volumes of chemistry and
waste disposal.

Environmental impact
All of the current SmartTurn process
degassing
and
decontamination
products are non-hazardous, water-

based, near-neutral-pH chemicals. The

components are not DOT regulated or
listed as CERCLA hazardous substances
and do not require reporting under Sara
Title III.3
The water-based chemistry is used at
low concentrations on waste that is
already being generated during the
steam-out phase of a shutdown and
processed in the plant wastewatertreatment facility. The materials are
biodegradable and have been formulated
to be compatible with biological
wastewater-treatment plants. They are
designed not to harm nitrifying bacteria.
In fact, they enhance bacteriological
growth.
The process may eliminate the need
for post-shutdown acid cleaning or
potassium permanganate treatments for
pyrophoric iron control and the
hazardous waste generated by them.
One problem often associated with
surfactant-based cleaners is foaming and
a tendency to form strong oil-in-water
emulsions. These water-based degassing
and surface-cleaning products are low
foaming and, although effective at
removing oils from the system, they do
not normally create long-lasting
emulsions. The ability to achieve rapid
oil/water separation helps to reduce
cleaning time, increase the volume of
reclaimed oil and reduce biological
demand on the wastewater-treatment
facility.
Another environmental factor is the
reclamation of the LCO/Parasol 370
dispersant volumes, including all the
dissolved/suspended heavy asphaltenic
material. This stage is drained from the
unit into a slop tank and processed with
the feed to an operating crude unit,
catalytic cracker or coker unit. The
product will generally not affect the
catalyst and/or allow the asphaltenes to
deposit in the unit. This can be a great
saving over the disposal of tons of RCRA
hazardous material from heat exchanger
cleaning and hydroblasting of the tower
bottoms, lines and trays.

Economic impact
As previously emphasised, the cleaning
process can save valuable time during
shutdown, and the steam-out phase
may be cut by as much as 50%. The
LCO/Parasol 370 dispersant cleaning
can
reduce
or
eliminate
the
hydroblasting required to remove the
heavy asphaltenic material from the
towers, piping and heat exchangers.
Adding the Ferrotrol 600 agent to the
degassing steam helps eliminate the
need for a special step to neutralise or
remove pyrophoric iron. Since each
system to be treated is unique, it is not
possible to quote specific time savings in
this article. However, some general
estimated time savings can be generated,
as follows:

www.eptq.com

Based on a $5.0 crack spread, a 100 000

bpd crude unit would expect to save in
production $3.07.5 million

Reduction in steam-out time is one to two days

Reduction in hydroblasting time is three to zero days
Elimination of pyrophoric iron step is two to three days.
The total time saved in many cases can be 615 days. The
crack spread is one of the tools used to determine the potential
production value for a barrel of crude produced. In the North
American Gulf Coast, a 3:2:1 ratio is normally used. The
spread is determined by subtracting the cost of 3.0 bbl of West
Texas intermediate from the wholesale price of 2.0 bbl of
unleaded gasoline and 1.0 bbl of wholesale diesel, then
dividing by 3.4. This ratio changes frequently and varies
over months and years. For the sake of this article, a crack
spread of $5.0 per bbl will be used as an example, which can
be easily changed to accommodate the crack spread at an
individual plant.
Based on a $5.0 crack spread, a 100 000 bpd crude unit
would expect to save in production $3.07.5 million by using
this process.
100 000 bpd x 6 days x $5 per bbl = $3 000 000
100 000 bpd x 15 days x $5 per bbl = $7 500 000
Savings will be realised in several other areas:
Reduced disposal of heavy asphaltenic material listed as
RCRA Hazardous Waste
Reduced disposal of hazardous chemical-cleaning waste
that must be treated off-site
Reduced turnaround day rate for cranes, scaffolding and
manpower
Eliminated need for repeated steam repurging to allow safe
entry.

VOCs, reducing the steaming time. The Ferrotrol 600 ironcontrol agent added to the degassing steam is designed to
remove the pyrophoric iron that is normally treated later in
the shutdown on critical path. In the majority of cases, all of
the work is performed using safe, environmentally friendly
chemistries.
The removal of asphaltenes and residual hydrocarbons
eliminates the source of VOC off-gassing that traditionally
requires repeated steam purging to maintain safe entry
conditions in the unit for inspection and hot work.
Depending upon plant practices and cleaning requirements,
615 days can be shaved from a plants turnaround schedule.
This potentially saves millions of pounds/dollars in increased
production time and reduced expenditure.

The article is based on a presentation (AM-07-57) from the 2007

NPRA Annual Meeting in San Antonio, Texas, USA. SmartTurn,
Parasol, Pentol-S3, PSol-542, HS-5, Ferrotrol 600, Parasol 370 are
marks of BJ Services.

References
1 Data supplied by Continuum Chemical Corporation.
2 International Program on Chemical Safety (IPCS) website.
3 Consolidated List of Chemicals Subject to Emergency Planning and
Community Right-to-Know Act (EPCRA) and Section 112 of the
Clean Air Act, EPA 550-B-01-003, October 2001.
4 New proxy for coking margins forget the crack spread, NPRA
AM-05-55.

Mike Lippold is operations co-ordinator, process services, at

BJ Process & Pipeline Services in Houston, Texas, USA. Lippold
graduated from Clarkson University in 1970 with a BS in
interdisciplinary engineering and management.
Email: mlippold@bjservices.com

Case history
A refinery was performing a major upgrade on an asphalt
plant. During the turnaround, it was planning to replace two
towers, heaters and major sections of piping to handle more
acidic crude oil. The modifications required over 400 cuts to
demolish the existing equipment. During past turnarounds,
the plant had experienced extensive delays for purging, as
hydrocarbon off-gassing repeatedly raised levels above the LEL
in its equipment.
The refinery elected to use the SmartTurn cleaning process
boundary-to-boundary to help facilitate the cutting process.
The cleaning was performed during the five-day shutdown
window without any delays. During the steam-out, H2S,
hydrocarbons, benzene and explosive vapour levels were
reduced to zero. Inspections showed the towers and piping in
the unit were cleaner than they had ever been. All exchangers
where the diesel/Parasol 370 dispersant had been circulated
were clean and did not require hydroblasting. The unit
remained free of hydrocarbons throughout the turnaround,
and none of the planned cuts were delayed due to high
hydrocarbon readings.

Conclusion
When fully implemented, chemical cleaning services can in
many cases take process unit cleaning off the turnarounds
critical path. Typically, all the cleaning takes place during the
shutdown/steam-out phase. Parasol 370 asphaltene dispersant
works to remove the heavy asphaltenic material that requires
extensive hydroblasting. The Pentol-S3 degassing agent
removes heavy oils, residual hydrocarbons and encapsulates

www.eptq.com

Kidextractor
Quarter page

After you have used

it the first time, the
Kid Extractor will
become your trusty
hydraulic-tubebundle-extractor

KIDExtractor Ltd.
P.O. Box 11, Zebbug
MALTA
Tel. 00356-21-462891
Fax. 00356-21-462755
MOBILE: 00356-94-20596
E-mail: idrojet@videobank.it
Website: www.kidextractor.com

PTQ REVAMPS 25

build on our foundation

Our depth of knowledge and experience gives UOP

customers a head start.
As the global leader in technology solutions for the petroleum
refinery industry since 1914, UOP understands what it takes to
help our customers achieve and sustain success. Today, with the
support of our new parent company, Honeywell, we reaffirm our
commitment to leadership in customer satisfaction and
innovation. From equipment design and consulting to process
technology and products like high-performance catalysts and adsorbents, UOP is the one
global company that can consistently add value to your project.

Process Technology Catalysts Adsorbents Performance Equipment Profitability Consulting

UOP LLC, 25 East Algonquin Road, Des Plaines, IL 60017-5017, USA phone: +1-847-391-2000 fax: +1-847-391-2253 www.uop.com
2007 UOP LLC. All Rights Reserved.

Revamping atmospheric
crude heaters
The revamp of an atmospheric crude unit heater, which was suffering from
coking caused by asphaltene precipitation and poor burner stability, resulted
in a significantly increased heater run length
Michael Whatley Navajo Refining Company
Scott Golden and Jason Nigg Process Consulting Services Inc

efinery atmospheric crude heaters

can experience rapid increases in
tube metal temperature (TMT),
requiring unplanned shutdowns for
decoking. In the case of the Navajo
Refining Company facility in Artesia,
New Mexico, USA, rapid increases in
the atmospheric crude heaters TMT
resulted in a shutdown every three to
six months. In this case, as well as others
observed in the industry, coke formation
was initiated by asphaltene precipitation
from unstable crudes.
Industry-wide, atmospheric crude
heater coking is an unusual problem,
with some heaters operating reliably at
an average radiant section heat flux of
1314 000 Btu/hr-ft2 or higher. However,
some crude oils, including those
produced from North American fields in
West Texas, New Mexico, Ohio/
Pennsylvania and Alberta are known to
be unstable when there is asphaltene
precipitation in certain areas of the
crude unit. The equipment in these
areas includes preheat exchangers, fired
heaters and atmospheric column flash
zone and stripping section internals.

Naphtha

Prefractionator

Crude

12

Naphtha

23

Desalted
crude
Atmospheric
crude column

Atmospheric crude
heater coking
The Artesia refinery was experiencing
chronic coking in its atmospheric
column heater, with periodic shutdowns
to remove coke at intervals as short as 90
days. Figure 1 shows the process flow
scheme with the heater located
downstream from the prefractionator
column. Flashed crude is charged to the
atmospheric crude heater, which feeds
the atmospheric crude column. The
heater was a vertical-tube hexagonalshaped four-pass design with 12 burners.
Prior to the revamp, the radiant sections
average heat flux rate was only 9400
Btu/hr-ft2 with an oil outlet temperature
of just 635F. Furthermore, the heater
had oil mass flux rates of 250 lb/sec-ft2
and relatively poor flame stability. This,
in combination with poor asphaltene
stability, caused a very short heater run
length even though the heater was

www.eptq.com

10

New impeller
and motor

Atmospheric
crude heater

Figure 1 Process flow diagram

operating at relatively mild conditions.
Some atmospheric crude heaters operate
at an average radiant section heat flux of
1314 000 Btu/hr-ft2 and oil outlet
temperatures of 730F or higher, while
meeting four- to five-year run lengths.

Heater coking
Coke forms because conditions in the
shock or radiant tubes cause the oil to
thermally decompose to coke and gas.
The TMT increases as coke lays down on
the inside of the tube. With rising TMTs,
heater firing must decrease or the TMTs
will progressively escalate until their
limit is reached. The heater must then
be shut down to remove the coke. Rapid

coke formation is caused by a

combination
of
high
oil
film
temperature, long oil residence time and
inherent oil stability. In the majority of
cases where atmospheric heater coking
occurs, the root cause is high average
heat flux, high localised heat flux or
flame impingement.

Oil stability
Oil thermal stability depends on crude
type. For example, some Canadian and
Venezuelan crude oils have poor thermal
stability and begin to generate gas at
heater outlet temperatures as low as
680F. At outlet temperatures much
above 700F, these same crudes begin to

PTQ REVAMPS 27

Naphtha

Prefractionator
Figure 2 Fouled exchanger tubes
asphaltene precipitation
deposit sufficient amounts of coke to
reduce heater runs to two years or less.
Another form of oil instability is
asphaltene precipitation. As the oil is
heated, the asphaltenes become less
soluble, depositing in low-velocity areas,
fouling crude preheat exchangers, heater
tubes or atmospheric column internals.
In some cases, the asphaltenes do not
drop out until they reach the bottom
of the atmospheric column or inside
the vacuum heater. With some Canadian
crude oils, especially the bitumenbased oil sands crudes, asphaltene
precipitation occurs inside the vacuum
heater tubes rather than in the
atmospheric heater. The same heater
design
parameters
that
improve
atmospheric heater performance also
increase vacuum heater run length.
When asphaltenes separate from the
crude oil, the material deposits inside
the tubes. This increases heat-transfer
resistance,
raising
asphaltene
temperature and TMTs. Furthermore,
when asphaltene deposits are widespread
in the convection or radiant sections,
heater firing must increase to meet the
targeted heater outlet temperature. This
leads to a higher localised heat flux,
further raising the temperature of the
asphaltenes deposited on the inside of
the tubes. The temperature of these
asphaltenes eventually exceeds their
thermal stability, resulting in coke
formation and even higher TMTs,
because the coke layer has lower thermal
conductivity than asphaltenes. Heater
TMTs eventually exceed metallurgical
limits, requiring a heater shutdown to
remove the coke. In this example, heater
run lengths were as low as 90 days
between piggings.

Asphaltene precipitation
Crude stability is a function of its source
and highly variable. However, the
designer can influence the process and
equipment design to minimise the effect
of poor asphaltene stability. In some
cases, the material deposits inside the
exchangers, piping, heater tubes or
fractionation column. The lower the
velocity, the more likely it is that
asphaltenes will precipitate. In this
example, the oil velocity inside the heater

28 PTQ REVAMPS

Crude

10
12

Naphtha

23

Desalted
crude

Atmospheric
crude column

New impeller
and motor

Atmospheric
crude heater

Figure 3 Increased system pressure drop

tubes was only 5.56 ft/s prior to the oil
vapourising, which corresponds to a 250
lb/sec-ft2 oil mass flux. At these velocities,
whether in an exchanger or heater tube,
asphaltenes will likely drop out. Heat
exchanger data gathered from hundreds
of operating exchangers shows the rate
of fouling and the ultimate fouling factor
are to a large extent determined by the
velocity of the crude flowing through
the tubes (assuming no shell-side
design problems).
Asphaltene precipitation in crude
preheat exchanger tubes is common.
Figure 2 shows asphaltene precipitation
inside the channel head and tubes in a
unit processing West Texas crudes. In
this case, the oil velocity inside the
tubes was less than 5 ft/s and severe

Maintaining a high
velocity in the equipment
minimises asphaltene
precipitation. Crude
preheat exchangers and
heater tubes should be
designed for oil velocities
of 810 ft/s or higher

fouling occurred. Exchangers operating

at higher velocities in the same unit had
less fouling. Moreover, the atmospheric
heater downstream of the fouled
exchanger had short heater runs, with
TMTs increasing at 1F/day. This rate of
TMT rise is similar to a delayed coker
heater. Crudes with poor asphaltene
stability are especially difficult to process
and the equipment must be carefully
designed.
Maintaining a high velocity in the
equipment
minimises
asphaltene
precipitation. Crude preheat exchangers
and heater tubes should be designed for
oil velocities of 810 ft/s or higher.
Experience
shows
significant
improvements in crude preheat and
heater reliability when velocities are
high. Since many designers set the
maximum allowable pressure drop
through exchangers and the heater as
design criteria, low velocities are often
the result of meeting pressure drop
specifications. Crude preheat exchangers
and fired heater designs should be based
on a higher velocity, with the pressure
drop simply a result of the design.

Heater revamp
In late 2005, Navajo revamped its
existing atmospheric heater and
installed a new parallel helper heater.
The helper heater was needed because
the existing heater burner spacing was

www.eptq.com

increased and the number of burners

decreased for improved flame stability
with Ultra-Low NOX Burners (ULNB),
resulting in lower design heater firing.
Yet, installing the parallel heater
without revamping the existing heater
would have decreased oil velocity to less
than 4 ft/s in the existing heater.
Moreover, Navajo wanted to increase
the
atmospheric
heater
outlet
temperature from 635670F to unload
the downstream columns because they
limited the crude charge rate. Since the
root cause of heater coking was
asphaltene precipitation inside the
heater, reducing heater firing alone
would not have improved the heater
run length. Furthermore, designing the
new parallel heater for similar tube
velocities as the existing heater would
have created a second reliability
problem.
The revamped heater included a new
convection section, completely retubed
radiant section, new ULNB burners and
floor. Prior to the revamp, heater
performance showed the convection
section was not performing well, radiant
section tubes were fouled and the
burners had poor flame stability. Poor
convection performance was caused
primarily by fin damage. Radiant section
fouling was caused by asphaltene
precipitation and poor burner stability
resulting from the burners being too
close together, causing adverse flame
interaction.
Total fired heater absorbed duty is
the sum of the convection and radiant
section
duties.
Maximising
the
convection section duty minimises the
radiant duty, which lowers the oil film
temperature, reducing the rate of coking.
In this example, the convection section
was replaced with a similar design,
except some of the tube fins were
upgraded from carbon steel to 1113 Cr
to avoid damage from high temperature
and
to
maintain
performance
throughout the run. The radiant section
was completely retubed with smallerdiameter tubes. Bulk oil velocities were
increased from 5.66 ft/s to almost 10
ft/s, which resulted in an oil mass
velocity of 460 lb/sec-ft2. The revamped
heater mass velocity is over twice the
rule-of-thumb (ROT) values of 150200
lb/sec-ft2 that have been used to design
atmospheric crude heaters. The smaller
tube diameter dramatically increased
the pressure drop, requiring a larger
pump impeller and motor in the flashed
crude pumps (Figure 3). It is not
uncommon for process design engineers
to specify the pump design before the
heater is designed because they assign a
maximum allowable pressure drop to
the heater. This approach expedites
design but causes heater reliability
problems once it is built.
Prior to the revamp, the burners had

www.eptq.com

96 radiant tubes,
24 per pass

Figure 4 Burner floor before revamp

120 radiant tubes,

30 per pass
New 6 Ultra-Low
NOx Burners

Figure 5 Burner floor after revamp

poor flame stability. The heater had 12
burners that were spaced too close
together (Figure 4). A completely new
floor was installed with six larger
burners, which eliminated flame
interaction, producing a stable flame
(Figure 5). These burners were latestgeneration ULNB burners.

Revamp results
The revamped heater and new helper
heater have been operating for 18
months without a shutdown. Crude
charge has been increased by 14% and
the heater outlet temperature has risen
from 635670F, with the revamped
heater operating at approximately
103% of the pre-revamp firing rate or
9700 btu/hr-ft2 average radiant section
heat flux. The heaters have not been

pigged and the TMTs have shown very

little rise since startup.

Michael Whatley is engineering manager

for the Navajo Refining Company refinery
in Artesia, New Mexico, USA.
Scott Golden is a chemical engineer with
Process Consulting Services Inc, Houston,
Texas, USA. He earned a BS degree in
chemical engineering from the University
of Maine and is a registered professional
engineer in the state of Texas.
Email: sgolden@revamps.com
Jason Nigg is a chemical engineer with
Process Consulting Services Inc, Houston,
Texas, USA. Nigg earned a BS degree in
chemical engineering from Kansas State
University. Email: jnigg@revamps.com

PTQ REVAMPS 29

Fluctuating ethylene and propylene prices affect your

bottom line. Do you have the flexibility necessary to adapt?
We do. KBRs SCORE and SUPERFLEX olefins process
technologies have enabled our clients to realize excellent
return on their investments. What are you waiting for?
To learn more, contact us at petrochemicals@kbr.com
or visit www.kbr.com.

2007 KBR. All Rights Reserved. KA436

PTQ2_07_Ad_(Full).indd 1

2/27/07 10:13:18 AM

Optimising turnaround
maintenance performance
Case study of a risk-based inspection technique where maintenance and inspection
activities were moved from a traditionally reactive or time-based approach to proactive
maintenance. The optimisation of both turnaround duration and interval is discussed
Salah Massoud Elemnifi Arabian Gulf Oil Company
Farag E Elfeituri University of Garyounis

urnarounds
are
the
major
maintenance activity for most
refining,
petrochemical
and
chemical plants. They are costly both
in terms of lost margin resulting from
plant downtime and the expense of the
turnaround inspection and maintenance.
The proactive methodologies endorsed
by the American Petroleum Institute
(API)
for
risk-based
inspection
provide guidelines for identifying and
quantifying degradation mechanisms
and risk to help in prioritising inspection
and care actions. Proactive maintenance
is characterised by the application
of methods, tools and techniques to
eliminate failures, extend component
life, mitigate consequences, minimise
downtime, and optimise all resources
through a systematic identification and
elimination of potential problems in
all aspects of reliability, availability and
maintainability.
The employment of proactive
maintenance at the Arabian Gulf Oil
Companys Sarir refinery in Libya
showed that the turnaround interval
and duration could be optimised based
on the risks associated with the
equipment; by removing items that
could be implemented as a routine
maintenance function from the
turnaround workscope; and the
application of online maintenance. Cost
analysis and evaluation revealed that
applying turnaround performance
optimisation could generate increased
profit and improved availability.
Refiners are under pressure to run
their plants safely, at minimum cost,
while maximising product throughput
over extended periods. Emphasis is
being placed on reducing operating cost
and improving plant availability. One
area for reducing cost is maintenance.
The term turnaround generally
refers to the major inspection outages,
whereas shutdown refers to the
planned
outages
between
the
turnarounds. Turnaround interval is
the time in between successive
turnarounds,
and
turnaround

www.eptq.com

Significant value can

be generated from
optimising turnarounds
and shutdowns. This
value is captured by
applying industry best
practice and experience in
turnaround and shutdown
management, planning
and execution
duration is the period from feed out
to the first products on grade (feed
out to product out). Significant value
can be generated from optimising
turnarounds and shutdowns. This value
is captured by applying industry best
practice and experience in turnaround
and shutdown management, planning
and execution at several levels in the
organisation.
Of all the Libyan refineries, the Sarir
refinery is the least attractive when it
comes to improving turnaround
performance. Although each refinery or
petrochemical complex is unique, this
uniqueness stems mainly from its
geographical location, which in
conjunction with the refinery process
configuration determines how efficiently
crude oils can be delivered to the
refinery and the products to markets
served by the refinery. The Sarir refinery
is the worst-case scenario for a number
of reasons:
Economical consideration Sarir is a
small refinery of 10 000 bpd capacity. It
is a simple refinery with a topping unit
(atmospheric distillation unit) and
reformer unit to produce motor gasoline.
Due to its location in the Sarir field and
its unique design, residuum from the
crude tower bottom is blended with
higher-priced crude oil for export.
Residuum is therefore sold at crude price

instead of blending with distillates to

produce lower-priced heavy fuel oil.
This situation makes Sarir the most
profitable refinery in the Libyan refining
industry. Shutting down Sarir for
maintenance will reduce the refining
margin and therefore decrease its
profitability significantly
Technical consideration
The Sarir refinery has only one crude
oil tank to receive crude oil. This makes
it more difficult to process crude in the
case of maintenance due to the high
wax content of the crude. In addition, it
has only one slop tank, which cannot be
maintained during normal operation
The crude preheat exchangers cannot
be bypassed in the case of plugging, as
in the case of other refineries like
Azzawia in Libya
Since the Sarir refinery is located in a
desert, it uses well water for cooling and
softened water for steam generation
The Sarir refinery consists of both a
topping plant and reformer unit, while
the Tobruk and Ras Lanuf refineries are
made up of a topping plant only
Administrative
consideration
Another problem in the Sarir refinery is
that most of the staff work in rotation
due its location.
The key to profitability is maximising
asset utilisation by operating the plant
at its optimum conditions for a
maximum period of time. The present
situation in the Sarir refinery often
includes the following unique features:
Frequency and duration of the
turnaround is set in stone the
turnaround takes 30 days and is done
every two years
Execution is adjusted to the
predetermined duration; there is no
incentive to do it in less time
It is better to work in the
turnaround, since the plant is shut
down and everything is idle, not
realising that if it were possible to do the
work as routine maintenance it would
cost about half of what it does during
turnaround, with better quality of
workmanship.

PTQ REVAMPS 31

Sarir refinery turnaround history

January February

March

1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006

Apri

May

June

July

August September October

Sarir

November

December

Sarir
Sarir

Sarir
Sarir
Sarir
Sarir
Sarir

Table 1

Percent

225

The objective is to move away from

planned
maintenance,
which
is
distinguished by schedules that are
followed whether the work is needed or
not. Planned maintenance sounds good
and has been widely practised in
industry for many years, but it can be
excessively expensive when unnecessary
maintenance is performed on lowpriority equipment.

210.3

150
110.4

100

92.8

99.9

80.7

Research questions

75

Questions often posed to management

in the face of turnaround performance
optimisation are:
What are the constraints and
strategies for extending the turnaround
interval and reducing the turnaround
duration?
What equipment should be included
in this turnaround?
What is the risk of removing certain
items from the turnaround schedule?
What groups of units should be
turned around at the same time?

0
1994

1997

1999

2001

2003

2005

Figure 1 Turnaround costs for Sarir refinery compared to 1994 costs

350
No turnaround
Turnaround

300

283

Percent

268
240

250
200

188

150
100

297

133
100

144

159

147
109
68

50
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Figure 2 Annual processing costs for Sarir refinery compared to 1994 costs

Study rationale
The term optimisation implies a
single point or goal of maximum plant
production capacity at a minimum cost.

32 PTQ REVAMPS

The goal of maintenance optimisation

would be to achieve the highest level of
reliability for the least investment in
parts and labour.

Research methodology
The
methodology
includes
an
assessment of current turnaround
performance and benchmarking against
the industry leaders by way of key
performance indicators. The results of
these initial steps are presented in the
form of a report. An improvement
plan
is
then
developed
and
implemented over the turnaround or
shutdown cycle.

Turnaround performance
assessment
The key drivers for managing and
improving turnaround performance are
availability, risk and cost management.
Availability is a function of reliability
and maintainability. Reliability is
increased as the frequency of outage is
reduced and the time between failures

www.eptq.com

Availability
95.55*
96.82
97.88
97.84
98.56

* Current availability

Table 2
or shutdowns is increased (turnaround
interval). Maintainability is increased as
the duration of the plant, subsystem or
equipment downtime is reduced
(turnaround duration).
The impact of turnarounds on overall
plant availability can be expressed in the
equation:
Availability =

Uptime
Uptime + downtime

This can be written as:

Availability =

Turnaround interval
Turnaround interval +
turnaround duration +
breakdown outage

The executed turnarounds at the Sarir

refinery are shown in Table 1. Table 2
illustrates the current availability of the
Sarir refinery and the suggested
availability for intervals of three and
four years, and durations of 31 and 21
days respectively. As can be seen from
this table, increasing the turnaround
interval from two to three or four years,
while keeping the same turnaround
duration of 31 days, will increase
availability from 95.55 to 96.82 or 97.88
respectively. The effect of reducing the
turnaround duration from 31 to 21 days
would result in further increased
availability of 97.84 and 98.56
respectively.
The risk of an unplanned outage due
to reliability issues generally increases
with time from the previous turnaround.
More frequent turnarounds reduce the
risk of forced shutdowns while
increasing revenues losses. Cost savings
due to extending run lengths are often
offset by an increased risk. Inspection
influences the uncertainty of the risk
associated
with
the
equipment,
primarily by improving awareness of the
state of deterioration and predictability
of the probability of failure. Although
inspection does not directly reduce risk,
it is a risk management activity that may
lead to risk reduction.
Figure 1 shows the total turnaround
cost in Libyan Dinars for the Sarir
refinery. It also illustrates the fluctuation
in total turnaround cost from year-toyear for the period from 1994 to 2005.

34 PTQ REVAMPS

Benchmarking
All Libyan the refineries except Sarir
extended turnaround intervals from two
to three years based on history. The bestin-class refineries ran their plants for
more than four years without any
outage. In addition, the duration of their
shutdowns (turnarounds) was shorter
than 23 days.

zone. Of the vessels that can be

inspected and maintained during
normal operation, six of these were in
the very low-risk zone, four in the lowrisk zone and only one in the moderaterisk zone.

Moving rotary equipment to

routine maintenance
All redundant equipment such as
pumps, boilers, gas compressors and air
compressors could be removed from the
turnaround workscope and maintained
as part of a routine maintenance plan.
The routine maintenance of pumps and
gas compressors, generally all redundant
equipment, can be implemented
without any threat to the operating
stability.

Conclusions

Improving performance
The technique known as risk-based
inspection (RBI) was applied to the
vessels in the refinery to improve
performance. It was determined that out
of the 71 vessels included in the study,
39 were rated in the very low-risk zone,
25 in the low-risk zone, five in the
moderate-risk zone, two in the high-risk
zone and no vessels in the very high-risk

The purpose of the study was to

optimise turnaround performance at
the Sarir refinery. The goal was to move
from planned maintenance to proactive
maintenance
by
increasing
the
turnaround interval and decreasing the
turnaround duration to increase
availability and lower maintenance
cost. The turnaround frequencies have
traditionally been time-based and

1000

751

Percent

Parameters
Duration
Interval
31
2
31
3
31
4
21
3
21
4

It can be seen from Figure 2 that the

processing cost is higher during the
turnaround years compared to other
years. Figure 3 shows the net profit for
the refinery during 2002 to 2005
compared to the 1994 profit. Figure 4
summarises the net profit and savings
for the Sarir refinery during the
turnarounds of 2003 and 2005 in the
case of turnaround optimisation.

540
500

142

100
0.0
2002

2003

2004

2005

Figure 3 Annual refining margin for Sarir refinery compared to 1994 margin

200

Turnaround
Saving

No turnaround

145

150

126
Percent

Sarir refinery availability

100

100

100

45

50

26
0
2003

2005

Figure 4 Refining margin for Sarir refinery in the case of turnaround optimisation

www.eptq.com

driven by conservative practices of

regulation and industry.
The first concern of the optimisation
process was that extending the interval
would increase the probability of
equipment failure and consequently the
turnaround workscope, which would
increase the turnaround duration. If
equipment
fails,
the
resulting
unscheduled outage is more expensive
than a scheduled outage. The cost is
much higher if the outage is due to a
catastrophic failure. To address this
concern, RBI was applied as a new
approach to increase productivity and
reliability, focus inspection resources,
reduce inspection and maintenance
costs, comply with regulatory or
insurance regulations, and improve
outage planning.
All of the 71 vessels in the three
process units were evaluated in the Sarir
refinery.
The
probability
and
consequences of failure for each vessel
was determined. A risk matrix was then
developed to provide a risk ranking for
each vessel.
The results of risk-based inspection
application at the Sarir refinery showed
that the turnaround interval and
duration could be optimised without
increasing the risks associated with
equipment failure. Turnaround duration
could be decreased by the application of
RBI as well as by removing items from

extended without any fouling threat or

adversity. Cost analysis and evaluation
of the 2003 and 2005 turnarounds have
revealed that increased profit can be
extracted by applying turnaround
performance
optimisation.
An
improvement in availability can also be
attained.
Industry
benchmarking
supports the feasibility of turnaround
optimisation at the Sarir refinery.

Increased profit can be

extracted by applying
turnaround performance
optimisation. An
improvement in
availability can also be
attained
the turnaround workscope to a routine
maintenance plan.
The other concern was that the
intervals between relief valve examinations needed to be extended. An
improvement process that would solve
this problem involves the application of
online maintenance as a new technique.
In addition, this problem can be solved
by traditional techniques that include a
partial shutdown to replace valves with
standby, twinned relief systems or
isolation valves in the relief stream,
enabling relief valves to be removed for
overhaul without shutting down equipment and using a rupture disk/relief
valve combination to increase reliability.
Data were collected for crude
exchangers working under severe
conditions through the bypass of the
crude feed tank. Analysis showed that
the turnaround interval could be

References
1 Dyke S, Optimizing plant turnarounds,
PTQ, 2004, 145151.
2 Fourth Libyan Refineries Meeting held at
Jowf Oil Technology, Benghazi, Libya, 1516 March 2006.
3 Williams J, Worldwide trends in RAM
improvement, PTQ, 2004, 100103.

Salah Massoud Elemnifi is a process

engineering specialist, refinery, Arabian
Gulf Oil Company, Benghazi, Libya.
Elemnifi holds a BSc degree in chemical
engineering and MSc from engineering
management department, The Academy
of Graduate Studies, Benghazi, Libya.
Email: salahmass@yahoo.com
Farag Elfeituri is an associate professor,
industrial and manufacturing systems
engineering, University of Garyounis,
Benghazi, Libya. Elfeituri received a
MASc and PhD from the industrial and
manufacturing systems engineering,
University of Windsor, Ontario, Canada.
Email: felfeituri@garyounis.edu

Aptech Engineering Services, Inc.

Providing Innovative and
Cost-Effective Solutions for
Total Asset Integrity Management
APTECH is a world leader in
providing Risk Based Inspection
(RBI) solutions and software,
Risk Assessment and Reliability
Modeling and Total Asset Integrity
Management Solutions
Let APTECH show you how our
SMARTTM Technology can help your
assets perform Safely, Reliably, and
Cost Effectively
Aptech Engineering Services, Inc.
Toll Free: 800.568.3201
Fax:832.593.0551
www.aptechtexas.com
E-mail: info@aptechtexas.com

www.eptq.com

Risk Based
Inspection, Risk
Assessment,
Software
Tools
Maintenance,
Reliability,
Equipment Life
Management

Metallurgical
Services, Failure
Analysis, Fitness
for Service

Total Asset
Integrity
Management
Mechanical
Integrity,
Process Safety,
Audits,
Training

Inspection
Programs,
Advanced NDE,
Corrosion
Studies
Design,
Operational
Support,
Materials
Selection

PTQ REVAMPS 35

ETHYLENE OXIDE CATALYSTS

| HYDROGEN & SPECIALTIES | STYRENE CATALYSTS | HYDROGENATION CATALYSTS | ENVIRONMENTAL CATALYSTS & SYSTEMS

With the men and women of CRI beside you, youll never take a wrong

With our expert catalyst

guidance, youll stay on

turn in catalysts. Well help you innovate when you need to move in a
new direction, and well be there with expert support down the road.
Whether you need catalysts for ethylene oxide, distributed hydrogen
products, styrene, environmental applications, selective hydrogenation

the path to success.

catalysts from KataLeuna, or custom products, the people of CRI will

keep you on track.

+1 281 874-2600
catalystsales@cri-criterion.com
www.cricatalyst.com
2007 CRI Catalyst Company. KataLeuna is a trademark of CRI Catalyst Company. (CCC706)

ccc706 hiker 210x297HE indd 1

7/23/07 10:19:56 AM

OSBL considerations for

expansion projects
Example cases of refinery revamp and expansion projects, where the aim is to identify
and compare the various OSBL requirements for each project. Guidelines are presented
for improving the accuracy of the OSBL definition early in project development
Dominic M Varraveto
Burns & McDonnell

s the demand for transportation

fuel has remained strong despite
rising prices, and refinery
utilisation hovers near maximum
sustainability, interest in refinery
expansion projects has increased. These
projects can be divided into two major
parts, distinguished by the industry as
inside battery limit (ISBL) and outside
battery limit (OSBL). The ISBL portion
typically focuses on the addition of new
or extra process units to the existing
refinery configuration, while the OSBL
portion consists of all the other necessary
supporting elements, including utility
systems,
interconnecting
process
and utility piping, feed and product
handling, and storage. In some cases,
the OSBL may also include an entire
unit for the production of a key process
reactant, such as hydrogen, and for the
safe handling and disposal of waste
streams and by-products like sour water
and sulphur.
Many projects include a revamp of
existing process units as part of the
OSBL. As such, this article focuses on the
OSBL portion of three example refinery
expansion projects consisting of:
Replacement of an existing vacuum
distillation unit
Adding a new vacuum distillation
unit and a new delayed coker (DCU)
Adding and integrating a new crude
distillation unit (CDU), DCU and a
hydrocracker (HDC) to a typical FCCbased gasoline refinery.
The aim of this article is to identify
and compare the various OSBL
requirements
for
each
example
expansion project and to discuss the
guidelines for improving the accuracy of
the OSBL definition early in the projects
development.
A meaningful discourse about OSBL
considerations for a refinery project and,
more specifically, an expansion project
relies on a clear definition of what is and
what is not meant by OSBL, as this
term does not have a universally
accepted definition. For different
departments in the same organisation,
the OSBL portion of a project often

www.eptq.com

LPG
HDT/
isom.
HDT/
isom.
Crude

Atm.

Gasoline
HDT

Gasoline

NHT/
reformer
Distillate
HDT
GO HDT/
FCC

Distillate

DCU

Coke

Vac.

Existing piping replaced

Existing unit replacement

Figure 1 Vacuum unit replacement

varies, especially during the projects
conception, and often even further into
its development. This lack of a common
definition can exist among organisations
too, such as between owner and
engineering/construction contractor.
The entire scope of a refinery project
can be split between ISBL and OSBL, and
the OSBL portion can be further divided
into subcategories according to the
nature of the work. Also to be examined
in detail are recent refinery projects and
how these projects are affected by
various factors, including return on
investment (ROI), growth in demand for
refined products, access to crude supplies
and the impact of environmental
compliance. Recently, solid operating
margins and continued profitability
have seen refinery owners come out of
the latest round of environmental
spending for low-sulphur gasoline (LSG)
and ultra-low-sulphur diesel (ULSD),
with an increased appetite for additional

refinery investment changing the

fundamental nature of the typical
refinery expansion project.

Definition of OSBL
A somewhat overly simplistic definition
of the OSBL relative to refinery
expansion projects is all aspects that are
not included in the ISBL portion of the
project. In other words, the OSBL is
everything else, including scope that
was unexpected and unplanned for
during the initial project development.
By this definition, the OSBL will be
incompletely defined until it presumably
can be defined in the latter stages of the
development process. A more specific
way to define the OSBL is to initially
identify what everything else includes
as early as possible in the project.
The OSBL will include more than
interconnecting piping and pipe racks.
Also
included
is
infrastructure,
secondary process units, revamps to

PTQ REVAMPS 37

OSBL classifications
Classification
Interconnecting systems
Infrastructure

Description
Pipe/pipe rack, power, controls
Utility system equipment, roads, tanks, truck and/or rail feed
and product unloading/loading, buildings
Secondary process units WWTP (wastewater-treatment plant), ARU (amine regen
unit), SWS (sour water stripper), SRU (sulphur recovery unit),
TGTU (tail gas-treating unit), SMR (hydrogen), air separation
Revamp of existing units Increase throughput and/or conversion
Financial
Owners cost, startup, working capital, interest, taxes
Other
Land, permits, fees

SRU/TGTU processes the concentrated

sour gases from the SWS and ARU,
producing molten sulphur by-product
and dischargeable vent gas. Other
process units that might be included in
the OSBL secondary process units are
facilities to produce a key reactant for a
primary process unit. Examples are a
steam/methane
reformer
(SMR)
producing hydrogen for a hydrotreater
or hydrocracker (HDC), and an airseparation plant producing oxygen for
sulphur plant oxygen enrichment or for
petroleum coke gasification.

Table 1
Examples of OSBL equipment for utility infrastructure
Utility
Steam
Water
Water
Air
Fuel
Power

System
Boiler
Boiler feed

Equipment
Deaerator, boiler, circulation pumps
Filters, ion exchange, chemical treatment, RO,
membrane preheat exchangers, pumps, tanks
Cooling
Cooling tower, circulation pumps, chemical treatment
Instrument/plant
Compressor, intercooler, drier
Fuel gas conditioning
Mix drum, condenser, liquid KO drum,
filter/coalescer, preheater, strainer(s)
Electric
Transformers, power distribution

Table 2
Examples of OSBL equipment for waste streams
Utility
Excess vapour/
overpressure protection

System

Equipment

Flare

Sewer

Process

Sewer
Condensate

Storm
Steam

Flare header, KO drum/pump, flare, flare

gas-recovery compressor
Sump, CPI (corrugated plate interceptor for
oil removal), pump(s), slop oil tank
Sump pump(s), tank or pond
Collection drum, pump(s), tank

Table 3
existing process units and financial
costs, such as interest and taxes. The
infrastructure includes not only utility
system equipment, but also roads, rail
spurs, docks, loading and unloading
equipment and buildings (warehouse,
power house, control room and office
buildings).1 Table 1 lists the classifications
that can be used to differentiate the
OSBL scope and summarises the items
included in each category.
The infrastructure is very general and
can be further categorised by considering
the process equipment to supply typical
utilities, including steam, water (boiler
feed and cooling) air (instrument and
plant), fuel (natural gas and refinery fuel
gas) and power (electric). Table 2 lists the
equipment frequently associated with
some of the more common OSBL
infrastructure.
Additional utility systems, not shown
in Table 2, which may require equipment
for reliable supply, might include
nitrogen, potable water, utility water
and fire water.

38 PTQ REVAMPS

The infrastructure also includes

systems and associated equipment to
handle return streams and wastes. For
refinery units, these systems include
steam condensate collection/return flares
and sewers. Table 3 lists the equipment
associated with these systems.
Some utility systems such as sewers
will require wastewater treatment,
which should be considered in the
OSBL classification of secondary or
supporting process units. Essential
secondary process units in todays
refinery for treating waste include a
wastewater-treatment plant (WWTP)
and sulphur-recovery facilities usually
comprising a sour water stripper (SWS),
amine regeneration unit (ARU), sulphurrecovery unit (SRU) and tail gas-treating
unit (TGTU).
The SWS and ARU process liquid
streams, removing hydrogen sulphide
(H2S) and ammonia (NH3) so that the
liquid streams can be reused (ie, lean
amine) or processed further and
discharged (ie, stripped sour water). The

Factors affecting
expansion projects
Until recently, it has been difficult for
refiners to attract investment capital
from both internal and external sources.
This was partly down to low ROI due to
poor margins, surplus capacity and
environmental regulation. In that kind
of economic environment, limited
capital was primarily appropriated for
environmental compliance to remain
viable. Investment for economic benefit
(ie, profitability) was limited or nonexistent. However, through creative and
clever improvement of existing facilities,
some refiners conceived and executed
low-cost projects that had significant
bottom line impact. These projects were
usually characterised by step-wise
debottlenecking over several years that
resulted in overall increases in refinery
throughput and/or conversion. These
projects were typically completed in
conjunction with regularly scheduled
maintenance turnarounds and did not
involve the addition of new process units
or improvement to existing infrastructure.
The OSBL portion of these projects was
small or almost insignificant.
For a time, this approach was adequate
to keep pace with growing demand.
However, during this time, many
refineries, especially smaller, less complex
facilities, were forced to close. These
closures were accelerated by the cost
of environmental compliance to
manufacture LSG and ULSD, and to
reduce the air and water pollution
originating in refineries. Over the same
period, growth in the demand for
transportation fuel increased steadily.
The combination of growing demand
and loss of capacity from refinery closures
eventually exceeded the increased output
from incremental debottlenecking and
created a much more balanced or, some
would say, precarious market for refined
transportation fuel. Margins and
profitability have improved dramatically,
especially for refiners capable of
processing heavier, higher sulphur crude
oil. With the bulk of the large
environmental expenditures already
made, at least for the near- to mid-term
(three to five years), there is an increasing
appetite for relatively large expansion

www.eptq.com

projects. These larger expansion projects

are now characterised by significant
capacity increases of 50100%. In some
instances, this involves a second parallel
processing train with new infrastructure
and secondary process units. The OSBL
portion for these projects is a significant
percentage of the overall project cost.

Case study 1
Vacuum unit replacement
The previously noted vacuum unit
(VAC) replacement project involved the
replacement of existing equipment,
including the vacuum unit furnace,
vacuum fractionation tower, overhead
ejector system, heavy vacuum gas oil
(HVGO) pumps and wash oil pumps,
and the addition of new HVGO vs crude
heat exchangers, as well as modifications
to the existing crude preheat system,
HVGO heat-removal system, vacuum
unit charge pumps, and crude and
vacuum section heat integration. The
project positioned the refinery for
greater flexibility in crude feed selection,
allowing heavier crudes with higher
percentages of gas oil to be processed.
The increased gas oil filled the excess
FCC capacity.2
The OSBL for this project was limited
and not identified separately from the
ISBL as described. By definition, the VAC
replacement excluded the revamp work
to existing process units. Spare FCC
capacity already existed and the design
basis defined the heavier crude so that
the gas oil content increased, but the
residual content (ie, the DCU feed) would
not increase as a result of the project. The
replacement of portions of the HVGO
product line to and around the gas oil
hydrotreater processing the FCC unit
feed was the only significant OSBL work.
The HVGO piping replacement was less
than 5% of the overall project and all the
OSBL work was less than about 10% of
the overall project.

Case study 2
New VAC/DCU
Reconfiguring an existing refinery to
process heavier crude and meet clean
fuel requirements was achieved in this
project by adding a new VAC and DCU
(Figure 2). The revamp of at least 11
existing process units was part of the
OSBL scope, including converting an
existing residual oil desulphuriser to a
gas oil hydrotreater; a shift in FCC
conversion to a higher gasoline yield;
upgrade of the light ends fractionation
and treating areas; and the modification
of two distillate hydrotreaters for
ULSD production. In addition, the OSBL
scope included upgrade and expansion
of the refinery amine circulation and
treating, sour water collection and
hydrogen distribution. A new hydrogen
plant was also included in the
OSBL scope.

www.eptq.com

LPG
Alky./
poly.
HDT/
isom.
Crude

Atm.

Gasoline
HDT

Gasoline

NHT/
reformer
Distillate
HDT
GO HDT/
FCC

Distillate

DCU

Coke

Vac.

Existing piping replaced

New process units
Revamp existing process units

Figure 2 New vacuum unit and delayed coker

The largest components of the
OSBL scope were two miles of pipe rack,
the corresponding interconnecting
process and utility piping, and over
500 tie-ins needed to install and support
the pipe. There were many projectspecific requirements, including the
extensive application of heat tracing,
both steam and electrical, to new and
existing interconnecting pipe. The
introduction of new intermediate
process streams from the DCU required a
significant modification to existing
tankage. These modifications varied for
each stream, but included new gas
blanketing, a new floating roof,
recommissioning of an existing tank
heater, new insulation, new heat tracing,
and the addition of water-draw
collection, piping manifolds and
pumps.
Overall, the OSBL was slightly above
30% of the total project cost. The revamps
comprised about one-third or 10%, and
the interconnecting piping and tankage
modifications were about two-thirds or
20% of the total project cost. Other than
tankage, the OSBL for this VAC/DCU
project did not have a significant
infrastructure component. This is
somewhat unexpected considering the
size and impact on the configuration of
the refinery. Possible reasons for this
relatively small infrastructure requirement
were because some of the typical OSBL
infrastructure, such as the coke-handling
facilities, were included in the DCU ISBL,
and the expansion project was based
on a shift to lower-cost crude and
increased conversion rather than higher
throughput.

Case study 3
New CDU/DCU/HDC
The addition of a new CDU/DCU/HDC
to an FCC-based refinery is the most
extensive project of the three presented
in this article. It combines a greater than
50% rise in throughput with an increase
in conversion for processing a
significantly heavier and higher sulphur
crude. In fact, the sulphur recovered
from the refinery will increase by nearly
an order of magnitude as a result of
this project.
Like the previous VAC/DCU example,
this project involves a new VAC unit and
DCU. In this example, the new VAC is
part of an entirely new CDU. The
differences end there, since this project
also includes a gas oil hydrocracker
(HDC), which removes a significant
amount of sulphur and converts a large
portion of gas oil, producing distillate
that meets ultra-low-sulphur (clean fuels)
requirements and also produces a
hydrotreated FCC feed stream. Applying
the definition established earlier, the
OSBL will include interconnecting pipe,
pipe racks, electric power supply and
distribution, infrastructure, secondary
process units and revamps. Figure 3
shows the major aspects of the ISBL
portion for this example. Not shown in
this figure are all the new secondary
process units. These new facilities are
part of the OSBL and include hydrogen
production, sour water collection and
processing, amine circulation and
regeneration, sulphur recovery, tail gas
treating and wastewater treatment.
There are also a significant number of
revamps associated with this example

PTQ REVAMPS 39

LPG
Alky./
poly.
HDT/
isom.
Crude

Atm.

Gasoline
HDT

Gasoline

NHT/
reformer
Distillate
HDT
FCC

Vac.

HDC

Distillate

DCU

Coke

Existing piping replaced

New process units
Revamp existing process units

Figure 3 New vacuum unit, delayed coker and hydrocracker

that will be part of the OSBL portion of
the project. They include upgrading the
existing crude unit to process heavier
crude, modifying the FCC for a higher
rate and shift in yield, expanding the
existing DHT for increased throughput
and higher-feed sulphur, increasing
throughput and adding the capability to
process DCU naphtha in the naphtha
hydrotreater (NHT) and reformer, and
expanding the light ends treating,
fractionation and processing in the
alkylation and polymerisation (ALKY/
POLY) area.

Recommendations: OSBL
definition quality
Why do the OSBL costs of refinery
expansion projects consistently exceed
expectations by large margins? There are
many possible reasons. Taken in
combination, the difference between
expectations and actual outcome will be
significant. Some of these reasons include
factors that affect both definition and
pricing of the OSBL scope.
The three previously discussed
example cases illustrate why many may
be unfamiliar with the size and
complexity
of
evolving
refinery
expansion projects. Until recently,
expansion project scope was limited to
the revamp of a single existing unit with
limited OSBL/infrastructure content,
such as case study 1. The goal for this
type of project was primarily to reduce
the raw material cost and improve the
operating margin, with less emphasis on

40 PTQ REVAMPS

large increases in throughput and

product volume. The more recent, larger
projects combine very large increases in
throughput coupled with a significant
shift in crude feed gravity and sulphur
content. These projects are relatively
recent and present a steep learning
curve to the inexperienced. Assembling
an experienced project team and
planning and organising a disciplined
project development effort are essential
first steps for better meeting project
expectations.
Some might say it is foolhardy to
estimate OSBL/infrastructure cost too
early in project development.1 On the
other hand, more time is required to
successfully define a large, complex
project. A lot of the early activity will
not require cost estimating. Early
activity should focus on gaining a
realistic understanding of the capacity
of existing infrastructure to support the
additional load and what must be
included in the scope to meet sitespecific needs. If necessary, elements of
the existing infrastructure should be
inspected and tested while in service
to prove suitability and capacity.
Determining the availability of open
space, the extent of demolition and site
clearing, plus the relative location of the
ISBL, OSBL and infrastructure is crucial.
OSBL development should be started
early, even while the ISBL is still being
developed. The initial site-specific OSBL
and infrastructure scope can be
developed early using similar or go by

ISBL designs. Utilities and infrastructure

needs can also be determined from the
go by designs using scaling factors.
The scaling factors can be validated later
in the development when a firm ISBL
definition emerges. Develop site-specific
OSBL and infrastructure scope, taking
into account the relative location of the
ISBL and infrastructure.
Conceptual cost estimating for
refineries and chemical plants is typically
done by applying installation cost factors
to sized equipment.3 While this
technique has proven successful for the
ISBL portion of projects, it is less than
satisfactory for OSBL, especially projects
containing
a
relatively
large
interconnecting pipe and pipe rack
component. For this type of OSBL scope,
using
technical
deliverables
like
preliminary layouts, pipe/rack routing
and electrical one lines to produce
material take-offs is a much more reliable
method for costing the non-equipment
component of the OSBL. Where
significant equipment-based infrastructure
is needed, installation cost factors can be
used. The scope definition should be
checked by comparison against other
similar projects.
Adjustments for differences related to
variation between projects due to siteand
project-specific
characteristics
should be taken into account when
making the comparison. In addition,
the use of project development check
lists specific to each project phase and
rating tools to quantitatively assess the
completeness of the scope definition
should be included near the end of each
project phase in the development
process. Finally, given the evolution of
refinery expansion projects from small
to large and simple to complex, time
and funding should be included in the
project development budget near the
completion of the OSBL development
effort for an independent assessment of
the costs outside project influence.4
This article is based on a presentation at the
2007 NPRA Annual Meeting in San Antonio,
Texas, USA.
References
1 Smith R S, Factored estimating guidelines
refinery/petrochemical projects, Cost
Engineering, 39, 4, April 1997, 3641.
2 Spangler R, et al, Flexibility in design
of crude unit revamp AM-06-14, NPRA
Annual Meeting, Salt Lake City, Utah,
March 2006.
3 Page, J S, Conceptual cost estimating manual,
2nd Edition, Section 18, Elsivier, 1996.
4 McIntire P, Cost estimating challenges face
frontier projects, Oil & Gas Journal, 99, 33,
August 2001.

Dominic M Varraveto is refinery process

manager at Burns & McDonnell in Kansas
City, Missourri, USA. Varraveto has a BS
in chemical engineering from University
of Notre Dame and a MS in engineering
management from the University of
Kansas. Email: dvarraveto@burnsmcd.com

www.eptq.com

see more

Greater opportunity requires a unique vantage point across your entire enterprise.
Honeywell is more than an industry leader in process automation, we also offer cutting-edge
technology and the services you need to oversee and optimize your plants operation. Our
applications go beyond the control system, allowing you to view your plant data in context
as well as integrate other relevant intelligence. Our offerings in cyber and physical security, wireless solutions,
and advanced applications like simulation technology, and planning and scheduling, better optimize your
facility while keeping your people safe. At Honeywell, we can help you see the information you need to
make better decisions faster for increased reliability, productivity and greater profit.

To see more Honeywell solutions, please call 1-877-466-3993 or visit www.honeywell.com/ps

2006 Honeywell International, Inc. All rights reserved.