ptq

Q3 2015

petroleum technology quarterly

refining
gas processing
petrochemicals

special features
PROCESS OPTIMISATION & CONTROL
ROTATING EQUIPMENT & FLUID FLOW

We create
chemistry
that makes
individual
refiners love
fueling the
world.

As the global leader in catalysis, BASFs drive is to

fuel the world by providing innovative technologies,
products, and services to enhance performance and
sustainability for the petroleum refining industry. We
create and deliver value to refineries through flexible
FCC catalysts and additive solutions, and advanced
technical service. When individual solutions solve global
problems, its because at BASF, we create chemistry.
www.catalysts.basf.com/refining

ptq
PETROLEUM TECHNOLOGY QUARTERLY

Q3 (Jul, Aug, Sep) 2015

www.eptq.com

3 A premium discount
Chris Cunningham
5 ptq&a
15 Managing fouling in refinery networks
Simon Pugh and Edward Ishiyama
IHS Energy
33 Dealing with depleting sour gas reserves: Part 1
Marco Van Son and Shashank Gujale
Jacobs Comprimo Sulfur Solutions
43 Revamping a hydrocrackers overhead condenser
Mustafa Gren and Ahmet Bebek
Tpras Kirikkale Refinery
51 Optimising preflash for light tight oil processing
Soun Ho Lee
GTC Technology
61 Splitting distributed control for a refinery expansion
Peter Helmersson and Sven Winberg
Preem AB
65 Cyber security in the refining world
Eric Knapp
Honeywell Process Solutions
71 Countering marine corrosion in instrumentation tube
Andreas Furukrona
Sandvik Materials Technology
77 Energy savings from electric capacity control
Klaus Stachel Hoerbiger
81 Revamp of a reciprocating compressor unit
Andreas Hahn
Neuman & Esser
91 Advanced turbomachinery controls improve ethylene plant yield
Medhat Zaghloul
Compressor Controls Corporation
97 Mixing in large crude tanks
Janakiramulu Adepu, Rahul C Patil, Ajay Gupta and Asit Kumar Das
Reliance Technology Group
Pravin Potdar, Jisna Raghavan and Kannan Srinivasan
Central Technical Services, Reliance Industries Limited
103 Improved feedstocks for ethane crackers
Ron Key
Linde Process Plants
107 Controlling the dew point of natural gas
Tek Sutikno
Fluor Enterprises
115 Testing and experience in FCC catalyst selection
Carel Pouwels Albemarle Catalysts Company
Ken Bruno Albemarle Corporation
131 Implementing real time optimisation technology
Hamid Ganji, Saeid Shokri, Mahdi Ahmadi Marvast, Soroo D Zahedi Abghari
and Behrouz Nonahal
Research Institute of Petroleum Industry (RIPI)
135 Technology in Action
LyondellBasells olefins manufacturing plant at Channelview, Texas.

Photo: LyondellBasell

2015. 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 and its supplements the publisher cannot be held
responsible for any statements, opinions or views or for any inaccuracies.

KBC Adv - PTQ Q2 2015.pdf 1 3/4/2015 12:12:35 PM

Advanced
Technologies
C

CM

MY

CY

CMY

MAXIMISE YOUR ASSETS POTENTIAL

KBC has just released new versions of its engineering software suite, which empower you with the industrys
leading tools for Maximising Asset Performance. From strategic planning and investment studies through unit
optimisation and performance monitoring throughout your value chain, KBC can help you make decisions with
confidence based on the fidelity and granularity of our technologies. KBCs superior platform enables
management of your complex, multi-dimensional challenges and adds identifiable value to your bottom line.

SEE WHATS NEW

www.kbcat.com/technology

KBC Advanced Technologies

AMER: +1 281 293 8200 APAC: +65 6735 5488 EMEA: +44 1932 242424

answers@kbcat.com

www.kbcat.com

blog.kbcat.com

ptq

A premium
discount

PETROLEUM TECHNOLOGY QUARTERLY

Vol 20 No 4

Q3 (Jul, Aug, Sep) 2015

Editor
Chris Cunningham
editor@petroleumtechnology.com
Production Editor
Rachel Storry
production@petroleumtechnology.com
Graphics Editor
Rob Fris
graphics@petroleumtechnology.com
Editorial
tel +44 844 5888 773
fax +44 844 5888 667
Business Development Director
Paul Mason
sales@petroleumtechnology.com
Advertising Sales Office
tel +44 844 5888 771
fax +44 844 5888 662
Publisher
Nic Allen
publisher@petroleumtechnology.com
Circulation
Jacki Watts
Louise Shaw
circulation@petroleumtechnology.com
Crambeth Allen Publishing Ltd
Hopesay, Craven Arms SY7 8HD, UK
tel +44 844 5888 776
fax +44 844 5888 667

PTQ (Petroleum Technology Quarterly) (ISSN

No: 1632-363X, USPS No: 014-781) is published
quarterly plus annual Catalysis edition by Crambeth
Allen Publishing Ltd and is distributed in the US
by SP/Asendia, 17B South Middlesex Avenue,
Monroe NJ 08831. Periodicals postage paid at New
Brunswick, NJ. Postmaster: send address changes to
PTQ (Petroleum Technology Quarterly), 17B South
Middlesex Avenue, Monroe NJ 08831.
Back numbers available from the Publisher
at $30 per copy inc postage.

ven at low ebb, the price of crude oil is a sore point for Indias refiners.
With crude accounting for around a third of Indian imports, it is a cause
of discomfort for the nation as a whole because it has to pay a premium
on the feedstocks it buys from the Middle East. The premium buys a guarantee
of supply and, with over 80% of its crude coming from OPEC countries, India
certainly needs such certainties.
Nonetheless, the premium is unfair, says the Indian government, and in reality serves to subsidise shipments of crude to OPECs more distant customers.
Instead of paying a premium, the nation should be enjoying the benefits of
customer loyalty in the shape of more favourable credit terms, for instance.
The low price of oil during 2015 has brought relief to Indias balance of payments deficit and a little more off the bill would do no harm to the nations
prospects for continued growth. Appeals to OPEC to remove the de facto surcharge have in the past made no progress. However, petroleums fast developing balance of geographical supply and demand took Indias Petroleum
Minister off to OPECs mid-year summit to press the case yet again.
The timing of Indias latest appeal for new terms could, at first glance, be
puzzling. Indias complaint is, after all, about a premium on crude at one of
the low points in its pricing history.
At the same time, the nations burgeoning refining industry is not only dedicated to meeting demand from its domestic market. The industrys growing
capacity is increasingly geared to exports of refined products to destinations
similarly targeted by OPEC members emerging fleet of mega-refineries. In
terms of balance of trade, it might be said that what Middle Eastern crude
exporters gave with one hand, Indian refiners might take from another.
However, as long as those producers continue to depend on their role as the
worlds biggest exporters of crude oil, Indias refiners are in an increasingly
strong negotiating position. The nation is expected this year to overtake Japan
as the worlds third largest consumer of crude. It processes its feedstock in
some of the worlds most advanced refining complexes. And rising import levels mean that India is a customer in an increasingly strong negotiating
position.
A further complication is the future position of Iranian oil exports, which are
embargoed as part of the sanctions imposed on Iran pending a general agreement on the development of the countrys nuclear industry. Should that dispute achieve resolution, OPEC members may have to deal with the matter of a
major competitor on its doorstep. In addition, Iran has a severely undernourished refining industry in need of external support. A future bilateral deal
between Irans and Indias respective producers and refiners may leave the
way open to very favourable terms for Indian imports of Iranian crude.
On this occasion, as before, OPECs members have remained intransigent on
crude exports to India. The organisation refused to budge when urged to drop
its output following the collapse in oil prices. Even heightened talk of crude
exports from the US fails to disturb the equilibrium, with Middle Eastern
exporters insisting that eastern US refiners will have to continue processing
their crude. In any event, the lowering of the Renewable Fuels Mandate in
Washington will mean that less domestic crude would be available for shipping out.
CHRIS CUNNINGHAM
PTQ Q3 2015

MOVE

Beyond

Limitations

Our scientists surpass the norm to create

catalysts that set the standard.
The next generation CENTERA catalyst is leading the industry in creating value all over the world.
CENTERA DC-2635 CoMo and CENTERA DN-3636 NiMo help refiners push beyond ULSD.
CENTERA DN-3651 and CENTERA DC-2650 drive FCC pretreatment to max performance.
CENTERA DN-3621 provides maximum hydrogenation at minimal nitrogen slip while processing
the toughest feeds.

Leading minds. Advanced technologies.

www.CRITERIONCatalysts.com

ptq&a
Q

Carryover from the amine unit is causing catalyst

contamination problems in our sulphur recovery unit. Is amine
directly causing this contamination and can we reduce the
foaming issues?

Berthold Otzisk, Kurita Europe GmbH, Otzisk@Kurita.de

Amine solutions contain many different compounds

such as heat stable salts, inorganic particles, metals
from corrosion attack, organic acids, cyanides, antifoam agents, and so on. Foaming is frequently
observed in amine systems, hence anti-foams are
needed. These defoamers are often silicone based products, which tend to coat metal surfaces, filters and
particles in the system. An unwanted carry-over of the
amine solutions also shifts these compounds into the
feed acid gas. Metals and silicone are well-known catalyst poisons, which can deactivate the active centres of
the catalyst.
The best foaming control to avoid amine carry-over
and amine losses is proper care of the amine solution,
which includes adequate filtration. The selection of the
defoamer should be done carefully because this can
have a significant impact on the quality of the treatment. While dimethyl polysiloxane defoamers can
increase the emulsification tendency of the amine, silicone esters such as Kurita E-702 are easily dispersed in
the amine solution. They do not create fouling or
adhere to activated carbon beds.

A W Jay Lechnick, Amine Process Specialist, UOP, A

Honeywell Company, William.Lechnick@honeywell.com
In this stream, amine can contaminate the sulphur
recovery unit. Even under normal operation, entrained
liquid droplets containing amine can be carried out of
the regenerator. To minimise entrainment, a properly
sized mesh blanket for collecting this liquid should be
installed at the top of the amine regenerator. Amine
condenses out of the vapour in the overhead
condenser, so operating the condenser at the lowest
possible temperature will also reduce contamination.
The lower condenser temperature only slightly
increases the required reboiler duty. If amine carryover is significant due to foaming, then the use of an
anti-foam agent is necessary. Pressure drop measurements across the regenerator will determine when
anti-foam is added. The pressure drop during normal
operation establishes a baseline for the unit. An antifoam agent is added whenever the pressure drop is
noticeably greater than the baseline value for an
extended period of time.
It is possible that the contamination is caused by

hydrocarbons that are co-absorbed by the amine in the

absorber. A large percentage of these hydrocarbons
enter the sulphur recovery unit, as they do not
condense in the regenerator overhead condenser. These
hydrocarbons are prevented from entering the sulphur
recovery unit using a rich solution flash drum
upstream of the regenerator. The hydrocarbon-rich
vapour from the rich solution flash drum is treated
and sent to the fuel gas header. The rich solution flash
drum is operated at the lowest allowable pressure to
minimise the amount of hydrocarbons that enter the
regenerator and, subsequently, the sulphur recovery
unit.

Q We suspect poor distribution of feed and catalyst in the

FCC riser as the cause of under-performance in the unit. Any
solutions to this problem?
A Cliff Avery, Global FCC Process Advisor, Albemarle,
Cliff.Avery@Albemarle.com
Selective FCC cracking is generally based on hydrocarbon vapour phase reactions. For the best performance,
an FCC unit usually should have a well-dispersed
vaporised feed that is in contact with a consistent catalyst density profile in the riser. Declining yields
(especially higher dry gas) can result when there is
poor feed vaporisation and dispersion. Non-uniform
regenerated catalyst to oil contacting will also affect
yields. The most common causes for improper feed
distribution include:
Damage to the feed nozzles (erosion, coking or wet
steam) or improper atomisation due to a change in
feed rate (velocities, vaporisation)
Change in feed quality (viscosity)
An improper feed injection angle.
Higher than design feed velocities and wet steam
will increase erosion. Coking occurs with heavier feeds
that do not vaporise well.
Periodic or regular pressure surveys of each nozzle
can identify if there is damage or plugging. Since FCC
feed nozzles are not retractable, the refiner generally
may continue to use the damaged nozzle(s) or increase
the feed to the other nozzles. Increasing feed to the
other nozzles, though, can increase the risk of erosion
due to higher velocities. Consult with your licensor to
identify erosion concerns and to identify the optimum
feed/steam ratio with the new operations. In addition
to erosion, the feed density profile may change if one
or more of the nozzles is blanked off. Besides observing higher dry gas, a higher delta coke may also occur.

Additional Q&A can be found at www.eptq.com/QandA

www.eptq.com

PTQ Q3 2015 5

Changing to a catalyst with lower delta coke can

counter the effects of the improper feed distribution.
Generally, the lower delta coke catalyst will have
lower rare earth or higher accessibility.

A Yorklin Yang, Global Process Technology Advisor,

Yorkin.Yang@BASF.com and Alexis Shackleford, Technical
Marketing Specialist, Alexis.Shackleford@BASF.com - BASF
Refining Catalyst
The riser mix zone in the FCC unit is the region where
the oil feed first contacts the hot catalyst from the
regenerator. The efficiency of this first oil catalyst
contact has a profound effect on the product slate from
the FCC unit. The performance of the mix zone is
affected by two main factors, namely feed atomisation
and the catalyst flow pattern prior to and after contacting with the atomised feed.
An ideal riser mix zone provides a homogeneous
cracking environment. Catalyst must be evenly distributed across the riser prior to contacting the oil feed.
The mix zone must rapidly bring the oil and catalyst
together, minimising any localised hot or cold spots in
the riser that can lead to sub-optimal product yields.

If the oil flow is restricted or

blocked to any one distributor,
more steam will flow through
that distributor
The catalyst and vaporised feed should flow up the
riser in plug flow, minimising any back-mixing and
ensuring that the feed only contacts freshly regenerated catalyst.
Proper feed injection is a key component of the reaction system and therefore optimum unit yields are
dependent on properly functioning feed nozzles. While
most advanced feed injection nozzles have proven to be
very reliable, there can be external factors that can lead
to problems with the feed injection system.
Poor distribution of feed into the catalyst or poor
distribution of catalyst into the feed can be crucial to
achieving optimum yields. To achieve optimum yields,
it is desirable to have uniform catalyst density in the
riser. If the catalyst is poorly mixed with the feed,
areas will be formed in the riser where there is either
too much feed and not enough catalyst and too much
catalyst and not enough feed.
If distribution of feed and catalyst in the FCC riser is
suspected as the cause of under-performance in the
FCC unit, the following actions should be considered
to improve or alleviate the conditions:
1. Increase or reduce feed dispersion steam if possible
The feed dispersion steam serves many purposes. It
provides the motive energy to generate smaller oil
droplets from the feed distributors, outlet velocity to
distribute the droplets into the catalyst and decreases
the hydrocarbon partial pressure at the feed injection

6 PTQ Q3 2015

elevation to increase the oil feed vaporisation rate. In

general, higher feed dispersion steam is desirable but
it is limited by its impact on erosion of the distributors
or internal riser refractory, riser vapour residence time,
main fractionator loading and overhead cooling duty
and sour water production.
2. Increase or reduce Wye or J-bend steam or lift gas rate
Steam or fuel gas may be injected into the base of the
Wye or J-bend sections. This can improve the distribution of catalyst flowing up to the feed distributors and
decrease the oil feed partial pressure at the feed injection elevation. As with feed dispersion steam, these
flows are limited by erosion potential, riser residence
time and main fractionator loading concerns.
3. Conduct pressure survey and review feed distributor
pressure drop for potential maldistribution
Most advanced feed distributor designs are pressure
balanced. The oil side supply pressure is dependent on
the pressure drop through the distributors. The pressure drop is primarily dependent on the steam flow
since it accounts for 70-80% of the total volume
through the distributors. If board mounted instruments are not available, a local, single gauge pressure
survey can determine if maldistribution of steam and
oil is occurring. If steam flow is restricted or blocked
to one distributor, this causes the controlled steam
flow to flow through one less nozzle. As such, the
pressure drop through those nozzles will be greater
and the oil flow will be lower to the open nozzles and
higher through the distributor with less or no steam. If
the oil flow is restricted or blocked to any one distributor, more steam will flow through that distributor since
there is no oil contributing to its pressure drop. As
such, less steam will flow to the other nozzles with
more oil. A pressure survey can be very useful
provided base readings are available when it is known
all distributors are open on both the oil and steam
side. This provides a reference pressure differential
between the steam and oil. If steam is restricted to one
or more distributors, then the steam supply pressure
will rise and the oil supply pressure drop and the
differential will increase. Conversely, if oil is restricted
to one or more distributors, then the steam pressure
will drop and the oil rise and the differential will
decrease.
4. Avoid feed nozzle plugging
While the most state of the art feed system is designed
to minimise problems, unforeseen circumstances in the
unit can occur that could potentially lead to a problem
with the feed injection system. While this is very
unusual, a thorough understanding of the potential
problems that can arise, and how to detect them if
they occur, may help minimise the time spent operating with a problem.
Nozzle plugging is very detrimental to nozzle
performance. For example, plugging of the internal
steam orifices decreases the steam flow to the plugged
nozzle while simultaneously increasing the steam flow

www.eptq.com

COMPLETE SOLUTIONS
FOR YOUR REFINERY
OPERATIONS
Whether youre dealing with tight oil, more stringent sulfur
limits or changing feedstock supplies, CB&I has the answers
to help refiners derive maximum value from every molecule.
Were with you through every stage of the process plant life
cycle, from feasibility studies through technology selection,
full-scope EPC, commissioning and start-up, to plant
optimization and upgrades.
Our broad portfolio of both refining and petrochemical
technologies, combined with our execution expertise, will
help you maximize unit flexibility and achieve margin benefits
in the widest range of scenarios.
PROCESS PLANNING AND DEVELOPMENT
LICENSED TECHNOLOGIES AND CATALYSTS
FULL-SCOPE EPFC SERVICES
PROJECT MANAGEMENT AND CONSULTING
AFTERMARKET SERVICES

A World of Solutions
Visit www.CBI.com

cbi_ptq_3q2015_ad_may_2015_press.indd 1

5/18/2015 3:57:44 PM

es
ot
N
Process
This exchanger is a victim of its design.

Opportunity or Annoyance?
Price differentials between conventional and opportunity crudes compel refiners to process increasing percentages of lower valued opportunity crude. However,
as many refiners have learned the hard way, opportunity crudes are tied to unique processing challenges.
Furthermore, existing crude unit configurations may
limit high-profit opportunity crude to a disappointingly
small proportion of the total unit blend.
Processing a changing slate of opportunity crudes of
varying compositions requires a CDU/VDU design that
is flexible, reliable, and commercially proven. A good
design must control chronic problems associated with
many of these crudes including corrosion, exchanger
fouling, tray plugging, vacuum heater and wash bed
coking, and unreliable product quality prediction. CDU/
VDU designs for challenging crudes should be based
on proven best-practices rather than simply on a lowCAPEX strategy. For example, a poorly designed heat
exchanger may operate well initially, but high fouling will quickly lower efficiency and eventually limit
throughput. Up front investment in engineering experience and know-how pays dividends when units meet
capacity and run-length targets.

Over the course of 20 years, Process Consulting Services has completed more than 130 revamp and grassroots designs supported by over 75 detailed test runs.
An extensive collection of test run equipment performance data and feed/product analyses enables confident prediction of real-world opportunity crude performance. PCS has enabled refiners worldwide to extend
crude unit run lengths from months to years while improving yields and operability.

Test run feed and product samples

3400 Bissonnet St.

Suite 130
Houston, TX 77005, USA

+1 (713) 665-7046
info@revamps.com
www.revamps.com

to the other nozzles. As a result, more of the oil flows

to the plugged nozzle in order to balance the oil side
pressure drops. Pressure drop calculations indicate that
up to half of the total feed will flow through a nozzle
with totally plugged steam orifices. The end effect is a
large amount of feed that enters the riser without being
finely atomised. The effect on gasoline yield and
conversion can be significant.

Oil side plugging diagnosis

Plugging on the oil side has been relatively infrequent. The most reliable indicator of oil side plugging
is an increase in oil header pressure. A sudden jump
in oil header pressure is typically an indication that
some debris has entered with the oil (for instance,
coke from a fired heater) and lodged in one or more
of the leads/feed nozzles. In most cases, the debris is
trapped by one or more of the check valves in the oil
leads. The pressure gauge downstream of a plugged
check valve reads noticeably lower than the oil
header. (Proper functioning of the pressure gauge in
the oil lead should be verified prior to taking any
corrective action.) Coke can also become lodged in the
nozzle itself, either in the internal feed pipe or the
nozzle discharge cap, forming a restriction in the
outlet slots of the cap.
The feed furnace is typically the culprit of coke
formation in the feed system. For this reason, it is
recommended for units with a feed furnace to consider
installing a strainer downstream of the furnace to
prevent from plugging the feed nozzles. Another
potential type of plugging that can occur is due to
catalyst backing up into the nozzles because of a
complete loss of steam and oil flow. This situation
would typically only occur during a shutdown where
steam was completely stopped with catalyst still in the
riser. This is why steam to the nozzles should never be
stopped when catalyst is in the unit. If plugging on the
oil side is suspected, and none of the gauges in the oil
leads read low, the best method to determine which
nozzles might be plugged is to block in the oil to one
feed nozzle at a time. Shutting in the feed to one
nozzle while maintaining a constant feed rate causes a
jump in the oil header pressure. The nozzles that are
plugged will cause a lower increase in header pressure
(less oil flow) than the nozzles that are open (more oil
flow). If a nozzle becomes completely plugged at the
cap, the steam will back pressure up the oil side to the
check valve, which will be indicated by a high oil lead
pressure gauge.
Steam side plugging diagnosis
Since steam is the source of most of the feed nozzle
pressure drop, a relatively small maldistribution of
steam causes a much larger maldistribution of oil. As a
result, feed is poorly atomised and unevenly distributed across the flowing catalyst stream. Fortunately,
relatively simple measures are available to prevent this
plugging/maldistribution. In addition, monitoring for
maldistribution and taking corrective actions, should it
occur, is also relatively simple.

www.eptq.com

In most cases, steam side plugging can be traced to

carry-over of dissolved solids or boiler feed water
treatment chemicals from a maloperating steam boiler.
Plugging of the nozzles steam orifices can be diagnosed by monitoring various system pressures and
pressure drops. Some indicators of steam side plugging are listed below:
Increasing pressure differential between steam and
oil headers
Increasing steam header pressure at constant steam
flow rate
Increasing steam control valve opening at constant
steam flow rate.
Falling oil header pressure at constant steam and oil
flow rates Plugging of the steam strainers due to boiler
feed water chemicals or pipe debris can easily be indicated by a lower steam lead pressure compared to
header pressure.

Plugging of the nozzles steam

orifices can be diagnosed by
monitoring various system
pressures and pressure drops
Use quality steam
Past experience indicates that carry-over of water from
the steam boiler into the dispersion steam often leads to
nozzle orifice plugging and the associated maldistribution. Apparently, the presence of hot feed vaporises the
water droplets, leaving behind deposits of dissolved
solids that partially plug the internal steam orifices.
Further reading
1 Miller R, Yang, Y Gbordzoe E, Johnson D, Mallo T A, New developments
in FCC feed injection and stripping technologies, AM-00-08, NPRA 2000
Annual Meeting.
2 Kalota S A, Rahmim I I, Solve the five most common FCC problems,
AIChE Spring National Meeting, 30 March-3 Apr 2003.

Q What developments are there in mild hydrocracking

catalysis to improve our ULSD production?
A Stefano Melis, Global VGO Application Specialist,
Albemarle, stefano.melis@albemarle.com
There are mainly two areas of catalyst development to
improve diesel production from existing assets: T95 shift
operations in ULSD units and MHC in VGO HT units.
For T95 shift operations, we feed the ULSD units
with heavy gasoil (HGO) and/or light vacuum gasoil
(LVGO). Processing of such feedstocks with conventional hydrotreating catalyst would invariably lead to
off-spec product in terms of both T95 distillation and
density. We can efficiently improve the product quality
by combining high-activity pretreatment catalysts with
selective cracking catalysts. Since the addition of cracking catalyst will come at the expense of less

PTQ Q3 2015 9

es
ot
N
Process
Crude units can be designed for reliability.

Maximize Reliability in
Grassroots Crude Units

DESALTING

Crude unit operators are far too familiar with a long list
of crude unit reliability problems including fouling in
heat exchangers and fired heaters, poor desalting, corrosion of piping and equipment, and coking in the vacuum column wash zone. Many millions of dollars have
been spent fighting these problems, yet they continue
to force unplanned shutdowns with depressing regularity.
Revamps must address reliability issues, but project
scope is hindered by the limitations of existing equipment. Grassroots design of crude and vacuum units
presents an opportunity to get everything right the first
time. Here are a few tips for designing a reliable and
profitable crude/vacuum unit.

HEAT EXCHANGER AND HEATER FOULING

High velocities in heat exchanger tubes produce high
shear at the walls, preventing foulants from accumulating. High shell-side velocities, coupled with exchanger designs that minimize dead zones in the flow,
eliminate shell-side fouling. In fired heaters, high mass
fluxes maximize wall shear, shorten residence time,
and lower wall film temperatures, all of which reduce
coking. Furthermore, reliable heaters must have correctly sized burners with proper burner-to-burner and
burner-to-tube spacing.

Desalter size is highly dependent on crude blend due

to dramatic variation in required centerline velocity.
A unit must be designed with the flexibility to carefully control desalter temperature, which can range
from 110C to 150C, by shifting heat from upstream to
downstream of the desalters. Vendors are often judged
on cost alone, which results in minimum sizing for the
design crudes and rates. Carefully consider whether
long-term crude trends will soon render these desalters inadequate.

CORROSION
In grassroots design, be realistic about metallurgy. Because modern refineries do not run a steady diet of the
same crude, consider the sulfur and TAN numbers of
potential crudes outside the units design blend. Chronic corrosion issues, or the inability to process high-margin opportunity crudes, will quickly overshadow the initial savings from choosing too low of a metallurgy.

VACUUM COLUMN WASH ZONE COKING

Wash zones are not for fractionation, they are for
de-entrainment! Pursuing fractionation efficiency by
specifying a deep bed with small crimp packing is a
recipe for rapid coking. The correct choice of packing
combined with the right wash rate and good distribution will properly de-entrain while preventing coke formation.

3400 Bissonnet St.

Suite 130
Houston, TX 77005, USA

+1 (713) 665-7046
info@revamps.com
www.revamps.com

pretreatment catalyst, we need to ensure that the unit

is still able to obtain the same performance in terms of
HDS and HDN. This is usually possible by utilising
the latest-generation pretreatment catalysts or resorting
to bulk metal catalysts, such as Nebula. We also need
cracking catalysts that selectively convert only the tail
of the feedstock in such a way that naphtha production is minimised. This requires an in-depth
knowledge of feed composition so that the most suitable cracking function is individuated. The wide
variety of available cracking catalysts can then be used
to selectively crack the molecules specifically present
in the tail, solving thus both T95 and density
constraints.
Another method of increasing diesel production is to
increase the conversion in FCC-PT units. New VGO
catalysts with enhanced HDS activity and stability can
permit increased severity of the operations, automatically leading to an increased conversion to diesel. Also,
when applying the new VGO catalyst, it can become
possible to achieve the required target with less catalyst volume, leaving room for the introduction in the
loading profile of an amorphous MHC conversion
catalyst. This ought to further enhance diesel
production.

When applying the new VGO

catalyst, it can become possible
to achieve the required target
with less catalyst volume
Q How do recombination problems involving a hydrotreater
effluent stream and H2S occur and what effect can it have on
ULSD product specification?
A

Don Wolfe, Separation Technology Manager, Baker

Hughes, don.wolfe@bakerhughes.com

Recombination is the process by which hydrogen

sulphide (H2S), formed during the reduction of other
sulphur species in the hydrotreater, reacts with olefinic
material in the reactor effluent to form mercaptans.
The resultant mercaptans, while less reactive than the
H2S, are more difficult to remove from the distillate
and can impact other ULSD product quality specifications in addition to possibly raising the total sulphur
content.
Conventional desulphurisation wisdom would
dictate that, for a given operation, increasing reactor
temperature will result in greater desulphurisation efficiency and lower total sulphur in the reactor effluent.
This maxim still holds true to a point; however, as the
heater transfer temperature increases, the total sulphur
content of the effluent reaches a minimum then begins
to increase. This degradation in total sulphur reduction
is brought about as mercaptans are formed from the
newly formed H2S and olefins in the effluent. Reactor

www.eptq.com

space velocity plays a similar role in that increasing the

space velocity can impact total sulphur and mercaptan
sulphur concentration.
From a fuel quality perspective, the conversion of
hydrogen sulphide to mercaptans adds a number of
difficulties. The most significant is that H2S, the very
compound that was deliberately created in order to
easily strip it from the reactor effluent, has been reintroduced to the fuel as a reaction product:
mercaptans. Due to the solubility of the mercaptans
and their resistance to stripping, their presence can
cause the distillate fuel to exceed the 15 ppm sulphur
ceiling. Additionally, mercaptans are a reactive
sulphur species and can potentially cause the fuel to
fail the copper corrosion specification (ASTM D130)
as well as negatively impact long term storage
stability.
Options for mitigating the impact of recombination
would be to (1) operate the hydrotreater at the lowest
possible temperature to achieve the maximum desulphurisation and minimise mercaptan formation or (2)
operate to saturate olefins. Because the literature
suggests that methyl mercaptan forming reactions are
more thermodynamically favourable than heavier
mercaptans, in some scenarios, the resultant increase in
methyl mercaptan may be amenable to treatment with
aqueous mercaptan scavengers. The scavengers react
with the mercaptan to form a water-soluble reaction
product that is then decanted from the ULSD, thereby
reducing the total sulphur content and improving
overall ULSD quality.

Q Can reducing atmospheric tower top temperatures

to maximise middle distillate production introduce new
corrosion problems. If so, how are they overcome?
A Joel Lack, Corrosion Technology Manager, Baker Hughes,
Joel.lack@bakerhughes.com
Corrosion problems in a tower top section with upper
pumparound or reflux circuits can be caused by the
condensation of water, formation of corrosive salts, or
the introduction of these to the tower via free water in
an overhead reflux stream. Corrosion caused by a wet
overhead reflux will be present regardless of the tower
top temperature and is overcome by methods that
succeed in drying the overhead reflux stream.
However, the condensation of water or the formation
of corrosive salts in the tower top is partly dependent
on the tower top temperature. Both can form when the
tower temperature is lowered below specific temperatures dictated by the vapour pressure of water or the
contaminants reacting to form the salt.
The only sure way to control corrosion is to maintain
the operating temperature above the temperature at
which these corrosive conditions form. Simulation
methods that accurately calculate dew point and salt
formation temperatures can be used to maximise
temperature reduction without increasing corrosion
risk. Experience has shown that operating the tower
top temperature at least 15C above the highest corro-

PTQ Q3 2015 11

sive phase formation temperature should overcome

local cooling at the reflux return on the top tray.
Additional process optimisation can be achieved
through methods that lower the corrosive phase
formation temperatures. Water dew point can be
lowered by better dehydration of desalted crude, lower
steam rates, or use of a prefractionation tower. Salt
formation temperatures can be lowered by decreasing
overhead chlorides through optimised desalter operation or caustic addition to desalted crude. Ammonium
chloride salt formation temperatures can be reduced
by keeping ammonia levels in stripped sour water
used for desalter wash at a minimum. Formation
temperatures of salts from neutraliser amine recycle or
boiler treatment amines can be lowered by switching
to amines that form salts at lower temperatures.
Additionally, formation temperatures of salts from
contaminant amines can be reduced by reducing the
source (such as contaminated crude or amine unit
slops) or reducing their concentration in the desalted
crude by lowering the pH of the desalter.
Sometimes, the financial rewards of lowering operating temperatures are great enough to entice refiners to
accept increased corrosion risk in the crude tower.
Many refiners have responded to tower top corrosion
with corrosion-resistant alloys, although even the best
materials available today will eventually succumb to
salt corrosion. Salt dispersant chemicals have been
used in numerous cases, but these have only shown

effectiveness against fouling and have had little impact

on corrosion. Chemical treatments designed to reduce
the corrosiveness of salts in towers are new to the
marketplace, but have so far shown promise in reducing the corrosion impacts of salt deposition.

Q We believe arsenic contamination may be the cause of

loss of activity in our diesel hydrotreater. How do we address
this?
A Robert Bliss, Mid-Distillates HPC Specialist, Albemarle,
Robert.Bliss@Albemarle.com
Arsenic concentrations in crude oils vary significantly
across the globe and crudes containing high levels
(>300 ppb) are typically associated with specific
geographical production regions. Some of the regions
known to have the highest concentrations of arsenic
include the Canadian Alberta Oil Sands bitumen
deposits, the Venezuelan Orinoco Belt extra heavy
crude deposits and Russian crudes from the Western
Urals. Theories suggest petroleum plays derived from
vast marine animal deposits are typically high in arsenic due to arsenic concentrating in marine species
thousands of times above the 1.5 ppb level typically
found in ocean water.1
Arsenic compounds found in petroleum are primarily derived from aryl and alkyl substituted arsine.
These organoarsenic compounds can vary greatly in

Our experience Your safety

400
- DN 1 0
6
5
N
3
D
- PN
6
1
N
P

In gas technology operations,

quality and safety standards are
the highest priority during planning, construction and operation.
Bhmer ball valves are continuously developed and adapted to
the ever changing conditions of
practical use.
3
3
3

fully welded/split body

meets the relevant standards
(e.g. EN 13774, EN 14141)
applicable in different areas:
in the gas storage, pipeline
construction, in transfer and
compressor stations etc.

BHMER GmbH
Gedulderweg 95, D-45549 Sprockhvel
http://www.boehmer.de, eMail boehmer@boehmer.de
Tel. +49 (0) 23 24 / 70 01-0, Fax +49 (0) 23 24 / 70 01-79
Bohmer NEU Gas GB 125x178 06-2015.indd 1

12 PTQ Q3 2015

01.06.15 08:50

www.eptq.com

reductions
in
regulations
willasbeatemperatures.
considered,
structure
andozone
molecular
weight;
result they
tower
temperature
and
de-sublimation
which
create
greater
uncertainty
in
terms
of getting
distribute
evenly
across
all
petroleum
distillate
fracChemical mitigation plans include use of Dorf
Ketals
permits
and
will
increase
compliance
costs
to
US
tions
(such
as
naphtha,
diesel,
VGO)
and
negatively
proprietary amine removal aids and top pumparound
business.
impact
hydroprocessing
units
across
all
application
filming corrosion inhibitors. When one operates at a
segments
(NHTs, DHTs,
HCUs).
These
and
others
US refiners
to
low
top regulations
temperature,
cold
refluxleave
temperature
further
Arsenic
is
one
of
the
most
severe
poisons
for
base
face
billions
of
dollars
in
investments
and
costs
(EPAs
reduces. Ensuring reflux is as dry as possible using
metal (NiMo,
CoMo)
hydroprocessing
catalysts
estimates
arehelps
that
ozone
NAAQScorrosion
regulations
alone(HPC).
drier/filters
in improving
control
inwill
the
Studies
have
shown
activity
losses
exceeding
50%
cost
as much
annually) due to newat
top section
withasa $100
lowerbillion
top temperature.
arsenic
loadings
as low asvery
0.8 wt%
catalyst. This is
rules
that
will provide
littleon environmental
A
very
severe
compared
to
a
moderate
poison
benefits.
list APT
of regulations
will
hurt
moreas
Celso This
Parejo,long
Manager
Refinery Applications
& such
Process
silicon
where
catalyst
activity
losses
do
not
exceed
50%
than
our industry,
they willUSA,
hit celso.pajaro@sulzer.com
every American citizen
Technology,
Sulzer Chemtech
until
loadings
in
excess
of
8
wt%
on
catalyst
are
realin
and otherthe
costs,
and put
In terms
order of
to higher
reducefuel,
top food
temperature,
column
will
ised.
Arsenic
isreflux.
a permanent
poison;
it is on
the
our
nation
at a competitive
require
a higher
Thisdisadvantage.
reduces
theonce
residence
time
catalyst
it
cannot
be
removed
and
it
will
prevent
catarecently and
alarming
is the
climate
deal
that
inMore
the overhead
drum
which can
lead
to poor
separalystofregeneration
at loadings
as
low
as 0.3
on
President
made
China
late
lastwt%
year.
tion
theObama
water from
thewith
naphtha.
In this
case,
water
catalyst.
Under
this over
agreement,
emission
ChinaOnce
will
can carry
with the
refluxlevels
to thefrom
column.
Arsenic
bondswill
directly
with
continue
unabated
for
the next
16 nickel
years,
essentially
inside,
the
water
vaporise,
leaving
theand
saltscobalt
that
promoter
atoms
on molybdenum
doubling
existing
emissions.
the same (MoS
time,
can
createtheir
under-deposit
corrosion.Atdisulphide
Coalescing
inter-2)
HPC
active
sites,reduce
forming
nickel
arsenides
(NiAsy)
we
will
further
carbon
emissions
by
nals
like
Mellaplate
have our
been
successfully
used
in
2025
by at least
26%, and
as accumulators
much as 28%,toover
2005
atmospheric
overhead
drum
improve
levels.
the water-naphtha separation.
Any
efforts
on the corrosion
part of theproblem
US to meaningfully
Another
potential
occurs when
the column temperature is low enough for water to
condense, forming concentrated acids such as HCl,
creating a highly corrosive environment. Often this
problem can be mitigated with the use of corrosion
resistant tray metallurgy such as Zeron 100, AL6XN,
254 SMO, Monel, Alloy 825, C-276 (Hastelloy), 625
(Inconel), C-22, and so on.
and cobalt arsenides (CoAsy). Once formed, these
compounds do not destroy the MoS2 strucAarsenide
Charles Radcliffe, Senior Tech Service Engineer-Europe,
ture, however they do alter its electronic state,
Johnson Matthey Process Technologies, Charles.Radcliffe@
resulting
in substantial
catalyst
deactivation.
Further,
impact
global
emissions will
be futile
if other nations
matthey.com
it
has
been
shown
that
arsenic
has
a
higher
affinity
do
not crude
simultaneously
their emissions.
Lower
tower topreduce
temperature
will move For
the
towards
nickel
than
cobalt.
Thus
catalysts
designed
example,
by the
2030,
India
will have
population
point where
acid
condensate
dewa point
occurs, that
and
with larger amounts of nickel will trap larger
exceeds
The
is shared
and
you willthat
needoftoChina.
identify
thisatmosphere
and check the
metallurgy
amounts of arsenic.
undeniably,
of Chinaand
cannot
holdamine
back
is suitable, the
andGreat
that Wall
neutralising
filming
Because catalyst deactivation from arsenic poisoning
emissions
produced
Beijing.
allowed
to main
proceed,
it
injection
points
are in
before
thisIfpoint.
The
probis so severe, it is critical in hydroprocessing applicacould
shut be
down
the dew
US economy.
lem
will
if
the
point
is
actually
moved
to
tions known to have arsenic contaminated feedstocks
Oil and
gas amine
have been proven
to raise able
the
within
the natural
tower.
The
to employ
an arsenic
specificsuppliers
trapping should
catalyst be
layer in
living
standards
of millions,
even Switch
billions,toaround
the
tothe
help
with
analysis
and
advice.
caustic
for
top bed guard system to protect the main bed HPC
world,
advancing
public
health
and
saving
andif
overheads
corrosion
control.
This
is
not
advisable
below. Albemarle has extensive experience in the
lengthening
lives
in FCC
the process.
Campaigns
intent the
on
feeding
to the
because
willour
increase
design resid
and application
of
arsenicthis
traps;
two latest
eliminating
their
usefeed.
will only hurt our poorest citizens
sodium
in
the
FCC
arsenic traps include KG 8, commercialised in 2011,
here at home and those in developing countries where
and KG 16-MAC (Maximum Arsenic Capture),
affordable food and fuel are badly needed. The goal
in 2014. The best way to determine
Qcommercialised
should
be CO
to raise
the living
standards
of the in
worlds
The
content
our SWS
gases results
a high
2
which
arsenic
trap ofand
the quantity
required
to
population
rather
than
inhibiting
real
progress
through
reagent
turnover
in the
caustic scrubber.
Solutions
address
the loss
of activity
in your unit
is to please.
utilise the
costly, counter-productive programs. Advancing
knowledge of your Albemarle technical
Aapplication
prosperity
for the betterment
of mankind
be the
Marcello
ITW, should
mferrara@itw
service
team toFerrara,
design President,
a catalyst system
solution that
primary
basis
upon
which
regulatory
measures
should
technologies.com
will exceed your objectives. Albemarle also offers
be
morallyturnover
justified.
Reagent
depends
basically
stoichiometry
analytical
services
to quantify
the on
level
of arsenic
Left
unchecked,
President
legacyiswill
be
and
scrubber efficiency.
WhileObamas
stoichiometry
related
contamination
in your
feedstock.
one
of
higher
energy
costs,
a
weakened
economy,
the
to gas composition only, operational scrubber effielimination
of jobs, and
lost competitive
ciency is affected
by afouling.
Fouling advantage.
of caustic
We
must
embrace
the
reality
of
our
energy
abundance
scrubbers is quite common, as the
composition
of
and
use it to
enhance
our economic
fortunesunsaturated
as well as
treated
gases
normally
contains
References
those
who
tocan
a similar
standard.
Unfortunately,
compounds.
ITW
helparsenic
solve
fouling
of
1 Irgolic
K J,aspire
Puri
B K,
Organic
compounds
inproblems
petroleum and
this
presidents
rhetoric
and
ideology
are
completely
caustic
scrubbers
by implementing
Cleaning
natural gas
metal speciation
in the environment, Online
NATO ASI
Series,
Vol 23,
unaligned
andthe
in equipment
direct conflict
energy
and
cleaning
in aswith
littlerational
as 24 hours
on
1990,
377-389.
policy
and reality.
a process-out/process-in
basis.

Feed to the col

Is your plant
ready for the future?

Top of the colu

Reflux drum
controller. Thi
columns oper
pressure chan
amount of fe
almost no reflu
amount of refl
valve which is
ler (LC). The
controlled with
ture indicator.
top temperatu
with a reflux
this study is
developing the

Studies have shown activity

losses exceeding 50% at arsenic
loadings
as low asoutcomes
0.8 wt% for
Future
regulatory
on refining
catalyst and petrochemical
the
manufacturers industries remain
unpredictable

www.eptq.com
www.eptq.com
www.eptq.com

The feed to t
other units. Th
11th tray by m
tant source of

Separation Ind

Reboiler duty
the
column
liquid rate. Th
sharpness of se

As the dr
the gas p
heavy en

Effective cut po

This feature is
column. The
concerned wit
pressure decre
ends and gas fl

Life-cycle optimalization

Proposed appr
Making critical plant information fully visible is just the
beginning of the vigilant cycle. Seeing clearly gives
you the knowledge necessary to anticipate the changes
required in your process. Knowing in advance brings you
the speed and flexibility to optimize your plant in real time.
And by acting with agility, you are able to adapt to the
ups and downs of your business environment. VigilantPlant
excels at bringing out the best in your plant and your
people - keeping them fully aware, well informed, and
ready to face the next challenge.

Please visit us at www.yokogawa.com/eu

or contact us at info@nl.yokogawa.com

Control_86x270_2014.indd
32 PTQ Q3 2014 1

In this sectio
explained in d
mended appro
similar process

Base layer enh

Here, the cont

fix the main
solved and ap
ters (Proporti
decided. In ad
taking into acc
column.

Data collection

The data for t

month before i
The control
these control e
PTQ Q3 2015 13
06/03/14
11:36
PTQ Q1
2015 11
PTQ Q2 2015 13

ITW

Innovative
Technologies
Worldwide

CHANGE YOUR MIND

ON FOULING REMOVAL

ARE THESE IMAGES FAMILIAR TO YOU ? FORGET ABOUT THEM !

Patented ITW Online Cleaning can remove any type of fouling from equipment, including the polymers, without the need to
open or enter hazardous process equipment.
An entire Process Unit can be cleaned by utilizing ITW Online Cleaning in as little as 24 hours on an oil-to-oil basis.
Online Cleaning can be applied at any time during the run of the Unit, in order to solve the problems when they start
appearing, rather than when they are no longer sustainable.
This will in turn avoid throughput reduction, giveaway and energy loss, associated with fouling.
The application of ITW Online Cleaning will be therefore driven by performance recovery and Opex improvement, rather than
the economics for placing a turnaround.

ITW Online Cleaning can be applied to all Refinery/Petrochemical/Gas Field/Oil Field production Units, as
well as storage tanks, and is the only technology which has an immediate Return On Investment (ROI).
Regular application of ITW Online Cleaning will target an increased run length under clean conditions, with a far higher ROI.
For Turnaround applications, ITW Online Cleaning can eliminate/dramatically reduce the need for mechanical cleaning,
thereby reducing downtime and improve operational HS&E.
In a Turnaround application, an additional ROI increase can be targeted by applying ITW Improved Degassing/
Decontamination to achieve quick and effective safe entry conditions.
Our patented chemistry does not create any emulsion and fluids can be easily handled by Waste Water Treatment Plant.

EVALUATE ITW ONLINE CLEANING AND ITW IMPROVED DEGASSING/DECONTAMINATION TODAY TO

IMPROVE YOUR PLANTS PROFITABILITY !
For more informations contact :

ITW S.r.l.- C.da S.Cusumano - 96011 Augusta - Italy

Tel. +39 (0931) 766011
E-mail: info@itwtechnologies.com
www.itwtechnologies.com

Now hiring
Professionals
Worldwide

Join ITW Team worldwide and send your Curriculum Vitae to : jobs@itwtechnologies.com

itw.indd 1

05/06/2014 19:38

Managing fouling in refinery networks

Modelling heat exchanger and other networks enables the use of dynamic fouling
models in performance assessment and predictive studies
SIMON PUGH and EDWARD ISHIYAMA
IHS Energy

ouling of crude and product

streams in oil refineries is an
inescapable
problem
that
causes very significant financial,
environmental and safety problems.
For example, refinery preheat train
(PHT) networks nearly always
suffer from fouling, resulting in
reduced throughput, increased
energy consumption and more
frequent shutdowns. The financial,
environmental and safety impact of
fouling on processing 73 million
barrels of crude oil per day worldwide is enormous. It is estimated
that crude fouling in PHTs costs
0.25% BOE of all refined crude, or
66 million barrels per year.
Prediction of fouling is an
increasingly challenging task, with
laboratory fouling studies able to
assess blend compatibility but often
little more than a first approximation of real plant performance.
Refinery networks such as PHTs
often exhibit fouling that builds up
relatively slowly and is often
managed by exchanger cleaning or
improved exchanger design, as well
as rapid fouling events. It is imperative for effective refinery operation
that rapid events are flagged up
quickly and this can only be done
where the fouling inside individual
shells can be inferred correctly from
plant measurements. These measurements are often inadequate for
spreadsheet-type analysis, where
the geometries of the individual
exchangers are in any case
over-simplified.

Pragmatic framework

A
comprehensive
research
programme into crude oil fouling
was completed in 2010 in the UK.

www.eptq.com

This programme was facilitated by

IHS and guided by the majority of
the super-major oil companies plus
several technology providers. The
group identified that a pragmatic
framework to apply the research
knowledge to networks such as
PHTs was needed. SmartPM (Smart
Predictive Maintenance) is the software tool that fulfils this purpose,
and is now assisting refinery operators, engineers and managers to
make decisions on how to alleviate
fouling in existing and new plant.
In addition to greatly improved
reconciliation of monitoring data
that can flag up rapid fouling
episodes,
the
detailed
heat
exchanger modelling in SmartPM
allows dynamic models for crude
blend and product stream chronic
fouling to be inferred from the
reconciled data. These models allow
operators to compare capex-heavy
solutions (revamps and so on) with
opex routes (such as regular cleaning). Using detailed (rigorous)
thermohydraulic exchanger modelling and dynamic fouling models,
SmartPM models correctly the
behaviour of networks of heat
exchangers. The wide range of
generic and proprietary heat
exchanger configurations, including
arrangements such as helical baffles
and tube inserts, is included.
Within a state-of-the-art, simpleto-use graphical interface, SmartPM
is a turnkey application that
connects operation, maintenance,
engineering and fouling research,
with each group accessing the
appropriate facilities of the software for the same refinery model.
For example, whilst operation and
maintenance run data reconciliation

and cleaning scheduling, engineering costs the benefits of revamp

options, researchers review the
fouling propensity of various crude
slates, perhaps in a range of refineries. All share the same goal, which
is maximising profit through
increased efficiency and reduced
maintenance.
SmartPM is a heat exchanger
network simulation software aimed
at tackling fouling related problems
encountered in oil refining. The
software couples thermohydraulic
network analysis with dynamic
fouling simulations, enabling the
user to reliably predict plant operational
limitations,
economic
penalties
and
environmental
impacts
resulting
from
heat
exchanger fouling. The software
was designed (but not necessarily
limited) to model heat exchanger
networks in crude oil refineries.

Missed opportunities

Many refineries use spreadsheet

based models or other approximate
models to evaluate plant operational data and assess heat transfer
efficiency.
These
rudimentary
models are often used for candidate
selection in exchanger cleaning.
There are many missed opportunities when using these types of
model as they fail to resolve key
issues:
Many heat exchanger networks
are poorly instrumented, with
important temperatures and even
flow rates missing, or at least
temporarily unavailable.
Intermediate
temperatures
between shells in series are rarely
measured yet temperature fields in
these shells are very different;

PTQ Q3 2015 15

Figure 1 Schematic of a case study preheat train. Label E denotes heat exchangers. FIT
and FOT represent furnace inlet and outlet temperature, respectively

hence the fouling in the shells

differs signicantly.
An increasing number of reneries are now using efcient bafe
types such as Helixchangers and
EMbafes; tube inserts such as
Turbotal and Spirelf are in frequent
use.
Online exchangers compensate
for loss of overall heat duty when
other exchangers are taken ofine
for cleaning.
Cleaning hottest shells can lead to
a loss of preheat in upstream shells
as the temperature of streams (for
example, streams such as vacuum
residue are often bridged between
exchangers
downstream
and
upstream of the desalter, where
temperature variation propagates
throughput the network when associated units are cleaned).

Increase in pressure drop with

fouling can result in throughput
reduction.
Hot end exchangers can experience crude-side boiling, which can
lead to throughput reductions.
SmartPM resolves these and
other issues by using full network
simulation,
detailed
exchanger
modelling and dynamic fouling
models in all modes of operation
from data reconciliation to cleaning
scheduling.

Application to preheat trains

A key application of the software is

in the modelling of the crude
preheating system. Fractional distillation of crude oil is an energy
intensive process. The process
requires the crude to be heated from
ambient temperature to around

8D

Fouling rate,
h.ft2.F.BTU1.h1

2.8
2.4

9A

8C

2.0

9DE

1.6

8B

9BC

1.2

8A

0.8

3
5

0.4
0
280

7
320

360

390

430

8 Crude-side velocity,
460

500

540

Crude-side film temperature, F

570

ft.s1

Figure 2 Comparison of fouling in exchangers downstream of the desalter. The surface

indicates a fit of a dynamic fouling model

16 PTQ Q3 2015

700F (370C). The required heat is

provided through a network of heat
exchangers called preheat trains
(PHTs) and furnaces. PHTs recover
heat from the product and pumparound
streams
of
the
distillation column. A simple example of a PHT based on an
Argentinian renery is shown in
Figure 1.1 Crude is pumped through
a set of heat exchangers and enters a
desalter. The desalter washes the
crude with water to remove the
inorganic, water soluble impurities
in the crude. The vapour component
of the crude is removed through a
ash column. Crude is further
heated downstream of the ash
through a set of heat exchangers
where it enters the furnace at the
furnace inlet temperature (FIT). The
crude is then heated in the furnace
to the furnace outlet temperature
(FOT), prior to entering the fractional distillation column.
Crude oil is a complex mixture of
petrochemicals and impurities and
is prone to fouling.1,2,3 US reneries
are reporting new fouling problems
that are a direct consequence of the
increasing amounts of domestic
tight oils in the crude slate. These
problems arise from the changing
nature of the slate (one renery for
example is seeing swings upwards
of 5-10 API points on a day-to-day
or week-to-week basis), additives in
the crudes and crude incompatibility issues.
Many tight oils have added
corrosion inhibitors (usually amine
based) and drag reducing agents.
This can lead to operating issues in
the
preheat
exchangers
and
furnaces, and in the atmospheric
distillation towers where the amine
salts are deposited, inhibiting heat
transfer and fouling the tower.
Increased deposition of sand and
particulates are observed in heat
exchangers
upstream
of
the
desalter. In usual PHT operation,
chemical reaction fouling dominates the fouling mechanism
downstream
of
the
desalter.
Operational irregularities such as
inefcient desalting can lead to
inorganic deposition. Such problems are exacerbated as tight oils
stabilise emulsions in the desalter
and increase heat exchanger fouling

www.eptq.com

Do you want your bottom of

the barrel to be this clear?

LC-SLURRY is built upon the commercially proven and well established

LC-FINING technology, in which Chevron Lummus Global has over 50 years of experience.
LC-SLURRY converts nearly 100% of heavy oils/SDA tar to high-value products. From
extensive, in-depth testing on various heavy residues in Richmond Technology Centers
labs, LC-SLURRY has consistently demonstrated 115 vol% liquid yield and up to 75 vol%
Euro V diesel (with ISOCRACKING catalysts). Instead of being stuck with heavy stock, let
LC-SLURRY handle it for you and increase your bottom line.
Please contact us at: techma@chevron.com

Two experts join forces.

The acquisition of the water and process management division of BK Giulini, including its
460 qualified employees, significantly strengthens Kuritas market position in Europe by
offering customers more than 250 years of combined industry experience. Our continuous
growth has further resulted in an enhanced line of products and services as well as a
strong international presence with a multitude of branch offices, production facilities and
laboratories. The benefit is clear: optimal on-site service worldwide and a systematic approach to innovative product development.
Kurita offers a broad portfolio of solutions for the chemical water and process treatment
for refineries and the petrochemical industry, all of which ensure a seamless, trouble-free
production process.
Process Additives
Cooling Water
Boiler Water
Waste Water
Reverse Osmosis

Kurita Europe GmbH . Industriering 43 . 41751 Viersen . Germany

Phone:+49 2162 95800 . www.kurita.de

4375_00KU_RZ_Anzeige_Kurita_030615_MS.indd 1

efficiency for industry

03.06.15 14:38

in downstream units Desalting

efciency can be further challenged
by wax precipitation in the cold
train exchangers.4
Figure 2 is an illustration of heat
exchanger fouling behaviour in a
crude preheating system introduced
in Figure 1.1 The gure shows the
sensitivity of the fouling rate with
the operating conditions of the heat
exchanger, namely the crude stream
lm temperature and the crude
stream
velocity.
Apart
from
exchangers E8D and E8C, the fouling rates have an increasing trend
with increased lm temperature
and reduced crude-side velocity.
This dynamic behaviour matches
reported experimental fouling literature where crude undergoes
chemical reaction fouling.5 The
outliers, E8D and E8C, are the
second and third heat exchangers
downstream of the desalter. These
units are suspected to experience
salt crystallisation due to inefcient
desalting with water carryover. Salt
crystallisation
might
not
be
observed in the exchanger immediately downstream of the desalter as
the
thermodynamic
conditions
might not be sufcient enough for
the water components to vaporise
at this stage. The surface plot in
Figure 2 represents a model t for
chemical reaction fouling.
In the last decade, identifying and
utilising dynamic fouling conditions
for best practices has gained importance
in
renery
operation,
engineering design and management decision-making.6,7 Recognising
the increasing importance of utilising fouling knowledge to practice,
IHS developed SmartPM combining
fouling process physics and proprietaryheat
exchanger
technology
developed over the last 40 years
by IHS with new, advanced simulation
and
cleaning
analysis
techniques developed at Cambridge
University. This article highlights
the use of SmartPM software in
renery wide fouling mitigation
studies.

Novel approach to fouling

management

Reneries are dynamic environments with operational parameters

varying with the type of crude

www.eptq.com

Full HEX and network modelling

Proprietary technologies

Detailed HEX simulations

Network thermo-hydraulics

Tube inserts (e.g. Turbotal)

Novel baffle arrangements

Research tools

SmartPM

Dynamic fouling models

Models for fouling rigs

Decision-making tools
Advanced data reconciliation
Cleaning scheduling
Retrofit and revamp

Figure 3 Key aspects of SmartPM software (HEX denotes heat exchanger)

processed to the type of products

required by the market. Heat recovery
systems
endure
these
operational dynamics, making the
prediction of heat exchanger fouling
a complex process. SmartPM
breaks the complex process to
manageable, smaller, problems,
unmasking the key parameters
responsible for fouling. The renerys store of historical raw
performance measurements are
assessed for viability and validity
and fouling physics models are
tted to track historical fouling
performance. This enables the operator with a wide range of planning
scenarios, from the development of
optimum predictive maintenance
schedules to the assessment of the
economic and environmental performance benets of revamps and even
network redesign. Even without
historical data, the fouling physics
models can be used to guide these
scenarios based on performance of
other reneries processing similar
crudes.
Key
components
of
SmartPM are shown in Figure 3.
Heat
exchanger
specication
sheets, P&ID diagrams and stream
thermophysical properties provide
the information to build a preheat
train model. Any information on
the locations of monitoring tags
and associated measurement are
also
entered,
either
through
connecting to the plants distributed control system (DCS) or
through importing data storage
les.
Data
reconciliation
is
performed to t the available data
to a physical model (a heat and
mass balance). During this analysis,
a full network simulation is
performed to generate all engineer-

ing geometrical and operational

parameters of the exchangers (for
instance, stream lm coefcients,
Reynolds numbers, surface shear
stresses and metal temperatures).
Fouling analysis is performed for
the reconciled data where fouling
propensity factors are obtained for
particular crude blends. Fouling
propensity factor is dened for a
crude blend and represents the
degree of fouling in a specic operating condition. An example form
of the rate of fouling of a heat
exchanger is shown by:
!

=

!

!"#$

(1)

Here, h is the lm transfer coefcient, E is the activation energy, R is

the gas constant, Tlm is the lm
temperature and s is the surface
shear stress. Dynamic fouling
models are then used for predicting
future
performances
including
performances based on changes to
the network (through retrot and
revamp projects) or performances
based on generated optimum cleaning schedules. These aspects will be
discussed in a series of case studies.

Case study 1: Advanced data

reconciliation

In SmartPM, monitoring data are

linked to the network simulator
model though the graphical attachment of tags to appropriate streams
(Figure 4 shows tags attached to a
UK renery crude preheating train).
These tags store time-dependent
ow rate, temperature, and/or pressure measurement data. Not all
streams have temperature/ow
measurements. The thermophysical

PTQ Q3 2015 19

Figure 4 SmartPM network with a snapshot highlight of the quality of monitoring data. The colours of the tags, green, orange and red,
attached to the stream indicates either the data is of good quality, handle data with caution, or incorrect

properties of the streams and the

actual available temperature/ow
measurements have uncertainties.
Data reconciliation approach is

based on three steps:

1. Generation of missing data
through simulation using full heat
exchanger modelling

Crude slate

100
80
60
40
20

Fouling resistance,
h.ft2.F.Btu1

0
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
Jan 2011

Apr 2011

Jul 2011

Oct 2011

Figure 5 Combination plot of the variation in crude slate and fouling resistance over
an operational period. The green area on the fouling resistance is the fouling resistance
predicted via Equation 1

20 PTQ Q3 2015

2. Filtering unreliable data through

a trusted heat balance
3. Grouping heat exchanger monitoring data into periods of
processed crude blends.
Figure 4 highlights a snapshot of
the agreement of the available plant
measurements with the calculated
temperature at the tag locations.
The tags are coloured green, orange
or red. Green implies that the
monitored data have a very close t
following the mass and heat
balance. Orange is an indication to
handle the data with caution due to
possible inaccuracies in thermophysical properties or monitoring
instruments. Red indicates data
outside a usual range, perhaps due
to the adjustment of the reconciled
temperature to ensure the heat
balance or an incorrect tag
measurement.
A key output is the historical
operational parameters (such as
variation in transfer coefcients,
stream velocities) including fouling

www.eptq.com

Case study 2:
When to clean which units

A common strategy to mitigate the

effects of fouling is to periodically
clean fouled units. Systematic cleaning of fouled heat exchangers is a
proven methodology to minimise
economic
penalties.6,8,9 Heuristic
cleaning scheduling formulations
implemented in SmartPM incorporate practical constraints such as:
1. Mechanical constraints: locations
of heat exchanger isolation valves
to identify if the heat exchanger can
be isolated individually or as a
group
2. Operational
constraints:
for
instance, operation at constant
pumparound heat duty/return
temperature, furnace ring capacity
limits
3. Economic
constraints:
for
instance, cleaning budget for the
nancial year.

www.eptq.com

Crude-side equivalent
shear, Pa
Heat duty,
MMBtu.h1
Fouling resistance,
h.ft2.F.Btu1

resistance (Rf) proles for each

exchanger with time. When a heat
exchanger is selected, a combination plot (see Figure 5) shows the
change in Rf prole with variation
in the crude blend.
The green area in the fouling
resistance graph (see Figure 5) is a
t of the dynamic fouling model
described by Equation 1. This
describes a fouling propensity of
450 h-1, for the crude blend, crude
A and crude blended in a specic
average proportion over the period
January 2011 to October 2011. The
sudden drop in fouling resistance
around August 2011 denotes a heat
exchanger cleaning event. The fouling resistance prole no longer
follows the trend given in Equation
1 using fouling propensity of 450
h-1 as the crude blend composition
has changed. A third crude has
been introduced from October 2011,
reducing the fouling trend. The
proles for heat duty and the
crude-side shear stress for this
exchanger are shown in Figure 6.
While the shear stress remains
constant, the heat duty has reduced
signicantly due to the increased
thermal resistance. The historical
cleaning action is superimposed on
the plot, denoting a minimum fouling resistance identied following a
cleaning event.

3
2
1
0
15
12
9
6
3
0

0.05
0.04
0.03
0.02
0.01
0
Jan 2011

Apr 2011

Jul 2011

Oct 2011

Figure 6 Combination plot of the variation in crude-side surface shear, heat duty and
fouling resistance. The vertical stripe indicates a historical cleaning event (hydro
blasting). Star on the fouling resistance graph identifies the minimum fouling resistance
after a cleaning action

Once the mechanical, operational

and economic constraints are specied together with the economic
parameters (cost of energy, cost of
lost opportunity, cost of cleaning),
the scheduling formulation generates an optimised cleaning schedule
giving the highest economic benets. The scheduling formulation
uses the predictive capability of
SmartPM
coupled
with
the
dynamic fouling models identied

for the crude slate being processed.

Table 1 is a simplied Gantt chart
representation of when to clean
which units. Specic details (date
of unit to be taken ofine, furnace
inlet temperature drop, and so on)
could be extracted from the software. The net benet for each
nancial year and the CO2 emission
savings are summarised in Table 2.
When there are product streams
connecting heat exchangers above

An example representation of when to clean which units. The number 7 denotes

the number of days taken to clean the unit
2015
2016
2017
Unit name 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3
E014
7
7
E015
7
7
E008
7
E001
7
7
7
E002
7
7
7
E056
7
7
E023
7
E024
7
E025
7
E018
7
E033
7
E034
7
E021
7
7
7
7
E016
7
7

Table 1

PTQ Q3 2015 21

340

330

Temperature, F

320
310 O
300

290

280
270

260
250

20
Ju
n

13

ar
M

15

15
20

14
20
22

ec
D
29

ct

20

20

14

14

Ju
l
15

Ap

r2

01

14
20
23

Ja
n

20
29

ov
N
7

16

Au

20

13

13

240

Figure 7 Schematic of an example desalter inlet temperature in a heat exchanger

network where the heat exchangers are undergoing cleaning
Annualised summary of the generated cleaning schedule
Year Start
date

End
date

No of Cleaning benefit Cleaning benefit Total cost Total net

cleans average furnace
CO2 saving,
cleaning, savings,
duty, MMbtu/hr
tons
US$
US$
1 01/01/2015 31/12/2015 8
18.31
8800
260 000 717 800
2 01/01/2016 31/12/2016 8
39.38
18 900
220 000 2 114 900

Table 2

Summary of benefits in descending order on exchangers cleaned today

Top 5 candidates for cleaning today (8 February 2015)
Unit name Annual net Cost of
Unit duty
Unit duty Network duty gain Initial FIT
benefit cleaning before cleaning, after cleaning, after cleaning, increase
MMBtu/hr
MMBtu/hr
MMBtu/hr
F
E014/E015 300 200 74 200
39.57
49.15
5.56
4.9
E008
179 600 28 200
16.17
28.22
3.77
5.3
E001/002 144 000 69 100
34.16
46.93
2.81
2.4
E056
141 100
22 900
2.89
6.42
1.88
1.6
E023/024 116 700 64 200
44.75
54.90
3.59
5.1

Table 3

Figure 8 Preheat train diagram with a list of tube-insert models available in SmartPM

22 PTQ Q3 2015

and below the desalter (see Figure

4), cleaning actions downstream of
the desalter could have a strong
influence on the desalter inlet
temperature. 7 is an illustration of
the desalter inlet temperature when
the heat exchanger network is
subject to a cleaning action. The
increases and decreases in the
desalter inlet temperature are
related to dynamic fouling and
scheduled cleaning:
O to A: The desalter inlet temperature gradually increases as the heat
exchangers downstream of the
desalter are fouling (faster than the
heat exchangers upstream of the
desalter) causing the product
stream to enter the upstream heat
exchangers at a higher temperature.
A to B: The heat exchanger downstream of the desalter has been
cleaned, causing the product stream
to eject more heat at the upstream
heat exchangers.
B to C: The heat exchanger
upstream of the desalter is taken
offline for cleaning.
C to D: The heat exchanger downstream of the desalter is back online
after being cleaned (recovering
more heat), reducing the heat
recovered on the heat exchangers
upstream of the desalter.
D to E: The heat exchanger
upstream of the desalter is back
online after being cleaned.
The refinery operator could
assess economic benefits of which
exchangers to clean today, based
on the descending order of benefit
(see Table 3). The network duty
gain is less than the individual heat
exchanger duty gain (difference of
unit duty after and before cleaning), as the individual thermal
energy gain diminishes as it moves
along the network. The annual net
benefit represents the benefit over a
one year operating period. The cost
of cleaning includes the cost of
cleaning the actual unit plus the
additional energy cost for the
furnace to accommodate the
exchanger being offline for cleaning. The operator could make
decisive economic evaluations in
situations where there will be a
plant shutdown in near future (for
example, in a six-month period).
When cleaning heat exchangers at

www.eptq.com

Good For Furniture,

Good For Floors,
Bad For Diesel.

ExxonMobil and Honeywells UOP

offer integrated solutions for high-quality
lubes production and maximum production
of premium distillate using an array of
commercially proven processes.

Wax is good for a lot of things diesel isnt one of them.

ExxonMobil and Honeywells UOP have brought together their leading
technologies to deliver the ultimate solution for enhanced diesel quality.
The combination of UOPs Unionfining and Unicracking processes with
ExxonMobils MIDW dewaxing technology ensures maximum
value for producers of clean diesel. So say hello to the
powerful alliance of UOP and ExxonMobil, and start
producing high yields of ultra-clean diesel.

www.exxonmobil.com/refiningtechnologies

www.uop.com/processing-solutions/refining

2014 Honeywell International, Inc. All rights reserved

SPM-UOP-60 Dewaxing Ad_210x297mm.indd 1

5/26/15 9:49 AM

WHEN CONDITIONS
HEAT UP
DONT LET CORROSION
SHUT YOU DOWN
Whether its higher temperatures, rising pressures or more acidic media, conditions in
oil refineries have never been more extreme. Tube and pipe corrosion are a constant
threat, causing as many as half of all major shutdowns. This is why hundreds of the
worlds most demanding petrochemicals refiners are turning to the next generation of
corrosion resistant alloys. Like one German oil refinery, which used Sandvik SAF 2707
HD hyper-duplex heat exchanger tubes to reduce the number of shutdowns from 8 to
1 over a period of four years. The result was massive savings on material replacement.
So as your tubes performance is pushed to new heights, find out how we can help
keep corrosion from shutting you down.
SMT.SANDVIK.COM

Scenario 1 with HEX cleaning

Scenario 1 without HEX cleaning

580
560
540

FIT, F

plant shutdowns, there would be

no additional energy cost for the
furnace but the cost of cleaning the
actual unit could be tripled due to
contractors charging a higher cost.
If the heat exchanger is to be
cleaned at the plant shutdown, the
network will lose the energy benefit
that could have been recovered
through cleaning the unit today.

520
500
480
460

Case study 3: Capex heavy solutions

Case study 4:
Limits on furnace operations

Furnaces (fired heaters) are also

known to undergo coking on the
crude side, where the rate of coking
is strongly influenced by the variation in FIT.10 The properties of the
coke deposits are further complicated as they are subject to ageing.11
Coke formation in furnace tubes
can increase the tube surface (skin)
temperature through impeding heat
transfer. The mechanical integrity
of the furnace could be hindered
through exposure of its tubes to
temperatures above the maximum
specified safety limits imposed by
SmartPM
the
manufacturer.12
explores the variation of the tube
skin temperature (TST) of a crude

www.eptq.com

440
420

10

20

30

40

30

40

Time, months
B

1160

Tube skin temperature, F

SmartPM enables the user to

explore retrofit and revamp options
in a series of stages. Heat exchanger
modifications (addition of tube
inserts, switching crude from shellside to tube-side, swapping out a
segmental baffle for a helical baffle,
and so on) can be assessed based
on individual performance and
network performance. SmartPM
includes models for proprietary
heat exchanger designs and tube
insert technologies which enables
the thermohydraulics to be evaluated with retrofit changes (see
Figure 8). Table 4 is a summary of
selected tube insert geometries and
annual energy benefit due to a
reduction in fouling and enhancement in heat transfer. The software
enables the assessment of thermal
energy gain with pressure drop
limitations of the stream.
Network re-piping and heat
exchanger rearrangements could
also be explored through introducing the changes to the heat
exchanger model.6

Scenario 1 with HEX cleaning

Scenario 1 without HEX cleaning
Maximum skin temperature

1150
1140
1130
1120
1110
1100
1090
1080
1070
1060

10

20

Time, months
Figure 9 Scenario 1, in the absence of furnace coking: (a) variation of FIT with time, and
(b) variation of TST with time. Solid line: no cleaning; discontinuous line: with cleaning
actions; horizontal bold discontinuous line in (b): maximum safety specification for TST
Insert type, geometry and energy benefit due to heat transfer enhancement and
fouling reduction
Insert type
Twisted tapes
Hitran
Petroval

Insert geometry
360 turn per 100 mm
Loop density of 1
4:1

Annual energy benefit, US$

304 000
370 000
463 000

Table 4

furnace throughout its operation.

FIT is directly influenced through
the degree of heat recovery in the
PHT. A typical crude furnace
consists of two main sections: a
convective section and the radiant
section. As the crude within the
furnace is in two-phase flow, the
fouling mechanism is complicated
and the dynamic behaviour varies
in both temporal and axial dimensions.10 Radiative heat transfer to
the tubes of the radiant section
consists of several factors including
heat transfer from the burner flame

to the tubes, the refractory wall to

the tubes and the gas to the tubes.
As only one side of the tubes is
facing the burner, the degree of
coke formation within the tube
internal is non-uniform.13 The case
study illustrated in this section uses
SmartPM to investigate the influence of heat exchanger cleans on
the variation in TST and the operating strategies for safe operation of
the furnace. The preheat train
network illustrated in Figure 1 is
utilised to describe the following
three scenarios.

PTQ Q3 2015 25

Scenario 2 with HEX cleaning

Scenario 2 without HEX cleaning
Maximum skin temperature

Tube skin temperature, F

1160
1150
1140
1130
1120
1110
1100
1090
1080
1070
1060

10

20

30

40

Time, months
Figure 10 Scenario 2: Variation in TST with time with furnace coking. Solid line: no
cleaning; horizontal bold discontinuous line: maximum safety specification for TST;
discontinuous line: with cleaning actions

Scenario 1: Influence of FIT on TST

in the absence of furnace coking

The performance of the preheat

train was simulated, (i) without any

Crude throughput, 1000lb.h1

cleaning actions, and (ii) with an

optimised
cleaning
schedule.
Without any cleaning actions, FIT
drops considerably (approximately

Scenario 3 with HEX cleaning

Scenario 3 without HEX cleaning

820
800

Scenario 2: Influence of FIT on TST

with furnace coking

780
760
740
720
700
680

10

20

30

40

Time, months

Tube skin temperature, F

1160

Maximum skin temperature

Scenario 3 without HEX cleaning
Scenario 3 with HEX cleaning

1150
1140
1130
1120
1110
1100

With furnace coking, the TST

would be higher as the coke material would have a lower thermal
conductivity than the tube material.
For illustration, Scenario 1 was
repeated including a constant
coking rate. Figure 10 illustrates the
variation in TST with operating
period. Without any cleaning
actions, TST has increased to 1160F
(625C) by the end of the three
years. This is ~50F higher than the
case in the absence of furnace
coking. It is also noticeable that
several cleaning actions force the
TST to exceed the maximum TST
limit during the periods when the
heat exchangers are offline (dashed
spikes in Figure 10).

1090

Scenario 3: Control of TST by

variation in crude throughput

1080
1070
1060

10

20

30

40

Time, months
Figure 11 Scenario 3: furnace operation with throughput control, (a) variation of
throughput with time, and (b) variation of TST with time. Solid line: no cleaning;
discontinuous line: with cleaning actions; horizontal bold discontinuous line (in (b)):
maximum safety specification for TST

26 PTQ Q3 2015

90F drop in three years; solid bold

line in Figure 9a). In actual operation, the heat exchangers are subject
to cleaning and would maintain a
higher FIT. The variation in FIT for
an optimum cleaning schedule is
plotted as the discontinuous line in
Figure 9a. The sharp drops in the
FIT correspond to heat exchanger
cleaning actions. Table 5 represents
the optimum schedule on when
and which units are cleaned. With
the reduction in FIT, to maintain a
constant FOT, the furnace has to
provide the additional duty. Higher
furnace duty results in increased
TST, depicted in Figure 9b. When
no cleaning actions are performed,
TST has increased to about 1140F
(615C) by the end of the simulation. This has increased beyond the
safety operating limit of 1103F
(595C), denoted in the horizontal
dashed line. Under the optimised
scheduling, TST increases considerably, corresponding to the period
when the heat exchanger is offline
for cleaning.

Several studies discuss illustrations

of imposing a maximum furnace
firing capacity in preheat train
scheduling studies.14,15 The furnace
in case studies 1 and 2 did not reach
its maximum furnace capacity
during the simulated period and
hence exhibited a constant through-

www.eptq.com

Global leader in
hydroprocessing
catalysts offering
the complete
range of catalysts
and services

artcatalysts.com
ART-ad-297mmx210mm.indd 1

2/23/15 3:55 PM

Representation of when and which units are cleaned in the optimum cleaning schedule

HEX 1
E8D
E9A
E9B
E9C
E90 X
E9E X
8A
8B
8C

X
X
X

X
X

Months
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X

Table 5

Case study 5: Beyond crude fouling

Table 6

put. TST is a variable which is not

always continuously monitored (or
calculated) by furnace operators. In
order to control the TST (that is, to
maintain the TST below its maximum operating limit), imposing a
lower furnace firing capacity (for
instance, 212 MM btu h-1) was
considered in the simulation. For the
case of no cleaning, the simulation
starts at a lower throughput (777
800 lb h-1; bold line in Figure 11a)
indicating that the furnace has
already reached the imposed firing
capacity. As the throughput is
reduced, the furnace duty is lower
and the TST falls below the maximum operating limit. When the heat
exchangers are subject to cleaning, a
steep reduction in the throughput is
observed in the periods when the
heat exchangers are offline (dashed
line in Figure 11a). The corresponding TST is illustrated in Figure 11b,
where TST is below the maximum
limit throughout the operation.
Scenario 1 illustrated an idealised
operation where there is no coking
in the furnace. In practice, furnaces
experience coking and the increase

28 PTQ Q3 2015

in TST above its specified maximum safety limit could be expected

(as in Scenario 2). Scenario 3 illustrated the reduction of TST through
reducing the crude throughput via
artificially specifying a furnace
firing capacity below its actual
burner capacity. For Scenario 3, the
throughput reduction was quantified by integrating the area of
Figure 11a and comparing with the
initial constant throughput of
801 600 lb h-1. The total amount of
unprocessed crude amounts to
approximately 350 000 bbl over the
three-year period. Considering a
margin of $2/bbl, $700 000 has been
lost through reduced production in
this case study. That is, $700 000 is
the cost over three years for operating the TST below its maximum
safety limit. Scenarios 2 and 3
considered a constant coking rate
for the furnace. However, furnace
coking exhibits dynamic behaviour.10 It would be useful to
combine the dynamic furnace
coking model to establish a more
accurate relationship between FIT,
coking rate and TST.

Fouling in streams other than crude

oil is frequently observed in the
refinery. An example is the fouling
of heat exchangers associated with
petroleum upgrade technologies.
Examples are visbreakers and high
conversion soaker cracking (HSC)
units, which are thermal processes
in refineries to upgrade residual
oils produced by atmospheric or
vacuum distillation of crude. The
associated process streams in
visbreaking and HSC units consist
of heavy hydrocarbons with very
high viscosities and impurities,
resulting in fouling of heat
exchangers used to cool or heat
these streams. Such fouling is not
necessarily limited to a chemical
reaction type fouling and a range
of fouling models are available in
the
SmartPM
software
(see
Table 6).
Fouling under two-phase flow
conditions could occur when the
associated stream pressure is not
high enough to keep the stream
under single phase flow (for example, for crude (Ishiyama and Pugh,
2015)).16 The main effect of stream
boiling on fouling rate is through
the high turbulence created by the
bubbles at the solid-fluid interface.
This turbulence enhances both heat
and mass transfer. Boiling reduces
the boundary layer resistance and
promotes the flow of fouling
precursor toward the heated
surface. Such fouling dynamics can
be explored through the PM; the
software utilises two-phase flow
thermohydraulics from standard
engineering literature (IHS ESDU,
1991).
Ageing of foulant deposits (for

www.eptq.com

relentless

UOP has been serving catalyst customers for

decades, and constant innovation ensures well
be there for decades to come.
Pilot plants. A true commitment to fast, thorough research and
development. A vast team of experienced, dedicated engineers
and scientists. The reasons to choose UOP catalysts are many,
but perhaps the biggest of all is that UOP never quits innovating. When you use UOP
catalysts, youll know youre using products specifically designed to help optimize
your operation. UOP invented the refining and petrochemical technology used in most
operations today, and that knowledge and expertise comes through in every catalyst we
develop. Youll always get the high stability and global support your business needs, and
youll be working with a company that will keep you on the leading edge year after year.

For more information about UOP catalysts, visit www.uop.com/catalysts.

2015 Honeywell International, Inc. All rights reserved

SPM-UOP-87 Adsorbents Ad_297x210mm.indd 1

2/18/15 9:47 AM

instance, coking) could also have

an influence on plant economics.
Deposit ageing is the process of
transformation of a fouling layer
following its initial formation,
whereby the deposit converts to
another, usually more cohesive,
form. Such ageing models17 are
available to use in SmartPM simulation studies.

Conclusion

SmartPM is a novel heat exchanger

network modelling tool enabling
the use of dynamic fouling models
in historical performance assessment and predictive studies. The
software provides a powerful, yet
easy to use platform for reconciling
heat exchanger network data,
dynamic modelling hydrocarbon
fouling, network simulation and
scheduling predictive cleaning.
Detailed modelling of thermohydraulic exchanger is performed,
including segmental, helical baffles,
and a range of commercial tube
inserts.
Accurate data reconciliation can
be performed with far fewer tags
than conventional monitoring software, with no manual intervention.
Predictive modelling is based on
dynamic models that correctly
account for the fouling physics of
hydrocarbons and crude slates in
shell-and-tube heat exchangers.
Cleaning scheduling takes into
account all network interactions
(compensation effects) and crude
fouling behaviour for correct estimates
of
future
economics.
Cleaning schedules are optimised
for minimum energy use, desalter
operation temperature bounds and
crude maximum throughput. Top
5 exchangers for cleaning today
are ranked in terms of descending
order of benefit, enabling the user
to make confident actions in decisive
situations.
All
cleaning
scheduling calculations incorporate
practical
constrains
reflecting,
mechanical, financial and operational limits.
Retrofit and revamp studies are
easily performed as the software
includes thermohydraulic models
for proprietary exchanger technologies and tube inserts, giving a clear

30 PTQ Q3 2015

quantification of individual and

network
performance
benefits
under dynamic fouling conditions.
Furnace operational limits such
as the firing capacity and the maximum tube skin temperature limits
can be evaluated by the software,
enabling an overview of the influence of cleaning schedule on the
furnace operations.
Dynamic fouling models for
streams other than crude are available in the software. This enables
simulation of fouling in both the
heating and cooling streams in a
heat exchanger and associated
networks. Such fouling situations
are observed in heat exchangers
associated with petroleum upgrading systems (visbreaker units).

References
1 Ishiyama E M, Heins A V, Paterson W R,
Spinelli L, Wilson D I, Scheduling cleaning in
a crude oil preheat train subject to fouling:
incorporating desalter control, Applied
Thermal Engineering, 30, 2010, 1852-1862.
doi:10.1016/j.applthermaleng.2010.04.027
2 Watkinson A P, Deposition from
crude oils in heat exchangers, Heat
Transfer Engineering, 28, 2007, 177184.
doi:10.1080/01457630601064413
3 Watkinson A P, Wilson D I, Chemical reaction
fouling: a review, Experimental Thermal and
Fluid Science, 14, 1997, 361374.
4 Benoit B, Zurlo J, Overcoming the challenges
of tight/shale oil refining (Processing shale
feedstocks), GE Waters, 2014.
5 Ebert W A, Panchal C B, Analysis of Exxon
crude-oil slip stream coking data, in: Fouling
Mitigation of Industrial Heat Exchange
Equipment, Begell House, New York, 1997,
451460.
6 Ishiyama E M, Pugh S J, Paterson W R, Polley
G T, Kennedy J, Wilson D I, Management of
crude preheat trains subject to fouling, Heat
Transfer Engineering, 34, 2013, 692701. doi:1
0.1080/01457632.2012.739036
7 Wilson D I, Polley G T, Pugh S J, Ten years
of Ebert, Panchal and the threshold fouling
concept, in: Heat Exchanger Fouling and
Cleaning, Kloster, Irsee, Germany, 2005.
8 Ishiyama E M, Paterson W R, Wilson D I,
Platform for techno-economic analysis of
fouling mitigation options in refinery preheat
trains, Energy and Fuels, 23, 2009, 1323-1337.
9 Smali F, Vassiliadis V S, Wilson D I,
Optimization of cleaning schedules in heat
exchanger networks subject to fouling,
Chemical Engineering Communications, 189,
2002, 1517-1549.
10 Morales-Fuentes A, Picn-Nez M,

Polley G T, Mndez-Daz S, Analysis of the

influence of operating conditions on fouling
rates in fired heaters, Applied Thermal
Engineering, 62, 2014, 777-784. doi:10.1016/j.
applthermaleng.2013.10.016
11 Fan Z,Watkinson A P, Aging of carbonaceous
deposits from heavy hydrocarbon vapors,
Industrial and Engineering Chemistry Research,
45, 2006, 6104-6110.
12 Moss C J, Barrien P, Walczynski A, Life
management of refinery furnace tubing,
International Journal of Pressure Vessels and
Piping, 77, 2000, 105-112. doi:10.1016/S03080161(99)00090-3
13 Jegla Z, Kohoutek J, Stehlk P, Design and
operating aspects influencing fouling inside
radiant coils of fired heaters operated in crude
oil distillation plants, in: Heat Exchanger Fouling
and Cleaning, Crete, Greece, 2013.
14 Ishiyama E M, Pugh S J, Kennedy J, Wilson
D I, Ogden-Quin A, Birch G, An industrial case
study on retrofitting heat exchangers and
revamping preheat trains subject to fouling, in:
Heat Exchanger Fouling and Cleaning, Budapest,
Hungary, 2013.
15 Lavaja J H, Bagajewicz M J, On a new MILP
model for the planning of heat-exchanger
network cleaning. Part II: Throughput loss
considerations, Industrial and Engineering
Chemistry Research, 44, 2005, 8046-8056.
16 Ishiyama E M, Pugh S J, Considering in-tube
crude oil boiling in assessing performance of
preheat trains subject to fouling, Heat Transfer
Engineering, 36, 2015, 632641. doi:10.1080/0
1457632.2015.954916
17 Ishiyama E M, Coletti F, Macchietto S,
Paterson W R, Wilson D I, Impact of deposit
ageing on thermal fouling: lumped parameter
model, AIChE Journal, 56, 2010, 531545.
doi:10.1002/aic.11978
Simon Pugh is a Director and Consultant
with IHS Energy Downstream Research
and is responsible for the development and
application of advanced software for heat
exchanger networks, with particular emphasis
on fouling management in refinery pre-heat
trains. Under his strategic direction, SmartPM
software was co-developed with Cambridge
University using technology transferred to
IHS from the UK Government funded crude
oil fouling research programme, CROF, where
IHS was coordinator and technology transfer
partner. He holds a BSc degree in mechanical
engineering from Brunel University, UK.
Edward Ishiyama is a Senior Research Engineer
and Consultant with IHS Energy Downstream
Research and is lead developer on SmartPM. His
current research into the application of heat
transfer, process control and thermodynamic
principles identifies and solves problems
associated with heat exchangers and heat
exchanger networks undergoing fouling. He
holds a PhD from the University of Cambridge,
UK.

www.eptq.com

Trying to win without

a full set of options?
You need every club in the bag for todays
opportunity feedstocks, variable economics,
and shifting conditions.
The new ACHIEVE series of FCC catalysts from Grace hands you
the versatility and functionality you need for todays dynamic refining
environment. Select the ACHIEVE advantage thats right for your
feeds, your process, and your business requirements.
ACHIEVE 100 High activity for light feedstocks
ACHIEVE 200 Coke-selective bottoms conversion
ACHIEVE 300 Propylene selectivity
ACHIEVE 400 Octane and butylene selectivity
ACHIEVE 800 Maximum metals tolerance and resid conversion
Contact us today to learn more about how we can tailor a formulation
to meet your most challenging requirements and give yourself a
clear shot to win.

grace.com/achieve

achieve-297mmx210mm.indd 1

2/26/15 3:18 PM

HARNESS THE POWER

OF MANUFACTURING INNOVATION

RENTECH engineers build unmatched power and performance into

every boiler we deliver. Our 80-acre manufacturing facilitythe industrys most
technologically advancedincludes heavy bay and light bay areas with direct access
to rail, cross-country trucking routes and shipping facilities. We master every detail to
deliver elemental power for clients worldwide. Take an expanded tour of our facilities
today at www.rentechboilers.com/facilities
HARNESS THE POWER WITH RENTECH.

HEAT RECOVERY STEAM GENERATORS

WASTE HEAT BOILERS
FIRED PACKAGED WATERTUBE BOILERS
SPECIALTY BOILERS

WWW.RENTECHBOILERS.COM

Dealing with depleting sour gas reserves:

Part 1
Depletion of sour gas production in western Canada means that gas plants will
have to evaluate their sulphur recovery units for future turndown operation
MARCO VAN SON and SHASHANK GUJALE
Jacobs Comprimo Sulfur Solutions

n recent years, most of the sour

gas reservoirs in Alberta have
been depleting and new wells
coming on-line are sweeter and
leaner, resulting in a reduction in
the quantity and the richness of the
acid gas available for existing
sulphur recovery units to process.
As a result, Jacobs Comprimo
Sulfur Solutions has been involved
in the investigation, evaluation and
recommendation of processing
options available to these sulphur
recovery units that will need to be
operated in severe turndown
conditions.
As most of the sulphur recovery
units in Alberta, Canada are older,
they were not necessarily designed
for turndowns as large as 10 to 1,
or for lean acid gas operation.
Different options were explored for
several gas plants and also one
refinery to allow the SRU to process
the predicted acid gas and still
maintain the regulatory requirements for the facility. These options
vary from replacement of individual equipment items in the existing
SRU, to operating with natural gas
co-firing, to catalyst replacement
and finally installation of a
complete new SRU.
This article in two parts will
describe plants where acid gas rates
and composition have become a
concern, what the proposed modifications were to ensure that the
processing objectives of the facilities could still be met in the coming
years. At the same time some
performance data from operating
with natural gas co-firing and its
impact on sulphur recovery will be
discussed. This first part of the article includes case studies for two

www.eptq.com

Alberta Sulfur Recovery Guidelines per EUB Interim Directive 2001-3

Sulphur inlet rate,

tonnes per day
1-5
5-10
10-50
50-2000
>2000

% of sulphur inlet that must be recovered

Design sulphur
Calendar quarter-year sulphur
recovery criteria
recovery guidelines
70
69.7
90
89.7
96.2
95.9
98.2-98.52
98.5-98.81
99.8
99.5

1. Recovery = 98.18185 + 0.187259 log(sulphur inlet rate)

2. Calendar quarter-year recovery = 97.88185 + 0.187259 log(sulphur inlet rate)

Table 1

gas plants; the second part will

include case studies for a further
gas plant and a refinery.
In Alberta, the maximum allowable regulatory requirements for
sulphur emissions are dictated by
the inlet sulphur rate to the plant.
The overall sulphur recovery
requirement can be calculated as
seen in Table 1.
The information in Table 1 is
specific to Alberta requirements
and the local legislated sulphur
recovery efficiencies need to be
evaluated on a case by case basis
depending on the location. As is
clear from Table 1, in the case of
reduced sulphur plant capacity, an
option that can be considered is
relicensing to allow lower sulphur
recovery efficiencies in the plant.

Case study: using co-firing

to produce mass flow

For the first case study, Jacobs

Comprimo Sulfur Solutions evaluated the options for a sulphur
recovery unit at a gas plant in
Alberta which was originally
designed for 550 t/d and expected
to have a future processing

capacity of less than 100 t/d. In

addition, the original plant was
designed for an acid gas quality of
75 vol% and it was expected that,
due to the processing of leaner sour
gas streams, the acid gas H2S
concentration will drop to 50 vol%.
The plant in question had severe
turndown limitations due to its
configuration which consisted of
two hot gas bypass reheaters and a
gas/gas exchanger. The configuration of the unit is provided in
Figure 1. In addition, the plant had
a two-pass waste heat boiler
producing 400 psig steam on the
tube side, contrary to the more
commonly
employed
firetube
design. The condensers of the facility produced 50 psig steam and had
a common steam drum. The original burner in the plant had been
replaced with a high intensity
burner.
The sulphur plant is an existing
three-stage Claus unit followed by
an 1100 t/d Sulfreen unit with an
overall sulphur recovery efficiency
of 99.0%. According to the Alberta
Sulfur Recovery guidelines, the
future regulatory sulphur recovery

PTQ Q3 2015 33

THIOPAQ O&G
stable by nature

The proven gas

desulphurisation technology.

How to reach highest value when treating natural gas streams for sulphur?
THIOPAQ O&G puts you in control of sulphur removal and sulphur recovery.
Perform well on safety, sustainability, reliability, cost and operability. Oil & Gas
companies worldwide rely on THIOPAQ O&G. See why on paqell.com/thiopaq.
Paqells THIOPAQ O&G - exceptional achievements in H2S removal.

paqell.com

015-0083 Adv PTQ 210x297_216x303mm

9 -12-14 15.00

HGBP

Sulfreen

HGBP

Steam
Converter

Waste heat
boiler

Converter
G/G

Steam
Combustion
chamber

Converter

Condenser

Steam

Condenser

Condenser
S

Steam

Steam

H2S
2
=
QC SO
1
2

Condenser
S

FrC

Feed gas

Air

Figure 1 Process line-up for the first case study

requirement for the plant is 98.5%

when operating at 100 t/d or less
inlet sulphur.
The obvious limitation of the
plant to operate at high turndown
is the reheater conguration. Based
on the plants experience, the
current minimum processing capacity of the unit corresponds with
approximately 120 t/d. The key
reason for the limitation is the
inability of the gas/gas exchanger
to heat the inlet gas to the third
converter sufciently to remain
above the sulphur dewpoint in the
catalyst bed. The plant indicated
that after a recent replacement of
the tubes in the waste heat boiler,
the turndown capability of the
plant had actually decreased,
resulting in more concerns with the
future operation of the plant. The
impact of replacing the tubes in the
waste heat boiler can be explained
by the lower temperature from the
rst pass of the waste heat boiler.
As the gas from the rst pass of the
waste heat boiler is used to heat the
process gas to the rst and second
converters, a higher efciency of
the waste heat boiler has a negative
effect as more gas is required to
meet the same inlet temperature
into the converter. This also results
in less condensation of sulphur in
the thermal condenser, thereby
applying additional sulphur load
on the second and third stage of
the SRU. This has an impact both

www.eptq.com

on recovery and on the sulphur

dewpoints in the subsequent
converters.
The following options were
considered to allow the plant to
process the expected future sulphur
processing capacity:
1. Install a new 100 t/d SRU
2. Replace the second and third
reheater with steam reheaters
3. Use co-ring with natural gas to
increase the mass ow through the
unit and use titania catalyst to
counteract the higher formation of
COS and CS2.
The rst option, to install a new
100 t/d SRU meeting 98.5% sulphur
recovery at an estimated cost of
$25-30 million, was deemed too
expensive by the client, so this
option was eliminated without
much review. In order to meet the
required 98.5% overall sulphur
recovery efciency, some form of
tail gas treatment would still be
required and hence the installation
of a grassroots unit was considered
cost prohibitive. It would also limit
the plant to being able to return to
higher capacities in the future in
case new sour wells were added to
the plant.
As the key limitation of the
plants
turndown
capabilities
appeared to be its conguration,
and more specically the third
stage gas/gas reheater, Jacobs evaluated the option to replace the
third stage reheater with a steam

reheater. A steam reheater was

selected as the client was not in
favour of adding additional burner
management systems to the plant
which would be required with the
installation of a direct red (either
acid gas or natural gas) reheater.
In order to evaluate the performance of the rst pass of the waste
heat boiler, historical data were
collected from the plant, where the
outlet temperatures of the rst and
second pass of the waste heat boiler
were plotted as a function of the
throughput of the unit. Using this
data, it was possible to estimate the
minimum temperature required for
the outlet of the rst pass of the
waste heat boiler to ensure sufcient hot gas would be available to
reheat the inlet gas for the second
converter, which would then
ensure that there is sufcient heat
available for the third converter.
In order to maximise turndown,
the decision was made to consider
both the second and third reheaters
for replacement with steam reheaters. The steam available for the
new reheaters was only 400 psig,
corresponding to a temperature of
445F (230C). This steam temperature would limit the capability of
the plant to perform a heat soak on
the second and third converters.
However, as the plant is currently
already not able to do a heat soak
in these beds, this was considered
acceptable. The rst stage hot gas

PTQ Q3 2015 35

UOP-EMRE-Ad-210mmx297mm_EM UOP Ad 210mmx297mm size UOP6587 5/26/15 8:24 AM Page 1

www.eptq.com

COS concentration,
mol% (dry basis)

bypass was maintained due to the

inability to increase the inlet
temperature into the first converter
with 400 psig steam.
In addition to the two new steam
reheaters to replace the second and
third reheater, Jacobs evaluated the
impact of having a leaner acid gas
composition with an expected
increase in BTEX content. Per
industry practices, a minimum
temperature of 1920F (1050C) is
required for proper BTEX destruction and Jacobs estimated that, with
an acid gas containing 50 vol% H2S,
additional means for attaining the
required temperature would be
necessary in the reaction furnace.
Jacobs considered two options to
increase the reaction furnace
temperature to deal with the higher
BTEX concentration in the acid gas:
Co-firing natural gas with the
current configuration
Installation of steam heated acid
gas and air preheaters.
In addition to meeting a minimum temperature, the destruction
of BTEX in an SRU reaction furnace
requires a minimum residence time
of one second. As the plant was
originally designed for 550 t/d, the
residence time was not considered
to be of concern for this plant.
The sulphur recovery unit
employed conventional alumina
catalyst in the first and the second
converter and, based on the available performance test data, the COS
and CS2 hydrolysis was 91% and
86% respectively in the first
converter and 60% and 20% respectively in the second converter. The
hydrolysis activity in the second
converter can be explained by its
relatively high operating temperature (507F, 264C), which is
required to maintain the third
converter
above
dewpoint
operation.
For the co-firing option, Jacobs
assumed that the COS and CS2
hydrolysis would remain identical
(on the assumption that the inlet
temperature to the converters
would not be revised) and estimated the impact of the increased
formation of these species in the
reaction furnace on the overall
sulphur recovery efficiency. Based
on the available models in the

1080

1100

1120

1140

1160

1180

1200

1220

1240

Reaction furnace temperature, C

Figure 2 COS formation as a function of furnace temperature

simulators used (Promax and

Jacobs simulator), it was evident
that it would not be possible to
meet the required sulphur recovery
efficiency due to the substantial
bypass of gas to the second
converter via the second hot gas
bypass and the low COS and CS2
conversion in the second converter.
Therefore the option to co-fire to
maintain the reaction furnace
temperature for proper BTEX
destruction was initially discarded.
As the plant indicated that sufficient 400 psig steam was available,
the option to use both acid gas and
air preheat was considered. By
increasing the acid gas and air
temperatures to 392F (200C) and
with a lean acid gas containing 50
vol% H2S, the reaction furnace
temperature could be increased to
1050C.
A cost estimate was done for the
replacement of the first and second
reheaters and the installation of an
acid gas and air preheater and was
presented to the client. This cost
was again in excess of the clients
expectations and Jacobs was
requested to further evaluate alternative options.
The first two considerations of
the study indicated to Jacobs that
there were two limitations at play
in this facility:
Mass flow to maintain a proper
heat profile through the plant to
maintain the third converter above
dewpoint
Overall sulphur recovery losses
due to COS and CS2.

As a side note, it should be

mentioned here that the impact of
the Sulfreen unit as a tail gas unit is
limited to improving the overall
sulphur recovery efficiency with
respect to H2S and SO2 only. Any
unconverted COS and CS2 from the
first and second converter do not
get converted in the Sulfreen unit.
As a third option, Jacobs then
evaluated how a minimum mass
flow through the unit can be maintained
under
all
turndown
scenarios. By maintaining a minimum mass flow through the unit,
the outlet temperature from the
second converter is sufficiently
high to enable the gas/gas
exchanger to maintain the third
converter above the sulphur
dewpoint. In order to increase the
mass flow, co-firing with natural
gas can be used as it requires additional air for the combustion of the
natural gas component. It is well
known though that the operation
with co-firing will result in additional formation of COS and CS2.
Although COS and CS2 formation
increases, the exact increase is less
known and varies widely in the
commercially available simulators.
Therefore, a test programme was
conceived with the client to determine the impact of co-firing on the
formation of COS and CS2 in the
reaction furnace.
The results of the testing are
provided in Figures 2 and 3. It was
determined that the COS increases
as a function of increased natural
gas co-firing, however the CS2 goes

PTQ Q3 2015 37

CS2 concentration,
mol% (dry basis)
1075

1125

1175

1225

1275

Reaction furnace temperature, C

Figure 3 CS2 formation as a function of furnace temperature

to a maximum as a function of
temperature.
The test data indicated that when
co-firing is considered to increase
the mass flow, there would be a
definite impact on the overall
sulphur recovery efficiency of the
plant to the point where the regulatory requirements would no longer
be met. Therefore it was clear that
co-firing alone would not meet the
requirement to meet both the
capacity and the overall sulphur
recovery targets. In order to overcome
this
limitation,
Jacobs
evaluated the option to install titania in the existing first and second
converters to improve the COS and
CS2 hydrolysis in these reactors,
limiting the slip of these components to the Sulfreen unit.
As the existing converters were
originally designed for 550 t/d, the
catalyst beds are very large and as
a result the space velocity at the
future capacity of 100 t/d will be
low. In discussion with the catalyst
supplier, the decision was made to
install a 85/15% split alumina/titania catalyst bed in the converters,
which would be adequate to ensure
high COS and CS2 hydrolysis. The
expected values for the hydrolysis
were 97% and 88% for COS and
89% and 55% for CS2 in the first
and second bed respectively.
One of the items that was identified as a potential concern with the
installation of the new catalyst
configuration was the low space
velocity in each of the converters.
Low space velocity in a catalyst bed

38 PTQ Q3 2015

can result in channelling of the gas

through the bed. This means that
the gas may follow a preferential
path through the catalyst thereby
using the catalyst bed inefficiently.
In order to overcome this concern,
the decision was made to select a
smaller bead size for the catalyst
(2-3 mm), resulting in a higher
pressure drop and thereby better
distribution of the gas.
The
high
intensity
burner
installed on the reaction furnace
was evaluated for operation in
co-firing mode as well. As back
burning is typically the concern
with burners, the addition of cofiring was found to be beneficial to
the operation of the burner. With
co-firing, the air demand of the
burner is increased as well, resulting in a higher pressure drop across
the burner. Therefore as long as the
control system was able to handle
the required natural gas flow rates,
the burner would not be a concern.
Based on Jacobs evaluation, the
limitation of co-firing is not in the
sizing of the equipment, but in the
limitation of the installed refractory. Co-firing results in higher
temperatures in the reaction
furnace and although this has benefits with respect to contaminant
destruction, the amount of co-firing
is limited by the refractory maximum service temperature.
The new catalyst configuration
was installed in the converters
during a turnaround and since that
time the plant has tested the capability to operate at lower acid gas

rates with co-firing. The results

showed that the addition of titania
in the first and second converters
has the expected hydrolysis effect
and the plant has been able to operate successfully with co-firing while
maintaining both the dewpoint
margin in the third converter and
the overall sulphur recovery efficiency above the required values.
The conclusion from this study
was that there can be low capital
cost options to resolve turndown
issues while maintaining overall
sulphur recovery targets. Of course
in this case there will be an operating cost associated with the
co-firing mode. However, due to
the low impact on capital modifications in the plant, this was
acceptable to the plant. It also
allows the plant to have the flexibility to determine when there is a
need for the co-firing mode to be
used to meet all of the processing
requirements.
The plant is currently considering
further decreases in capacity and
Jacobs is involved in determining
the lower range of the turndown of
this particular facility. A combination of co-firing and optimisation of
the hot gas bypass operation is
currently under evaluation. The
plant is planning to do additional
performance tests at these low turndowns to understand the impact
and options when the maximum
co-firing rates are reached.

Case study: high turndown and poor

acid gas quality

In a second case study, located in

Alberta, Canada as well, the existing sulphur recovery unit was
designed to process 400 t/d of
sulphur and future predictions
indicated that the capacity would
eventually be reduced to 22 t/d
with an acid gas quality of close to
10 vol% H2S. The existing unit was
designed for an acid gas quality of
70 vol% H2S.
Processing an acid gas stream
with a concentration below 10%
H2S adds substantial complications
to the operation of the sulphur
recovery unit. Therefore Jacobs
focused first on the amine system
to determine what options were
available to improve the acid gas

www.eptq.com

Advanced Solutions for

the Worlds Toughest
Energy Challenges
KEY TECHNOLOGIES:

Premium Lubricants

Resid Upgrading

Methanol to Gasoline
Premium Diesel

Acid Gas Clean Up

ExxonMobil Global Leader in Fuels and

Lubes Process Technologies
Enabled by Proprietary Catalysts and Solvents
ExxonMobil Technologies are applied across our corporation and by licensees worldwide in
a growing list of process industry applications. Leverage our vast experience and ongoing
commitment to continuously improve our industry-leading technologies.

www.exxonmobil.com/tsl

FP.Technology_4PTQ.indd 1

11/5/14 10:52 AM

the conditions for the turndown of

this plant are much more severe.
The key parameter that really limits
the capability to reuse the existing
equipment and consider similar
modifications as for the first case
study is the poor acid gas quality.
Therefore in discussion with the
client, the decision was made to
evaluate the installation of a new
grassroots 49.9 t/d sulphur recovery unit. Per Table 1, this is the
maximum capacity for a SRU
before a tail gas technology needs
to be considered. The required
regulatory sulphur recovery efficiency considered was 96.2%.
In order to overcome the issues
associated with the leanness of the
acid gas, Jacobs proposed the
following configuration for the new
SRU:
Three stage Claus unit with
steam reheaters
Front/side split reaction furnace
Air and acid gas preheat.
In this particular case, it was not
possible to consider co-firing as in
the first case study as the lean acid
gas composition already resulted in
a very low sulphur recovery efficiency. If co-firing was added to the
plant, it would not be possible to
meet the required regulatory
sulphur recovery efficiency of
96.2%. In addition, the plant

quality of the plant. As the raw gas

supplied to the plant was relatively
sweet with a concentration of 1.68%
CO2 and 0.55% H2S, it was determined that there would be some
benefit to convert the amine unit
from DEA to MDEA. As the pipeline specification allows a slip of
2% CO2 into the sweetened gas it
was determined for this particular
plant that for the worst case
scenario it would be possible to
increase the acid gas quality to 22%
H2S. This is still considered a lean
acid gas although of substantially
better quality than the originally
predicted 10% H2S. Therefore as
part of the project to deal with the
turndown of the gas plant, the decision was made to convert the
amine unit to MDEA.
As for the sulphur plant, the existing equipment, including the
existing combustion air blowers,
were substantially oversized for the
new capacity. As a result, Jacobs
Comprimo Sulfur Solutions did not
recommend reusing any of the existing equipment, with the exception
of the existing sulphur collection pit.
The efficiency of the blowers and
continuous blow off to maintain
operation above the surge point
would make the operation very inefficient and high in operating cost.
Compared to the first case study,

Reheater

receives a significant quantity of

methanol in the acid gas which
results in problems such as fouling
of the catalyst beds and poor
sulphur recovery. To limit the
impact of the presence of methanol,
Jacobs recommended the installation of a wash water tower
upstream of the new SRU.
The existing incinerator and stack
were deemed too large to be able to
process the tail gas from the new
SRU. In the case of a stack, it is
essential to maintain the velocities
of the gas leaving the top of the
stack above a minimum to prevent
ingress of rain and also help with
the dispersion of the gases to the
atmosphere. This is important,
especially in a sulphur plant environment as the ingress of rain or
snow can lead to cold spots on the
shell of the stack and thereby lead
to increased corrosion. Jacobs
recommended that the new sulphur
plant have a new incinerator and
stack.
One can conclude from this
particular case study that for some
instances the only option is to build
a new, smaller sulphur recovery
unit. It is crucial though to consider
the impact of the amine unit in the
evaluation of the turndown of the
facility. Switching amine solutions
can provide a better acid gas qual-

Reheater

Reheater
Incinerator

Steam
Waste heat
boiler
Steam

HC
Reaction
furnace

FY

Condenser

Steam

Converter

Steam

Condenser

Condenser
S

Upstream
pressure signal

Converter

Converter

Steam

H2S
2
=
QC SO
1
2

Condenser
S

FrC

Acid gas
preheater

Air
preheater

Methanol
scrubber

Feed gas

Air

Figure 4 Process line-up for the second case study

40 PTQ Q3 2015

www.eptq.com

ity and hence reduce the size of the

SRU, allowing for better operation
and destruction of any of the
contaminants in the acid gas. In
this particular case, a temperature
of 1050C (1922F) was targeted.

Conclusion

As a result of the continued depletion of existing sour wells in

Alberta and the current cost prohibition to develop new sour gas
elds, it is expected that more and
more gas plants will have to evaluate their sulphur recovery units for
future turndown operation. In
order to evaluate the options for
these units, it is imperative to
understand what the limitations of
the existing equipment are and
what the regulatory requirements
are in the case of lower capacities.
In most cases, there is likely an
optimum with respect to capital
and operating expenditure to determine the best modications to the
Anzeige-CS-09-2014b.qxp
15:49 Seiteopera1
plant to be 10.09.14
able to maintain
tion. The modications could be as
simple as only changing catalyst to
requiring the installation of a

complete new smaller SRU next to

the old existing large unit. The
reuse of existing equipment will
largely depend on the known turndown limitations of equipment
which are back burning for burners, channelling for catalyst beds
and sulphur fogging in the case of
condensers. However, even with
turndowns close to 10:1 it may still
be possible to utilise a lot of the
existing equipment. In the end the
combination of acid gas ow, acid
gas composition and the presence
of contaminants will determine
what may or may not be possible.
As well, the presence of acid gas
enrichment in a plant or an existing
tail gas unit could save the day
when it comes to the turndown
situation.

Further reading
1 Royan T S, Loiselle C E, High sulfur recovery
achieved from lean acid gas, The Oil and Gas
Journal, 29 Jan 1979.
2 Basham S, Luinstra E, The effects of
methanol on the Claus Process, Brimstone
Sulfur Symposium, Vail, Colorado, Sept 2000.

3 Ismailian F, Proudfoot G, Ng K, Pushing

the Claus envelope. 10% H2S acid gas with
significant aromatics, LRGCC proceeds, Feb
2003.
4 Paskall H G, Reaction furnace chemistry and
operational modes, Sulphur Recovery Textbook,
Western Research, 1988.

Marco van Son is the Technical Manager of

Comprimo Sulfur Solutions with Jacobs Canada
Inc, Calgary, Canada. He has close to 20 years
experience with sulphur, sour water and
amine technologies and has published several
papers on these topics. He holds MSc degrees
in chemical engineering from the Technical
University Eindhoven, The Netherlands, and in
refining, engineering and gas processing from
ENSPM, Rueil-Malmaison, France.
Shashank Gujale is a Senior Process
Technologist/Specialist
with
Comprimo
Sulfur Solutions, Jacobs Canada Inc., Calgary,
Canada where he has been involved in various
sulphur recovery unit, amine recovery unit and
sour water stripper debottlenecking studies,
and in the development of process design
packages. He holds a bachelor of chemical
engineering degree from the Institute of
Chemical Technology (formerly UDCT, Mumbai
University), Mumbai, India.

your experts
f r burners and combustors
fo

A Member of Unitherm Cemcon Group

CS Combustion Solutions GmbH
Businesspark Marximum
Modecenterstrasse 17/Objekt 4/7/A
A-1110 Vienna Austria
T +43 1 907 44 16
F +43 1 907 44 16 500
sales@comb-sol.com
www.comb-sol.com

n
n
n
n
n

ACID/TAIL GAS
LIQUID SULFUR
SPENT ACID
CHEMICAL/PETROCHEMICAL RESIDUES
OXISPRAY & OXIJET NEW!
Solutions for boosting your acid process

The team of CS Combustion Solutions is experienced

in engineering, supply, commissioning and improvement
of burners and combustors for more than 20 years.

www.eptq.com

PTQ Q3 2015 41

Our focus on Planning

borders on the obsessive.
AltairStrickland doesnt dabble in turnarounds and revampswe specialize in them.
That laser focus involves a process of engineering/constructability/heavy-lift/safety/project
planning that uses AutoCAD, 3D surveys, even full-scale models to explore what-if scenarios
to reduce costs throughout. We want it done right. And nothing will distract us from that.

How can we focus on

your turnaround?
altairstrickland.com
emcor_info@emcor.net
866.890.7794
CRITICAL PROJECT EXECUTION FOR THE REFINERY AND PETROCHEMICAL INDUSTRIES

AS_PTQ_150601.indd 1

5/29/15 1:53 PM

Revamping a hydrocrackers overhead

condenser
A stepwise approach solved seasonal issues affecting plans to increase a
hydrocrackers capacity
MUSTAFA GREN and AHMET BEBEK
Tpra Kirikkale Refinery

he hydrocracking unit is one

of the most important units in
a refinery because of its high
potential for profitability compared
with other units. Tpra Kirikkale
Refinerys hydrocracking unit was
commissioned in 1993 with a capacity of 2305 m3/day.
A capacity test run was
conducted in 2010 to determine the
units potential for increased capacity. The results of the test run
demonstrated that unit capacity
could be increased from 2305 m3/
day to 2500 m3/day. However, the
capacity test run was carried out in
winter when weather conditions
meant that there were no cooling
problems in the fractionator section.
In the summer, the hydrocracking
units feed rate has to be adjusted,
depending on the capacity limit of
the fractionator sections overhead
condenser. In addition, kerosene
and diesel yields can be changed,
depending on demand for hydrocracking operations. When demand
for diesel increases, its yield
increases, but kerosene yield is
reduced as a consequence by leaving kerosene in the diesel draw
tray. In this case, the kerosene
bottom reflux flow rate needs to be
higher while the kerosene top
reflux rate becomes lower and the
amount of heat released from the
column by the kerosene top reflux
is diminished. Previously, this
could not be achieved effectively
because of limited column overhead condenser capacity. This
article explains how those problems
were solved, the steps needed to
achieve this, and how the cooling
capacity of the fractionator column
overhead was increased stepwise.

www.eptq.com

Tempered
water outlet

EC1

3"
3"

E1

3"

3"

Crude unit
clean water
tank
4"
4"

UCO
to tank
UCO
inlet

3"

D1

3"

G1

Tempered
water inlet

4"

Tempered
water outlet

3"
3"

Figure 1 Original tempered water cooler

layout
3"

Phase 1: Using sour water stripper

clean water for the unconverted
oil cooler

The fractionator column bottom

product is called unconverted oil
(UCO); this is the unconverted
component of the feed to the
hydrocracking unit. UCO has to be
routed to its tankage area at
temperatures above 75C in view of
its high freezing point. Originally,
it was pumped to the kerosene

After commissioning
of the new lines,
0.165 Mkcal/h of
energy was saved
by increasing the
temperature of the
desalter wash water
stripper reboiler, then to the fractionator
closed
loop
water
circulation system, or tempered
water cooler. This step was

3"

E1
3"

UCO
to tank
UCO
inlet

Tempered
water inlet
SWS
clean water

Figure 2 Reconfigured tempered water

cooler layout

designed to keep the UCO

rundown temperature stable at
75-80C. The tempered water cooler
is shown in Figure 1; demineralised
water is circulated through a heat
exchanger cooling the UCO product, then routed to an air cooler to
lower its temperature to 50C.
The sour water stripper (SWS)
units clean stripped water temperature is around 45-50C. This clean
water is used as desalter wash water
in the crude unit. After a check of
the design parameters and limits of
heat exchanger E1 in the tempered
water cooler (see Figure 1), it was
calculated that the cooling capacity
of SWS clean water is suitable for
the UCO product heat exchanger.
This new design has two advantages: first, it preheats SWS clean
water to increase the crude heater

PTQ Q3 2015 43

D2 operating temperature
D2 operating pressure

D2 temperature
PV1 pressure

0.6

100

Temperature, C

80
70

0.4

60
50
40

0.2

30
20

Pressure, kg/cm2g

90

10

5/

7/

/2

20

13

3
01

3
/6
30

/2

01

3
25

/6

/2

01

3
/6
20

/2

01

3
15

/6

/2

01

13

/6
10

20
6/

01

3
5/

/2
/5
31

26

/5

/2

01

Figure 3 Fractionation column drum operating temperature, C

2012 mv

2011 mv

2013 mv

110

PV2 opening, %

100
90
80
70
60
50
40
1
45
89
13
3
17
7
22
1
26
5
30
9
35
3
39
7
44
3
48
5
52
9
57
3
61
7
66
1
70
5
74
9
79
3
83
7
88
1
92
5
96
10 9
1
10 3
5
11 7
0
11 1
4
11 5
89
12
3
12 3
7
13 7
2
13 1
6
14 5
0
14 9
53

30

Hours/summer
Figure 4 Pressure control valve opening, %

inlet temperature before it is injected

as desalter wash water. Since the
tempered water cooler is not used
for cooling the UCO product, the air
cooler and pump shown in Figure 1
are not needed to cool tempered
water circulating in the system.
After commissioning of the new
lines, 0.165 Mkcal/h of energy was
saved by increasing the temperature

of the desalter wash water and,

thereby, the crude heater inlet. The
new system, which cools UCO
product with SWS clean water, is
shown in Figure 2.

Phase 2: Using the tempered water

systems air cooler to cool the
kerosene top reflux stream

During the summer, the fractiona-

Design values and new service requirements

Pressure, kg/cm2g
Temperature, C
Flow rate, m3/h
Hazardous area classification
Service

Table 1

44 PTQ Q3 2015

EC-1 design values

12.2
120
31
Zone-2-IIB-T3/H2S
Demineralised water

New service requirements

5.3
110
30-35
IIA-T3/kerosene
Kerosene

tor columns overhead condenser

lacked the capacity to maintain the
overhead reux drum at its design
operating temperature of 71C. The
temperature, pressure and opening
of the pressure control valve PV1 in
the years 2011-2013 are shown in
Figures 3 and 4 for the summer
season, to show the capacity limitation at the overhead cooling
capacity of the fractionator column.
Figure 3 shows that the pressure of
the reux drum was increased to
1 kg/cm2g whereas the design
operating pressure of the reux
drum is 0.73 kg/cm2g. Therefore
the separation index of the column
was inuenced by operating the
column at a pressure higher than
0.73 kg/cm2g. It can be seen from
Figure 4 that opening of the pressure control valve on the reux
drum was nearly 100% because of a
temperature of around 90C
(compared with the design operating temperature of 71C).
After the commissioning of Phase
1, the air cooler of the tempered
water closed loop system could be
utilised to increase the fractionator
overhead cooling capacity. The
kerosene pumparound network of
the fractionator column is shown in
Figure 5. Here, further cooling of
the kerosene top reux can be
maintained to decrease the column
reux ow rate so that the column
overhead ow is decreased. A
management of change study was
carried out in Kirikkale Renery
regarding Phase 2; this involved a
team comprising an inspection
engineer, a mechanical engineer, a
process engineer and operations
engineer. The hazardous area classication of the air cooler was
checked and declared suitable for
kerosene service. The operating
parameters for kerosene service
and design values for the tempered
water air cooler are shown in
Table 1.
The study showed that the air
cooler could be used for this type of
service, and that the metallurgy of
the air cooler is also suitable for
kerosene service. The cooling capacity of the tempered water air cooler
is 0.65 gcal/h and, after using the
air cooler for cooling the kerosene
top reux, the fractionator overhead

www.eptq.com

Sulzer Chemtech
Tower Technical Bulletin

Anti-fouling trays maximize coker main fractionator profitability

Background
The Coker Main Fractionator is systematically subjected to harsh
operating conditions that can lead to deteriorating efficiency and
performance due to coking and fouling. Poor reliability results in
loss of profitable coking margins for the refinery.

strong, uniform liquid flow across the tray deck that moves
the froth and associated coke particles forward, minimizing
stagnant zones that could allow fouling to accumulate. Fixed,
large open area V-GridTM valves are also used for fouling
resistance. SVGTM or LVGTM valves force the vapor laterally
across the deck in a sweeping motion to prevent fouling.

Sulzer VG AF TM with anti-fouling design features

Coker main fractionator general schematic

Extreme Operating Conditions

The main fractionator vapor feed from the coke drum overhead
is characterized by a tendency for further cracking, as well as
the presence of fine coke particles carried from the drums. The
hottest zones of the column (bottom quench, wash, and heavy
gas oil sections) are particularly susceptible to accumulation of
coke, ultimately forcing frequent shutdowns for maintenance.
Internals specifically designed to resist fouling and increase
run length can positively affect unit profitability.
Anti-Fouling Design Strategy
Sulzer VG AF TM trays can be particularly useful in fouling
services like the heavy coker gas oil section, above the wash
zone. While the wash sprays are designed to knock down coke
particulate and quench the unstable feed, coke still often ends
up in the HCGO trays. While an anti-fouling tray cannot prevent
coking, its unique features can substantially delay the onset
of coke-related problems. Specifically, push valves ensure a

VG AF TM trays typically use sloped or stepped outlet weirs

in HCGO service to allow particulates to be swept into the
downcomer from the active zone. Highly sloped downcomers
can also be used to minimize residence time in the downcomer,
thus reducing the risk of further cracking.
These advanced fouling-resistant technologies have been
successfully applied in a variety of commercial applications.
The net result is a significant increase of unit on-stream time
and a reduction in operating and maintenance expenses.

The Sulzer Applications Group

Sulzer has over 150 years of in-house operating
and design experience in process applications. We
understand your process and your economic drivers.
Sulzer has the know-how and the technology to design
internals with reliable, high performance.

Sulzer Chemtech, USA, Inc.

8505 E. North Belt Drive | Humble, TX 77396
Phone: (281) 604-4100 | Fax: (281) 540-2777
Email: TowerTech.CTUS@sulzer.com
www.sulzer.com

Legal Notice: The information contained in this publication is believed to be accurate and reliable, but is not to be construed as implying any warranty or guarantee
of performance. Sulzer Chemtech waives any liability and indemnity for effects resulting from its application.

condenser duty is decreased in line

with the capacity of the tempered
water air cooler. There was one
concern: the hydraulic network of
the kerosene pumparound would be
changed because one air cooler was
added to the kerosene pumparounds hydraulic network (see
Figure 5). Therefore, a final decision
was taken on condition that the flow
rate of the kerosene pumparound
network was checked after the
tempered water air cooler was
commissioned and, if necessary, the
impeller size of the kerosene pumparound pump would be increased
accordingly. A simple scheme of
Phase 2 is shown in Figure 5. The
pressure difference in the new lines
was calculated at 0.9 kg/cm2, based
on 3in lines. Therefore the pressure
difference in the lines is recalculated
with 4in lines and found to be 0.3
kg/cm2g. Based on these hydraulic
calculations, the line sizes were
increased from 3in to 4in in order
not to lose further flow rate from
the kerosene pumparound pump.
The kerosene pumparound flow
is split into two components: the
kerosene top and bottom refluxes.
Further cooling by the tempered
water air cooler was applied to the
kerosene top reflux because the
kerosene bottom reflux directly
affects the tray temperature of the
kerosene draw and the kerosene
product specifications, whereas the
kerosene top reflux directly influences the fractionator column
overhead flow and reflux rate.
Therefore, the decision was taken
to further cool the kerosene top
reflux stream. Hysys was used to
calculate the outcome of using the
tempered water air cooler to further
cool the kerosene pumparound top
reflux. The results obtained from
the model are shown in Table 2.
The column reflux rate and internal
flows are decreased by further
cooling the kerosene top reflux, and
the heat released from air cooler
EC1 is directly reflected in
the capacity decrease from column
overhead condenser EC2. The
final temperature of the kerosene
top reflux before entering the
column, 62C, can also be seen in
Table 2.
In order to establish the necessary

www.eptq.com

To
D1

PV-1

3"

LP flare

4"

4"

EC2

From
G1

EC1

4"

D2

3"

FV-3

FV-1
FV-2

3"

FV-4
C-1 Fractionator
column
To
debutaniser
column

E3
From
kerosene
reflux pump

Figure 5 Kerosene pumparound network before S/D change

line connections to the kerosene top

reflux, the shutdown of the hydrocracking
unit
was
required.
Therefore, the connection points
were changed to make the necessary
connections
without
a
shutdown (see Figure 6). The
connection points, which are
upstream and downstream of the
tempered water air cooler (EC1),
were changed from the configuration in Figure 6 back to the original
orientation (see Figure 5) at the first
shutdown of the unit because of the
reliable operation of FV1.
First of all, the block valves of
the control valve were checked for
leakage, then T connections were
manufactured for the necessary
line connections with their block
valves. Afterwards, the kerosene
top reflux control valve was closed

and all the necessary mechanical

isolations were carried out for one
hour while the two T connections
were installed. After the construction work was completed, tests and
inspection of the new lines were
carried out and the system was
commissioned.

Short test run

A short, two-hour test run was

conducted to investigate the effects
of further cooling of the kerosene
top reflux. The new air cooler was
blocked
and
operation
was
switched back to its previous
condition. The kerosene top reflux
flow rate was to be increased
because the pressure difference of
the hydraulic circuit was decreased.
As can be seen in Figure 7, the
fractionator column reflux flow rate

Hysys results

Frac. column overhead condenser duty, gcal/h

Frac. column overhead flow rate, kmol/h
Frac. column reflux flow rate, m3/h
Kerosen PA total flow rate, m3/h
Kerosen top reflux flow rate, m3/h

Before
4.9
537
51.5
68
38

After air cooler addition

4.3
473
43.2
60
30

Table 2

PTQ Q3 2015 47

To
D1

PV-1

From
G1

3"

LP flare

4"

4"

EC2

EC1

4"

D2

3"

FV-3

FV-1
3"

FV-2

FV-4

E3

C-1 Fractionator
column
To
debutaniser
column

From
kerosene
reflux pump

Phase 3: Increasing the impeller size

of the kerosene pumparound pump

Figure 6 Kerosene pumparound network for normal operation change

Column reflux rate

48

24

46

23

44

22

42

21

40

20

38

19

36

18

34

17

32

16

30
09:36

10:48

12:00

13:12

14:24

Column reflux rate, m3/h

Kerosene reflux rate, m3/h

Kerosene top reflux

15
15:36

Time on 9/12/2014, h:m

48
46
44
42
40
38
36
Kerosene top reflux
Column overhead temperature
10:48

12:00

13:12

Time on 9/12/2014, h:m

Figure 8 Column overhead temperature, C

48 PTQ Q3 2015

14:24

34
32

30
15:36

Kerosene top reflux,

m3/h

Temperature, C

Figure 7 Column reflux flow rate, m3/h

134.0
133.5
133.0
132.5
132.0
131.5
131.0
130.5
130.0
129.5
129.0
09:36

was increased. Since the temperature of the kerosene top reux was
increased, the column internal
liquid and vapour ows tended to
increase. Thus, the column overhead ow rate was increased. This
affects the load of column overhead
condenser EC2 and the column
reux rate was increased to control
the column overhead temperature.
Until the reux ow control valve
stabilised the column overhead
temperature, it was increased for a
while and then returned to its set
temperature value (see Figure 8).
Figure 9 shows that, once the
reux ow rate was increased to
control the column overhead
temperature, the column overhead
ow rate was increased. Since the
column overhead condenser has
limited capacity and is at its maximum load, the reux drum
temperature
was
increased
gradually.

After commissioning of the new

kerosene top reux lines, the air
cooler was commissioned to cool
the kerosene top reux. The total
ow of the kerosene pumparound
was decreased by 8 m3/h and the
pump discharge pressure was
increased by 0.3 kg/cm2. The operating parameters after Phase 2 are
shown in Table 3.
In order to use the tempered
water air cooler effectively, it was
decided that the kerosene pumparound ow should be increased
since the management of change
study determined that the size of
the kerosene pumparound pump
impeller could be increased without a unit S/D. The pump curve
was investigated and it was
concluded that there was room to
increase the pump discharge ow
by increasing the impeller size of
the pump to its maximum diameter. The pump curve and its
maximum impeller size curve are
shown in Figure 10; if the impeller
size is increased to its maximum
diameter, the total ow of the kerosene pumparound increases by
around 5-6 m3/h. After the necessary calculations and checks of the
pressure increase on the kerosene

www.eptq.com

D2 temperature

Temperature, C

48
46

76

44

74

42
40

72

38

70

36
34

68

32

66

:3
6
15

14

Time on 9 Dec 2014, h:m

Figure 9 Once the reflux flow rate was increased, the column overhead flow rate increased
Impeller change

Kerosene PA pump discharge pressure, kg/cm2g

Impeller size, mm
Kerosene PA flow rate, m3/h

Initial condition After Phase 2

6.2
6.5
211
211
68
60

After Phase 3
6.7
218
66

Table 3

60
50
40
30
Max impeller head
H~O
Min impeller head
NPSH
Efficiency min impeller
Efficiency max impeller
BHP

After all steps were nalised, the

heat released from air cooler EC1
was almost 15% of the overall duty
of the column overhead condenser,
which means that the capacity of
the fractionator overhead condenser
had increased by 15% over its
design cooling capacity. The fractionator column reux drum
pressure can be controlled at its
operating design limits, so the
hydrocracker feed rate need not be
adjusted, based on the fractionator
overhead condensers capacity. In
addition, kerosene can be left
in the diesel draw tray and kerosene and diesel yield shift can
be easily carried out without
problems in column operation.

Figure 10 Kerosene pumparound pump curve

Mustafa
Gren
is
the
Operations
Superintendent of the Hydrocracking Unit
complex in Tpras Kirikkale Refinery. He has
eight years refining experience in continuous
catalytic reforming, Benfree and Unifiner units

as a Process Engineer. He has a BSc in chemical

engineering from Bogazici University and a
MSc in engineering management from METU.
Ahmet Bebek is the Production Manager in
the Production Department of Tpras Batman

Efficiency, %

90

8
7
6
5
4
3

80

70

60

NPSHR, m

20

Conclusion

www.eptq.com

:2
4

:1
2
13

:0
0
12

:4
8
10

9:
36

30

Kerosene reflux rate, m3/h

Kerosene top reflux

78

Head, m

discharge lines of the pumparound

pump, it was seen that the design
pressure of the pump discharge
lines were 11.5 kg/cm2g, far above
the pressure after the impeller
size was changed, which was
intended to be 6.7 kg/cm2g. The
expected increase in the pump
discharge pressure was 0.2 kg/
cm2g because of the 5-6 m3/h
increase in ow in the new kerosene top reux lines.
The spare kerosene pumparound
pump was stopped and mechanically isolated. Within a week, the
impeller size, previously 211 mm,
was now 218 mm. After this change
in impeller size, the kerosene
pumparound pump was started up
and the increase in kerosene pumparound ow was seen to be nearly
the same as the calculation. The
ow rate of the kerosene pumparound was increased from 60 m3/h
to 66 m3/h and 6 m3/h was gained
by the change to the pumps impeller size.
The increase in the pump ow
supply cools the column more than
in Phase 1 via EC1. This further
cooling effect gained by the
increased impeller size decreased
the total heat load of the column
overhead condenser by around 0.1
gcal/h, calculated by Hysys.

50
40
30
20
10
0

Capacity, m3/min

Refinery, Batman, Turkey. He has eight years

refining experience in R&D, process technology
and operations. He holds a BSc in chemical
engineering and MScs in process systems
engineering and engineering management.

PTQ Q3 2015 49

SUPPORT GRID
SCALLOPS

OUTER BASKET

CENTERPIPE

OUTLET COLLECTOR

BILFINGER IS JOHNSON SCREENS PRODUCTS

From design to manufacturing and field installation, the Johnson Screens product line from Bilfinger provides an extensive
offering of products and services for the hydrocarbon processing industry.
The Johnson Screens product line offers a complete solution to your hydrocarbon processing industrial needs, including:
Down Flow Reactor Internals
Radial Flow Reactor Internals
Installation of Internals

Hydrocarbon Refining Reactor Internals

Petrochemical Processing Reactor Internals
Gas Processing Column Internals

Please visit our website to see the full line of Johnson Screens products, or find a reactor internals specialist engineer at one
of our global manufacturing plants to assist you in finding the proper solution for your application.
As Bilfinger, we look forward to providing the products you have come to expect when specifying Johnson Screens products.

BILFINGER WATER TECHNOLOGIES

www.water.bilfinger.com
Asia Pacific
Phone +61 7 3867 5555
Fax +61 7 3265 2768
asiapacific.water@bilfinger.com

Europe, Middle East & Africa

Phone +33 5 4902 1600
Fax +33 5 4902 1616
france.water@bilfinger.com

North & South America

Phone +1 651 636 3900
Fax +1 651 638 3281
usa.water@bilfinger.com

Optimising preflash for light tight oil

processing
For a crude unit handling light tight oil feedstock, a semi-preflash column can deliver
significant advantages in energy saving, capacity gain and revamp economics
SOUN HO LEE
GTC Technology

he basic function of a crude

distillation unit is to provide
the initial separation of crude
oil feed into the desired fractions to
feed downstream units. Now,
newly introduced light tight oil
feed processing may limit existing
crude distillation unit capacity. The
addition of a feed preflashing
option may improve capacity and/
or energy efficiency for light tight
oil processing.
Two preflash options which are
commonly implemented for crude
distillation units are a preflash
drum and a preflash column. Each
has strengths and weaknesses.
This
article
discusses
an
improved semi-preflash column
design that optimises the traditional preflash drum and preflash
column options. A case study
includes revamp economics for the
arrangement.

Light tight oil

Light tight oil describes petroleum

fractions produced from low
permeability formations such as
shale or tight sandstone.1 Improved
hydraulic fracturing technology
with
directional
drilling
has
increased light tight oil production
in North America, where processing light tight oil can be a
game-changer for refiners. Light
tight oil slates produced in North
America have typically light and
sweet natures. However, its characteristics vary significantly.
Adapting light tight oil as a new
feedstock may require a significant
retrofit in a refinery layout which
was originally configured with traditional crude oil feedstock. Capacity
imbalance in downstream conver-

www.eptq.com

sion units and insufficient overhead

train processing capacity are already
well recognised issues with refiners.
This imbalance issue is more exaggerated in refineries which are
designed with heavy crude slates
and a high conversion ratio. These
mismatches between new light tight
oil feedstock and current refinery
configurations are initiated in the
crude
distillation
unit
which
provides initial separation of feedstock for entire downstream refinery
processing units.

Implementing crude
feed preflashing
has been one of the
common options to
debottleneck crude
distillation units
There has been extensive research
to try and identify magic juice,
blending ratios of light tight oil and
traditional heavy crude oil slates to
fit
existing
refinery
layouts.
Meanwhile, it has been discovered
that blending between light tight
oil and traditional heavy crude oil
can cause asphaltene destabilisation
due to the paraffinic nature of light
tight oil.2 This destabilisation can
cause unexpected fouling issues.
This article focuses on debottlenecking and enhancing crude
distillation unit capacity and energy
consumption using various preflash
options. Other retrofit strategies for
light tight oil processing are not
discussed.

Crude distillation unit preflashing

Implementing crude feed preflashing has been one of the common

options to debottleneck crude
distillation units. Switching feedstock to a lighter crude slate can
cause undesired feed vaporisation
at the feed furnace pass control
valve and limit capacities in the
crude atmospheric column and
overhead
condensing
circuits.
Adding a preflash drum or column
can reduce the charge rate to the
feed furnace. In addition, transfer
line vibration due to capacity limitation can be resolved through
adapting feed preflashing.
Two preflash options that are
commonly implemented in a crude
distillation unit are the preflash
drum
and
preflash
column.
Another preflash option, GT-IPS is
a semi-preflash column design that
may be optimised between traditional preflash drum and preflash
column options.

Preflash drum option

The preflash drum is the simplest

option for crude preflashing.
Complex modification of equipment
and instrumentation is not required
for this option compared to the
preflash column, making it the most
economical option with regards to
capital expenditure (Capex).
One weakness of the preflash
drum is that this option does not
have the capability to control
preflashed vapour quality in foaming prone service. Therefore, the
preflash drum is prone to have
challenges
with
entrainment.
Entrainment can be accelerated in a
foaming environment. Improper
location of preflashed vapour feed-

PTQ Q3 2015 51

Case study crude distillation unit

Case
Parameter
Feed
Crude slate
Crude charge, BPD
API
-300F norminal cut fraction, LV%
300-450F norminal cut fraction, LV%
450-650F norminal cut fraction, LV%
+ 650F norminal cut fraction, LV%
Operating parameter
Furnace coil outlet temperature, F
Crude atmospheric column top pressure, psi, g
Unit bottom stripping steam rate, Lb/BBL
Unit kerosene stripping steam rate, Lb/BBL
Unit diesel stripping steam rate, Lb/BBL
Fractionation performance
Naphtha D86 (95 LV%), F
Kerosene flash point, F
Kerosene freezing point, F
Diesel D86 (90 LV%), F
Reduced crude D86 (5 LV%), F

Base

100% Eagle Ford

50 000
60.7
58.2
14.7
13.4
13.7
674
24
7.9
12.0
6.2
291
100.8
- 58.0
649
617

Table 1

ing and/or undesirably entrained

heavy oil boiling range materials
can contaminate rundown product
qualities.
Introducing preflashed vapour to
a flash zone is the safest configuration
to
prevent
product
contamination. However, lower
temperature of preflash vapour
containing entrainment can quench
the main feed vapour and can
reduce feed lifting.3
Although recent technology such
as Vortex Tube Cluster (VTC) helps
to discourage the chances of foaming, a preflash drum should be
sized large enough to prevent
foaming. 4

Preflash column option

The
preflash
column
option
produces
preflashed
products.
Although it requires higher Capex
than a preflash drum, this option is
especially effective when the
column top section and overhead
condenser capacities are limited.
A rectifier is a common configuration for a preflash column. Crude
feed is introduced at the preflash
column bottom and preflashed
vapour is rectified by reflux liquid.
Overhead vapour can be condensed
through a crude atmospheric
column overhead condenser if the
condenser has ample capacity. An
exclusive preflash column overhead

52 PTQ Q3 2015

condenser needs to be installed if a

crude atmospheric column overhead condenser is limited. Some
deluxe preflash columns have a
bottom stripping section to control
the front end of the crude atmospheric column top product.
High Capex is a major drawback
of a preflash column. Moreover, a
large plot space is generally needed
to add the equipment.

GT-IPS semi-preflash column option

GT-IPS (Improved Preflash System)

has a semi-preflash column
arrangement compared to the traditional preflash column. Like the
preflash drum option, the semipreflash column does not produce
any independent product. This
column functions as a preflash
column but the additional overhead
condenser and receiver are not
equipped for the preflash column.
Maintaining a desired preflash
vapour quality is one of the most
critical issues. To wash out the
entrained components, some liquid
is withdrawn from the crude
atmospheric column and routed to
the semi-preflash column as reflux.
Then controlled preflashed vapour
is safely fed to the middle of the
atmospheric column.
One of the benefits of the GT-IPS
concept is its flexibility. The
preflash column reflux source and

the preflashed vapour routing can

be varied depending on process
conditions, including crude slate
and target performance, and refinery layout.
To compare the three preflash
options in more detail, the revamping of a crude distillation unit for
light tight oil processing is included
as a case study.

Case study: unit description and

study basis

Eagle Ford, a common light tight

oil in the US, was selected as the
feedstock. Eagle Ford tight oil has
light and sweet characteristics with
approximately 0.1 wt% sulphur
content and is in a highly variable
slate. Published gravity varies
between 40 and 62 API.
The case studys crude distillation
unit receives 100% 60.7 API Eagle
Ford; 50 000 b/d of light tight oils
are charged and heated through
pre-heat trains and furnaces, and
then introduced into the crude
atmospheric column. This column
separates the charged oils to intermediate products: unstabilised
naphtha, kerosene, diesel and
reduced crude (R/C).
The crude atmospheric column
equips three pumparound circuits:
top pumparound, kerosene and
diesel pumparound circuits. A top
reflux stream which is recycled
from the unstabilised naphtha
(overhead distillate) is combined
with the top pumparound stream
before returning to the crude
atmospheric column. The amount
of the top reflux stream can be
adjusted relative to the unstabilised
naphtha boiling range of the
processed crudes. The wash section
is equipped with structured packing while the rest of the sections
consist of trays. Two side strippers
which belong to the crude atmospheric column are operated with
stripping steam. Figure 1 illustrates
the base case configuration of the
crude distillation unit.
To
evaluate
the
optimum
preflashing option, three case studies are conducted and compared to
the base case. Light tight oil charge
rate, quality, and product yields are
maintained among the four cases.
Since
preflash
equipment

www.eptq.com

Offgas

Top P/A
Unstabilised
naphtha
to LER

Kerosene
P/A

Steam
Kerosene

Diesel P/A

Steam

Desalted
crude

Diesel
Steam

Crude
atmospheric
column

Reduced
crude

pressure is floated by crude atmospheric column operating pressure,

in each case the preflash system
pressure is set through a pressure
drop prediction between the
preflash and hydraulic end point in
the crude atmospheric column.5 In
addition, no extra capacities are
available in the crude atmospheric
column overhead condenser and
feed furnace. The base cases crude
atmospheric
column
overhead
condenser and absorbed feed
furnace duties are defined as maximum. Since preflashing options let
down preheat train pressure, an
additional
booster
pump
is
required to transport the remaining
feed to the feed furnace and crude
atmospheric column. Additional
feed booster pump discharge pressure is set to prevent vaporisation
through the entire preheat train
and the feed furnace pass control
valve. Table 1 depicts the process
conditions of the crude distillation
unit.

Case study A: preflash drum

Figure 1 Crude distillation unit: base configuration

Offgas

Top P/A
Unstabilised
naphtha
to LER

Kerosene
P/A

Steam
Kerosene

Diesel P/A
Desalted
crude

Preflash
drum

Steam
Diesel
Steam

Crude
atmospheric
column
Figure 2 Crude distillation unit: case study A preflash drum configuration

54 PTQ Q3 2015

Reduced
crude

Case study A illustrates how

preflash drum implementation is
conducted. A modified configuration is illustrated in Figure 2.
Considering
that
potential
rundown product contamination
can
occur
by
entrainment,
preflashed vapour is introduced to
the crude atmospheric column flash
zone. The preflash drum operating
pressure is determined per the
crude atmospheric column flash
sone pressure and the pressure
drop through the preflash vapour
line circuit.
Preflashing requires a higher
furnace coil outlet temperature to
maintain the current total distillate
yields. Higher internal reflux ratios
for fractionation sections are also
required to achieve target distillate
product yields and qualities. The
preflash drum operating temperature is selected based on reusing
atmospheric
column
overhead
condenser duty. A higher preflash
temperature can increase the
amount of preflashing but it
requires a higher furnace coil outlet
temperature and a simulated overhead condenser duty that exceeds
the current value. Therefore, the

www.eptq.com

position of the preflash drum in this

case study is determined by overhead condenser duty limitation.

Offgas
Offgas

Case study B: preflash column

Case study B gives an overview of

implementing a preflash column.
The preflash column is configured
as a rectifier in this case study, and
is equipped with an independent
overhead receiver and overhead
condenser. The preflash column is
positioned between the preheat
train and furnace to maximise the
amount of preflashing.
The preflash column overhead
vapour streams are split into two
streams. The amount of preflashed
vapour stream charged to the crude
atmospheric
column
overhead
condenser is quantified by the
crude atmospheric column overhead condenser limitation. The
remaining
preflashed
vapour
stream is condensed through an
additional preflash column overhead condenser. The reflux rate for
the preflash column is driven from
the new preflash condenser and
overhead
receiver.
Produced
preflash overhead liquid distillate is
combined with crude atmospheric
overhead distillate and transported
to a naphtha stabiliser for further
separation. A modified configuration for this case is illustrated in
Figure 3.

Top P/A
Unstabilised
naphtha
to LER

Kerosene
P/A
Preflash
column

Desalted
crude

Steam
Kerosene

Diesel P/A

Steam
Diesel
Steam

Crude
atmospheric
column

Reduced
crude

Figure 3 Crude distillation unit: case study B preflash column configuration

Offgas

Top P/A
Unstabilised
naphtha
to LER

Case study C: GT-IPS semi-preflash

column option

Case study C describes implementation of the GT-IPS semi-preflash

column. A modified configuration
for this case is illustrated in Figure 4.
Prior to being sent to the crude
atmospheric column, produced
preflashed vapour is rectified by
the reflux stream recycled from the
crude atmospheric column. The
reflux source for the semi-preflash
column is the crude atmospheric
column overhead distillate in this
particular case study. This reflux
routing configuration is determined
based on the amount of preflashed
vapour and quality. Case study
results show that the rectified
preflashed vapour is suitable for
mixing with the unstabilised
naphtha
stream.
Nevertheless,
preflashed vapour is introduced

www.eptq.com

Kerosene
P/A
Semipreflash
column
Desalted
crude

Steam
Kerosene

Diesel P/A

Steam
Diesel
Steam

Crude
atmospheric
column

Reduced
crude

Figure 4 Crude distillation unit: case study C GT-IPS configuration

PTQ Q3 2015 55

The WINNING

EDGE

AMACS YOUR GLOBAL TOWER INTERNALS PIT CREW

Mass Transfer and Separations Technologies for all your project,
turnaround, and maintenance needs.

WERE AVAILABLE 24/7

TOLL-FREE: 1.800.231.0077 | 24 HOUR EMERGENCY: 281.716.1179 | AMACS@AMACS.COM | WWW.AMACS.COM

Case study results

The three case study results are

summarised and compared with
base case conditions in Table 2.
Since feed preflashing results in a
lower preflash drum or column
bottom temperature, the temperature drop through each preflash
option is quantified. The crude
atmospheric column pumparound
balance for each case is adjusted to
maintain base case product yields
and product qualities. These
temperature drop values through
preflashing and shifted pumparound balances are reflected in
preheat train temperature profiles
and furnace inlet temperature
prediction. Identification of the
cracking tendency of the chosen
light tight oil slate can help optimise furnace coil outlet temperature
further.
The capacities of crude atmospheric
column
distillation
equipment are calculated with
simulated traffic. Calculated capacities are based on the packed wash
section and fixed valve trays for the
rest of the section. The calculated
capacity for each section is plotted
in Figure 5.

www.eptq.com

100

Base
Preflash drum
Preflash column
GT-IPS

90
80
70
60
50
40
30
20

To
p

W
as
h

P/
A
N
a
p
fra h
ct th
io a/
na ke
tio ro
n
Ke
ro
se
ne
P/
A
K
fra er
c t o/ d
io ie
na s
tio el
n
D
ie
se
lP
/A

Distillation equipment capacity,

%

underneath the naphtha/kerosene

fractionation section, taking floating
semi-preflash column pressure into
consideration.
The semi-preflash column operating pressure is determined by the
crude atmospheric column operating pressure and the preflash
vapour line pressure drop. The
column operating temperature is
selected based on reusing atmospheric column overhead condenser
duty. A higher preflash temperature can increase the amount of
preflashing but it requires a higher
furnace coil outlet temperature and
simulated overhead condenser duty
that exceeds the current value.
Like the case study A preflash
drum, the semi-preflash column is
also located in the middle of the
preheat train. However, a higher
preflash percentage through the
semi-preflash column allows the
semi-preflash column to be located
in a higher temperature zone of the
preheat train, compared to case
study A.

Figure 5 Crude atmospheric column distillation equipment capacity

The case study A preflash drum

does not provide significant furnace
duty savings. Moreover, the calculated capacities of this preflash
drum case are similar to those in
the base case. Higher internal reflux
ratios for fractionation sections
erode gain in distillation capacity.
A lower preflash drum temperature
requires the higher discharge pressure of a new booster pump
installation compared to the other
two cases: the preflash column and
the semi-preflash column options.

Table 2 shows that the case B

preflash column option achieves
the highest preflashing percentage
among three cases. Meanwhile, the
highest
temperature
drop
is
observed in the case B preflash
option, resulting in the lowest
furnace
inlet
temperature.
Nevertheless, the lowest net
amount of furnace charge provides
the minimum absorbed furnace
duty. The preflash column also
provides the highest distillation
equipment capacity gain.

Case study result summary

Case
Parameter
Preflash condition
Preflash inlet temperature1, F
Preflash vapour/product %, vol% of feed
Yield structure
Crude charge, BPD
Unstabilised naphtha, BPD
Kerosene, BPD
Atmospheric gas oil, BPD
Reduced crude (R/C), BPD
Operating parameter
Furnace inlet temperature, F
Furnace coil outlet temperature, F
Absorbed furnace duty, MMBTU/hr
Preflash condenser duty, MMBTU/hr
Overhead condenser duty, MMBTU/hr
Top pumparound duty, MMBTU/hr
Kerosene pumparound duty, MMBTU/hr
Diesel pumparound duty, MMBTU/hr
Fractionation performance
Kerosene 5% -naphtha 95%2, F
Diesel 5% - kerosene 95%2, F
Reduced crude 5% -Diesel 95%2, F
Kerosene flash point, F

Base

Preflash drum

Preflash column

GT-IPS

282
16%

395
41%

353
32%

Base
Base
Base
Base
Base

0%
0%
0%
0%
0%

0%
0%
0%
0%
0%

0%
0%
0%
0%
0%

Base
Base
100%
100%
100%
100%
100%
100%

- 14F
+ 31F
97%
100%
100%
86%
76%

- 68F
+ 45F
82%
Additional
100%
45%
100%
100%

- 38F
+ 42F
87%
100%
71%
77%
91%

Base
Base
Base
Base

0F
0F
0F
0F

+ 2F
0F
0F
+ 1F

0F
0F
0F
0F

1. Prior to preflashing 2. ASTM D86 (LV%)

Table 2

PTQ Q3 2015 57

10
EFO: $55/BBL
EFO: $35/BBL
EFO: $25/BBL

Pay-out per year

8
7
6
5
4
3

Conclusion

2
1
0

Preflash drum

Preflash column

GT-IPS

Figure 6 Profitability index payback period

In case C, the GT-IPS semipreflash column option shows

lower furnace duty saving and
distillation equipment capacity gain
compared to the preflash column
option. However, furnace duty
saving and crude atmospheric
column capacity gain are still
substantial compared to the base
case and the case A preflash drum.

Case study: retrofit economic

evaluations

The case A preflash drum option

provides the simplest unit modification among the three cases. The
required plot for the preflash drum
implementation is also smaller than
for the other cases. However, the
aforementioned higher discharge
pressure of the new booster pump
increases Capex substantially.
The case B preflash column
option provides the best performance improvement but it requires
an additional overhead condenser
and receiver for the preflash
column and complex piping modification which influence retrofit
cost. The highest Capex is identified for the case B preflash column
option. The larger plot requirement
for the modification is another
drawback in case B. If space is
limited in the existing unit,
this case is not viable for
implementation.
In case C, GT-IPS does not necessitate substantial addition of
equipment as is required in the
case B preflash column. Also, new
booster pump differential pressure
and capacity requirements in case

58 PTQ Q3 2015

heating values are used for heating

medium consumption predictions.
Figure 6 shows that case Cs
GT-IPS is identified to provide
better retrofit economics compared
to the other two cases. Adding
extra profit through capacity gain
can
further
improve
retrofit
profitability.

C are lower than those in case A.

Calculations of pay-back period
are conducted to gauge retrofit
economics. Gains in crude atmospheric column capacity and/or
furnace/transfer
line
capacity
significantly improve retrofit profitability. However, profits through
these factors vary noticeably per
refinery layout and equipment limitations. Therefore, these factors are
not considered in the pay-back

The GT-IPS semipreflash column can

deliver significant
energy savings and
capacity gains
period calculations and only
furnace heating medium saving is
used for economic evaluations.
Profitability
indexes
are
expressed with regard to pay-back
period. These indexes are shown in
Figure 6. The modification costs are
estimated based on the US Gulf
region. Most US refiners utilise
both refinery fuel gas and natural
gas as furnace heating mediums.
However, the mixing ratio between
two different utilities depends
upon the particular refinery.
Therefore, a single equivalent fuel
oil (EFO) basis is selected. Three
different fuel oil prices are selected
through the last four-year fuel gas
and natural gas trend.6 Typical
furnace efficiency and EFO lower

The technical results and economic

evaluations from three case studies
show that the GT-IPS semi-preflash
column can deliver significant
energy savings and capacity gains
compared to the preflash column
option. Meanwhile, the lower
investment
cost
for
GT-IPS
provides better retrofit economics
compared to the preflash drum
case. These benefits show that
GT-IPS is a viable option for efficient light tight oil processing.
Actual preflashing options and
designs can be customised after
reviewing the specific circumstances of a crude distillation unit
configuration and refinery layout.
Retrofit economics can be precisely
gauged after reflecting actual profit
gains.
This article is an updated version of a
presentation given at AIChE 2015 Spring
Meeting Kister Distillation Symposium, 27-30
April 2015, Austin, Texas.
GT-IPS is a mark of GTC Technology US, LLC.
References
1 Mills R, The myth of the oil crisis: overcoming
the challenges of depletion, geopolitics, and
global warming, Greenwood Publishing Group,
2008.
2 Sandu C, et al, Innovative solutions for
processing shale oils, Hydrocarbon Processing,
July 2013.
3 Golden S W, Prevent preflash drum foaming,
Hydrocarbon Processing, May 1997.
4 Golden S W, Crude unit preflash drums and
columns, PTQ Revamps, 2005.
5 Libermann N P, Troubleshooting Process
Operations, PennWell Books, 4th Edition.
6 US Energy Information Administration Web
Site.
Soun Ho Lee is Manager of Refining
Application with GTC Technology in
Euless, Texas, specialising in process design,
simulation modelling, energy saving design
and troubleshooting for refining and aromatic
applications.
Email: sounho@gtctech.com

www.eptq.com

CRIES-0020_3Q PTQ Ad_Chess_210x297mm.pdf

6/3/15

9:54 AM

choose wisely.
When exploring solutions for emission reduction, the moves you make now can have a large impact on your
environmental compliance and plant operations.
Our team of experts at CRI Catalyst Company, which is part of CRI/Criterion, Inc., the global catalyst technology
company of the Shell Group, offers energy-efficient solutions for catalytic reduction of emissions. CRIs
proprietary Shell DeNOx System, Shell Dioxin Destruction System and the CRI N2O Abatement Technology System
can be designed to achieve emission reductions in excess of 99+%. All systems feature our unique Lateral Flow
Reactor design and low-temperature operation enabling your facility to exceed clean air regulations and reduce
greenhouse gas emissions. Reduction of greenhouse gases could potentially earn valuable carbon credits.
It is all part of our commitment to delivering innovation.

cricatalyst.com

Its time to

LOOK AT PROJECTS DIFFERENTLY

Looking for a way to avoid unforeseen and unrecoverable risks, budget overruns, schedule slippage, and unreliable partners?
With Emersons global project services team providing guidance, you will reduce cost and schedule risk throughout the project.

PROJECT SCHEDULE RISK

$135

65

56%

million

INEFFECTIVE

AT RISK

communication

projects over $1B

around the world

projects under $500M

FAIL

FAIL

For every $1 billion spend on a capital project,

$135 million is at risk. 56% of that ($75 million)
is at risk due to ineffective communication.

35

A project is considered to have failed if the schedule slips or the project

overspends by more than 25%, the execution time is 50% longer, or there are
severe and continuing operational problems into the second year of the project.

2013 Pulse of the Profession, Project Management Institute.

Speed Kills, Klaver, Ali. 2012 Project Manager Magazine.

40 percent of projects in the oil

and gas industry are subject to
budget and schedule overruns.
Capital Project Execution in the Oil and Gas Industry.
M. McKenna, H. Wilczynski, D. VanderSchee.
2006 Booz Allen Hamilton survey from 2006
of 20 companies (super-majors,
independents and
EPC firms).

budget & schedule

OVERRUNS

40

20

30%

cost savings

anticipated value

DISAPPEARS

schedule

Up to 30 percent of anticipated
value disappears during the turnover/
commissioning and ramp-up phases
of new asset lifecycles.
Deloitte. Effective Operational Readiness of Large Mining Capital
Projects - Avoiding value leakage in the transition from project
execution into operations. Article, 2012.

PERSONNEL

39

67

Good front end

REDUCTION
planning leads to
as much as 20% cost
savings and 39% schedule
reduction for total project
design and construction.
Construction Industry Institute: Adding Value
Through Front End Planning. CII Special Publication 268-3.

TO

50

YEARS
expected to

RETIRE

50% of experienced and managerial

personnel in national and international oil
gas processing companies are expected to
retire in the coming decade.
Society of Petroleum Engineers, The Great Crew Change:
A Challenge for Oil Company Profitability, April 16, 2011.

It takes an average of six to seven years to develop

new employees into autonomous petrotechnical
professionals who can make non-standard, original
technical decisions.
2010 SBC Oil & Gas HR Benchmark, Schlumberger Business Consulting
Energy Institute, March 2011.

Bring the most complex projects to

SUCCESSFUL IMPLEMENTATION
www.emersonprocess.com/projectcertainty

E M E R S O N . C O N S I D E R I T S O LV E D .

116349_ins 5 PSS_Platform Infographic ad_PTQ.indd 1

18/05/2015 09:29

Splitting distributed control for a

refinery expansion
A refiners expansion plan required a second distributed control system, achieved
by splitting the original system without a shutdown
PETER HELMERSSON and SVEN WINBERG
Preem AB

reem, with headquarters in

Stockholm, Sweden, is the
countrys largest supplier of
petroleum products, and is responsible for 80% of national refinery
capacity and 30% of Nordic capacity. Total refinery capacity is more
than 18 million tonnes of crude oil
at the companys two refineries in
Gothenburg and Lysekil.
The Lysekil facility, located on
the south western coast, can
process about 11 million t/y, shipping the majority of its output by
sea.
In the past, the refinery (see
Figure 1) had been controlled by a
single DeltaV distributed control
system (DCS) with nearly 15 000
device signal tags (DST), 56 redundant controllers, one wireless
gateway, 28 operator stations and
nine servers. The company has
several projects and expansions in
mind, yet over time it became
apparent that such expansions
would be difficult with a single
system. The best way forward
would be to split the DeltaV system
into two. The challenge was to do
this without a shutdown and without disrupting production.
Consulting with Emerson Process
Management, Preem decided to
split the system into two systems
joined by DeltaV Zones, which is
an architecture designed to operate
and manage large systems. Because
the communication across zones
was process critical, Preem chose to
implement redundant inter-zone
servers. Figure 2 shows two zones,
each representing a DeltaV system,
defined by a ProfessionalPlus
(ProPlus) Station, with necessary
data exchange between domains

www.eptq.com

Figure 1 Preems Lysekil refinery split its single DeltaV control system into two to allow
for expansion

handled seamlessly by the redundant DeltaV Inter-Zone Network

using Inter-Zone servers. Any operator workstation within the unified
architecture can view and control
the areas assigned to that workstation, regardless of domain.
The
immediate
questions
concerned where precisely to make
the split and which controllers,
with their I/O and associated
equipment, would be assigned to
which of the two new systems.
The first consideration in this
process was to avoid splitting apart

any controllers that are sharing

data for control purposes; the ideal
would be to find a logical point
where the interfaces between
controllers are at an absolute minimum. Logically, the split point
should fall somewhere in the
middle of the facility, based on unit
operations.
To aid in choosing the split point,
Preem engineers undertook a study
that included reference visits to
other companies that use Zones
architecture. The refinery is already
divided into two processing areas,

Figure 2 Because the communication across zones was process critical, Preem chose to
implement redundant inter-zone servers

PTQ Q3 2015 61

CADWorx

2015
Plant Professional

UP TO 30%

FASTER

MODELING
& SMARTER

DESIGN
Intergraph CADWorx Plant Professional
makes the creation of intelligent plant
designs quick and easy. It has helped rms
produce the high-quality deliverables their
customers have come to rely on for over
fteen years.

Capabilities
AutoCAD -Based
Intelligent 3D Piping Design
Specication-driven Design
On-the-y Collision Checking
Structural Steel
Equipment
Ducting/Cable Trays
ISOGEN Isometrics
P&ID Creation and Links
Links to Stress Analysis
Design Review

www.intergraph.com/go/cadworx

2014 Integraph Corporation. All rights reserved. Intergraph

is part of Hexagon. Intergraph and the Intergraph logo are
registered trademarks of IntergraphCorporation or its subsidiaries
in the United States and in other countries. AutoCAD is a registered
trademarkof Autodesk, Inc.

62 PTQ Q3 2015

called Area 2 and Area 3, and it

was decided to make the split
between them, with each to get its
own DeltaV system. It was further
decided that the existing system
would be used by Area 3 because it
contained equipment that could not
be stopped during the turnaround,
and the new system would be
assigned to Area 2.

Preparations

Preem commissioned Emerson

Process Management Sweden to
carry out the project, including
planning, factory testing, implementation and site acceptance
testing. In order to keep everyone
up to date, meetings between
Emerson and Preem personnel
were scheduled at two-week
intervals.
The Emerson engineers analysed
the database on the existing DeltaV
system, looking at the configuration
to see which controllers were
exchanging information with others
to try to identify clean break lines
places where the systems could be
split between the two systems without having to do a great deal of
software modification. They were
able to identify a number of these;
in the small number of instances
where this was not possible because
controllers contained modules from
different areas, communications
between one destination system
and the other were run through the
Inter-Zone Network.

Setting up the test system

Such a profound change in the

refinerys control system would, of
course, require complete off-line
testing to ensure that, when implemented, the split would not
produce any unexpected results
that would disrupt production or,
worse, create any safety hazards.
While the configuration changes
to the controllers themselves were
straightforward, the testing was
considerably more about making
sure that everything that was
supposed to happen did indeed
happen, and that the operational
speeds were satisfactory.
Because this was a live split on a
plant running at reduced capacity,
one of the biggest and most impor-

tant tasks was to ensure that data

that was moving between controllers in the original system would be
able to go through the zone server
as they were split into new
networks.
Emerson Sweden set up a test
system at its office in Karlstad. The
user configuration database from
the existing systems ProPlus workstation was exported and the
relevant portions were imported to
the test system ProPlus workstation
(ProPlus 2), which was to become
the workstation for the second
DeltaV system after the split. After
that, the controllers and operator
stations were moved from one
system to the other as required,
and the inter-zone server was
set up.
The next step was to run tests
with a set of field instruments
maintained at the test site, check
out operation of the Foundation
Fieldbus segments, and check the
status of communications.
It was also necessary to verify
whether a problem during operation would trigger a redundant
inter-zone server switch-over to its
standby partner which is
supposed to be completely transparent without creating any
issues for the process. A considerable effort was put into the necessary
configuration changes to make
this work.
Next, the factory acceptance test
(FAT) took place. This was essentially a repeat of the items
previously mentioned, with Preem
personnel observing and signing
off.

Software freeze

One week before starting the split,

a software freeze was implemented.
No further configuration changes
could be carried out because that
could cause unpredictable system
behaviour when the split was
implemented.
A full back-up of the system was
sent to the Emerson office in
Karlstad, where it was cleaned up
to remove any items that were
remaining in the new Area 3 zone,
based on the original system. This
left a clean database to utilise for
the new Area 2 zone.

www.eptq.com

After several iterations and modifications to plans based on

information gathered during the
testing, it was time to move on to
implementation.

Implementation

Emerson was given full responsibility for the split, with the work done
by two of its system engineers and
two system engineers from Preem.
The work was performed during
the first two weeks of a turnaround
period of five weeks. The actual
splitting of the hardware was
completed in the first week, while
the second week was devoted to
cleaning up Area 3 and installing
Emersons AMS Suite predictive
maintenance software in the second
zone. This was done following
what was essentially a standard
installation guide.
This was followed by the site
acceptance test (SAT), which was a
repeat of the FAT using the actual
production system, and about two
days of work to extensively check
out the health of the controllers
and
of
the
communications
between the systems. The team
made sure everything had been
configured properly to account for
the split.

Building the new control room

To house the new distributed

control system, Preem built a stateof-the-art control room in a
hardened, blast-proof building,
which is considerably sturdier than
the one it replaced. Preem also built
two IT-style computer rooms, one
for each half of the new system. In
addition, some of the equipment,
such as the redundant inter-zone
server, is split between the two
system rooms. In many cases, even
if something goes completely
wrong in one of those computer
rooms, there is a standby available
from the alternate system. This
means much of the plant can
continue to operate even if one of
the computer rooms is lost.

Summary

The two smaller systems are more

closely aligned with the operational
philosophy of the facility, increasing the serviceability of the DeltaV

www.eptq.com

system. The risk assessments that

must be performed as part of
project work are confined to a
smaller system containing fewer
plant assets. The new control
system will also make future
expansion easier.

Lessons learned

Several important lessons were

learned during the project, including
the
need
for
detailed
preparation and assessment of the
proper place to split the system.
Preem needed to account for
Guardian Support which proactively provides critical, relevant,
system-specific information to keep
the plant current and licensing
for the new system. And the refiner

Several important
lessons were learned,
including the need for
detailed preparation
and assessment of
the proper place to
split the system
needed to properly track the time
and effort needed to upgrade
graphics to bring in data from the
other zone.
It was also important that Preem
kept Emerson on site until
everything was up and running
smoothly, so there would be no
surprises without the appropriate
people on hand.

Non-intrusive flow
measurement

up to 400 C
Trouble free operation at
extreme pipe temperatures
No clogging, no abrasion,
no pressure losses
Installation and maintenance
without process interruption
Independent of fluid or pressure
Hazardous area approved

Field-Proven at Refineries
Heavy Crude Oil
Atmospheric Distillation
Vacuum Distillation
Coker & Visbreaker Feed
Fluidized Catalytic Cracker
Bitumen

Peter Helmersson is an Automation Engineer

with Preem AB. He has 30 years of automation
experience, previously as an instrument
technician, electrician, and project engineer. He
has 20 years of experience within refining.

refinery.flexim.com

Sven Winberg is an Automation Engineer with

Preem AB. He has 40 years of experience within
refining as an instrument technician, working
with radio communications, as a network
administrator and automation engineer for
15 years.

PTQ Q3 2015 63

You certainly cant be there all the time.

Its a good thing Magnetrol can.

For accurate, reliable level and flow control solutions at the wellhead,
nothing works like MAGNETROL instruments.
With over 80 years of expertise in the oil and gas industry, its no wonder producers install MAGNETROL controls in challenging wellhead
applications. They rely on intelligently designed instruments that run on low power to extend battery life, reduce production cost with
more precise tank capacity control, store data for efficient remote monitoring, and carry the industrys highest reliability ratings.

Make MAGNETROL Level and Flow Control Solutions Part of Your Best-In-Class
Wellhead Automation System. Because MAGNETROL Works.
magnetrol.com

magnetrol.com 1-630-969-4000 info@magnetrol.com

2015 Magnetrol International, Incorporated

Cyber security in the refining world

Protection of industrial control assets requires ongoing risk assessment,
well-defined security policies and an aggressive overall strategy
ERIC KNAPP
Honeywell Process Solutions

eadlines about cyber breaches

in the energy sector have
gained the attention of senior
executives and board members of
companies around the world,
resulting in greater scrutiny of their
own risk management. For many
petroleum companies, however,
while technology and security
demands continue to increase,
skilled resources and time to apply
critical security technology is often
decreasing.
The same advances in process
automation systems for oil refineries and petrochemical plants that
bring higher efficiencies and
increased output can also bring
greater cyber security risks. These
can include concerns about personnel safety, damage to expensive
infrastructure, loss of production
and negative impacts on company
reputation.
By understanding what contributes to the risk of a cyber incident,
petroleum facilities can work to
minimise them and achieve a
higher level of performance and
predictability of process systems
and networks. This approach
includes preventing possible business outages, and diminishing the
threat of lost revenue due to serious safety, environmental and
personnel catastrophes.

Growing security threats

The evolution of cyber threats and

exploitation of data vulnerability
continues to advance at a rapid and
increasing pace. Malware has
become more sophisticated, harder
to eliminate and better at evading
detection. Newly engineered strains
of malware continue to surface, and

www.eptq.com

new variations of existing malware

appear at an amazing rate, as many
as six per second, or over a half a
million per day.1 While there may
be efforts by malicious parties to
cause physical disruption to industrial
operations,
the
sheer
prevalence of malware creates an
inherent risk to any interconnected
system. Whether accidental or
intentional, as automation technology advances and becomes more
and more dependent upon digital
communications, the exposure to a
cyber threat has become a critical
issue
for
petroleum
refiners
worldwide.

The evolution of
cyber threats and
exploitation of
data vulnerability
continue to advance
at a rapid pace
Information is a necessary ingredient to any successful modern
business, and the need to collect
and share information within all
areas of the energy industry is only
going to increase. The automation
infrastructure for refineries and
petrochemical plants is more
advanced and complicated than
ever before. Plant networks are
converging
with
enterprise
networks at an increasing rate. The
unfortunate side effect of this is an
increasingly large attack surface,
with a growing number of potential
attack vectors. It is now easier for a

cyber threat to reach and penetrate

an unprotected industrial system
than ever before.
Information derived from the
industrial automation and control
system (IACS) at facilities in energy
sectors is invaluable to plant
managers to ensure efficient
processes and safe operations.
However, it can be just as valuable
to bad actors, or those malicious
individuals or groups that might
pose a threat. Information can be
ransomed for money. It can be used
to replicate process knowledge and
steal intellectual property. It can be
used to alter or disrupt operations
up to and including catastrophic
damage and the loss of life. Just as
information assists company executives in decision making to advance
their profitability and gain an
advantage in a competitive marketplace, information can be used by
cyber criminals to craft even more
sophisticated and targeted cyber
threats, controlled by even more
expansive and evasive command
and control infrastructures.
In summary, the increasing
dependence upon data acquisition
coupled with the adoption of open
architectures increases the exposure
and vulnerability of a control
network to a wide range of cyber
threats. At the same time, the
threats are increasing in both
frequency and complexity, ranging
from
malware
infections
to
targeted, sponsored attacks.
Increased
vulnerability
plus
increased threat leads to much
greater risk. This risk is significant
and pressing, so understanding it
needs to be a key part of any
successful cyber security strategy.

PTQ Q3 2015 65

Understanding risk requires a

knowledge of threats and vulnerabilities, which means additional
resources will be required to collect
cyber security information, make
sense of it, and act.
When considering the refining,
natural gas and petrochemical
industries,
the
risks
are
compounded by additional factors:
insufficient manpower to manage
security efforts; a greater emphasis
on uptime and reliability; tighter
industry and government regulations; and, of course, an even
greater need for more data accessibility, to drive more aggressive and
demanding business goals.
Another challenge, perhaps one
of the greatest challenges in
improving cyber security practices
for the IACS, is that while the
threats facing the modern IACS are
derived from a common computing
platform shared by many industries, many of the practices and
techniques
used
for
general
purpose cyber security are not
entirely applicable to industrial use.

While specialised cyber security

products do exist for the protection
of industrial control systems, the
development of new tools still lags
that of traditional enterprise cyber
security development. In order to
ensure that the available tools are
used to their maximum potential,
the tools must be thoroughly tested
for efficacy, and special considerations must be made during
deployment in order to ensure
plant reliability.

Mitigating the root causes of cyber

security risk

Despite the increasing sophistication of the cyber threat, the basic

tenets of cyber security are still
applicable. Risk is a function of
threat, vulnerability and consequence, so when the threat level is
high, risk can be reduced by the
reduction of vulnerability and
consequence.
Consequence is best mitigated
through the design of robust industrial control systems. Vulnerability
is best addressed through the

implementation of strong access

and authorisation controls. Strong
passwords, network encryption,
properly configured firewalls and
role-based
authentication
are
among the primary methods to
protect against an exploit reaching
its target. While specific software
vulnerabilities might exist, such as
an application or operation that is
susceptible to remote code execution, there are additional layers of
vulnerability to the software. For
the typical refinery, cyber security
issues can result from:
Lack of defined security zones,
and unsecure conduits/connections
between zones
Unsecured staff access for internal maintenance, retrieval of
historian data, and so on.
Remote third-party contractor
and vendor access
Removable media brought onto
the site, including external hard
drives and even smart phones
Out of date malware signatures
Obsolete or unpatched operating
systems

A reputation for safety

thats recognized
internationally.
Streamlights ATEX and IECEx certified ProPolymer
lights are impact resistant and waterproof.

2AA PROPOLYMER HAZ-LO 3AA PROPOLYMER HAZ-LO 3C PROPOLYMER HAZ-LO

Ultra lightweight and compact

Runs up to 24 hours
Zone 0 ATEX

120 lumens
Ergonomic design
Zone 0 ATEX

150 lumens
Runs up to 18 hours
Zone 0 ATEX

CONNECT WITH US 01.610.631.0600 | WWW.STREAMLIGHT.COM

66 PTQ Q3 2015

www.eptq.com

 Poorly configured, inadequate

or
missing
cyber
security
countermeasures.
Attacks on a facility may involve
only the cyber components and
their operation, but their impact
can extend into the physical, business, human, and environmental
systems to which they are
connected. A cyber event, whether
initiated externally, internally or
due to inadequate policies and
procedures, can lead to a loss of
system control and the corresponding negative consequences.
The reduction of vulnerability is
complicated by the sophisticated
nature of the threat itself. Malware
can often detect and evade known
cyber security controls, and so
layers of defence and compensating
measures need to be taken into
account for a given security control
being circumvented.
The complexity of the cyber security threat cannot be overstated.
The widely reported Stuxnet virus,
which was designed to attack
industrial
programmable
logic
controllers (PLCs), was the first of
its kind, representing a significant
advance in malware complexity.
Since then, increasingly complex
threats have emerged. Flame,
Havex, DyreWolf, Dark Energy,
and other threats have long
surpassed Stuxnet. Malware can no
longer be thought of as an isolated
virus or Trojan, but rather as
complex modular suites of commercial grade malicious software.
Malware today can consist of tens
of megabytes of code, able to
mutate and evade detection, either
on its own or via remote command
and control. Malware is also
increasingly clever, finding new
ways to propagate into and
through industrial environments. It
is no longer effective to depend
upon a few simple cyber security
tools to protect production and
operations.
Therefore, an important aspect of
cyber security for critical infrastructure protection focuses on a basic
understanding and awareness of
real-world risks existing within an
integrated IACS architecture. This
includes potential attacks on the
distributed control system (DCS)

www.eptq.com

and supervisory control and data

acquisition (SCADA) system that
comprise most industrial environments, and impact not only the
common IT infrastructure like
Windows based computers and
network appliances (switches, routers and firewalls), but also
embedded proprietary equipment
such as the programmable logic
controller (PLC), remote terminal
unit (RTU), intelligent electrical
device (IED), basic process control
system (BPCS), safety instrumented
systems (SIS), operator panels, and
ancillary systems.

Improving enterprise protection

Petroleum companies can mitigate

their information security risks by
implementing an overall cyber
security programme for the corporate and plant enterprise. This

It is no longer
effective to depend
upon a few simple
cyber security tools
to protect production
and operations
programme should combine all
aspects of managing security, ranging from the definition and
communication of policies through
implementation of best industry
practices and ongoing operation
and auditing. It should be built
upon a framework that includes
technology (hardware and software),
process
(policies
and
procedures) and people (training
and awareness).
Many firms create cross-functional
teams
to
develop
a
company-wide
cyber
security
programme and ensure the operational and risk differences between
automation and IT systems are
addressed.
Typical
members
include plant managers and operators, control engineers, a member of
the organisations IT staff, and a
risk or compliance manager.
Since industrial cyber security
requires specific expertise and

knowledge of the IACS environment, its often necessary to use

specialists. Some manufacturers
end up stalled because these
specialists can be difficult to find,
and approaching industrial cyber
security with insufficient knowledge could cause more harm than
good. Likewise, using professionals
that arent as well versed in cyber
security may maintain reliability,
but will likely do so at the cost of
strong security. This is typically
where managed industrial cyber
security services are suggested as a
way to bring in the necessary
expertise to optimise programme
results. Unfortunately, once again
there are special considerations,
especially around personnel assurances and the need to provide
secure remote access to the service
provider. If implemented poorly, a
solution can simply be another
attack vector for the ever-evolving
threat. But if executed properly, the
services can take the uncertainty,
complication and effort of industrial cyber security off the task list
of process control engineers and
put it into the hands of process
control security experts.
The most critical step in any
cyber security programme is to
actually take a first step. Once that
step is taken and a plan is developed and put into action, the most
important step is the next step
realising that cyber security is a
cyclical process and not a single
effort, and continuing to move
forward. Cyber security is dynamic
and reactionary and while blanket
recommendations are easy, theyre
not typically the most helpful. For
example, you can start by evaluating vulnerability and risk, then
recommend appropriate changes,
develop design improvements for
specific requirements, implement
the design, and finish up by overseeing continuous management of
the IACS environment. This is a
good process to follow, but the
truth is that theres no such thing
as a one size fits all security
programme. Steps will sometimes
need to be skipped, or revisited
and refined. The important thing is
to understand that it is a cycle, and
you can never stop looking forward

PTQ Q3 2015 67

Figure 1 Continuous monitoring of activity is a key component of situational awareness

to the next step that needs to be

taken, whatever that is. It starts
with an understanding of risk
(because understanding risk allows
any changes to made where theyre
needed most) and understanding
threats (because youll need to be
reactionary if an incident occurs).
An end-to-end cyber security
solution facilitating efficient and
safe operations encompasses:
Assessments and audits For refineries,
assessing
assets
and
vulnerabilities against industry
standards and best practices
provides a roadmap to eliminating
or diminishing revealed areas of
risk. Assessments are often seen as
a necessary first step in the security
lifecycle, but in reality they are a
continuous need. A comprehensive
cyber security assessment can
uncover passive attacks and identify
vulnerabilities
such
as
unauthorised access and other
non-compliance issues, but they
also allow cyber security teams to
track their milestones and the
maturity
of
their
security
programme over time to indicate
progress toward achieving their
desired assurance level. Similarly,
audits are often seen as the last step
to take, but again the need to audit
is continuous. Assessments identify
and evaluate the current security
posture, while audits provide
necessary validation against current
security goals. Together they help
to prioritise efforts and reduce
risks.
Architecture and design Decisions
on architecture design determine
system configurations and impact

68 PTQ Q3 2015

the performance characteristics of a

plant automation platform. For this
reason, refineries need a secure
architecture approach, serving as a
long-term baseline for system availability, reliability and safety.
Architecture and design could be
considered anywhere from the
design and procurement phase of a
greenfield project, to a mitigation
step made in response to a risk
assessment of a brownfield project.
Whenever this step is planned,
plant management should understand the pros and cons of possible

The rapidly evolving

nature of cyber
threats means a
strong security
posture today could
turn into a weak one
tomorrow
cyber security architectures, including
the
advantages
and
disadvantages of each topology
based on proper segmentation of
security zones and conduits.
Network security Network security is the first line of defence in
providing industrial cyber security
to an entire plant or site, but it can
also be the last line of defence in
protecting vulnerable devices. Any
number of potential pitfalls may
arise if it is not implemented properly. An effective network security

strategy will employ cyber security

counter-measures such as firewalls,
threat detection, and security
analytics, but will also establish
policies and procedures that restrict
unauthorised access to system
resources and specify authorised
users, programmes, processes, or
other systems. Controls should be
set up to manage access, including
establishing, activating, modifying,
reviewing, disabling and removing
access privileges. This step can be
taken proactively or reactively, and
should be considered necessary as
it is needed.
Endpoint protection Just like with
network security, the protection of
end nodes within an industrial
control system will be dependent
upon the particular risks being
faced. Every device is a potential
entry point for security threats, so
proactive endpoint protection
such as device hardening and
application whitelisting is necessary to minimise the effective attack
surface. The endpoints are also the
ultimate target of the cyber threat
which we have already established
can be complex and clever, so reactionary counter-measures such as
patching applications and antivirus protection are also required.
Situational awareness Improving
situational awareness is critical
with cyber security because its
impossible to adapt a security plan
without first understanding the
conditions to which it needs to
adapt. Making sure everyone is
able to quickly assess a security
situation is a constant goal and
refinery personnel need ongoing
situational awareness of new cyber
vulnerabilities and the latest
advanced persistent threats (APTs).
This awareness starts with interpreting and understanding activity
on the control network through
continuous monitoring, compliance
and reporting, security analytics,
security information and event
management, and security awareness training (see Figure 1). Without
awareness, it Is impossible to know
which next step should be taken,
and it Is much more likely that a
security programme will stall and
become ineffective.
Response and recovery Response

www.eptq.com

vulnerabilities consolidated into a

single view for better visibility.
These tools help firms determine
the potential impact of a risk and
100
95.7
possible resolutions in88.8
the context
90
of the IACS, so 83.4
plant personnel
have an 80
indication
of remediation
73.4
steps. They
are
also
useful
in trans70
lating complex
cyber
security
60
indicators50 into simple measure46.9
ments,
providing
essential
40
information for ongoing situational
30 and prevention of secuawareness
20
rity incidents.
Performance, %

synthesized
blends
of
and
recovery from
perhaps
the last
commercial
alcohols.
Theseaninhibistep
to be taken
if/when
incitors areoccurs
thermally
and lack
dent
unstable
must also
be
oil solubility.
alsosecurity
highly
considered
as aThey
stepare
in the
acidic which
complicatesoperations
delivery
process.
Industrial
to process
streams
and can
break
should
develop
a sound
back-up
downrestore
sulphide
scales.toTanscient
is
and
strategy
help minia polymeric
phosphate
with
mise
data loss
in the triester
event of
a
a synthesized
cyber
incident custom
or otheralcohol
outage.which
This
increasesencompasses
thermal stability
andand
oil
activity
incident
solubility. recovery
Ethoxylation
of the phosdisaster
planning
to
phate ester
a triester
renders
it
increase
thetosites
resilience
and
low in acidity
which means it is
recovery
capability.
easier
to feed and
not attack
As experience
hasdoes
demonstrated,
the protective
layer need
formed
at the
refining
operations
a strong
metaldynamic
surface.cyber security plan. It
and
The proactively
difference monitor,
is thermal
stabilmust
measure
ity; Table
4 compares
Tanscient
and
manage
cyber security
riskstofora
conventional
phosphate
ester
and a
control
systems,
as well as
real-time
thiophosphate.
visibility,
understanding and deciFigure
4 showstothat
the phosphate
sion
support
tackle
potential
ester loses
overvulnerabilities.
50% of its phosphothreats
and
The
rus within
rst ve minutes
of
plan
must the
alsoreactively
mitigate
exposureand
to high
temperature
and
threats
reduce
vulnerabilities
the thiophosphate
loses 50%
based
on that understanding.
within approximately
30 minutes.
Development
of powerful
risk
Conversely, tools
thermally
stable
management
enables real-time
Tanscient retains
over threats
90% ofand
its
assessment
of cyber
phosphorus after two hours.

security policies and an aggressive

overall security strategy. They must
also remain vigilant, as the consequences of cyber attacks on
99.4heavily
regulated refining
operations
are
91.4
84.6
too great to
ignore. 80.6
73.0
57.5

Reference
1 McAfee Labs Threats Report, Feb 2015.

10

Eric D Knapp is Global Director of Cyber

Security Solutions and Technology with
and 26.7
a recognised
Regular
occurrences
1.8 3.6 4.8 of 6.0 data
2.6 Honeywell
3.0 3.6 Process
4.8 Systems
6.0 13.9
in industrial control
systems cyber
Tanscient
Conventional
breaches prove that
no company Competitive
is expert thiophosphate
ester
years of experience
immune to a cyber attack, and the security. He has over 20 phosphate
in information
technology, specialising
Passivation,
ppm
rapidly evolving nature of cyber
industrial automation technologies,
Type of inchemistry
threats means a strong security
infrastructure security, and applied Ethernet
posture today could turn into a
protocols, as well as the design and
weak one tomorrow.
implementation of intrusion prevention
Figure
5 Corrosion
performance
comparison
different HTCI products
In order
to protect
their
enter- ofsystems
and security information and event
prises,
petroleum
companies management systems in both enterprise and
Performance
in corrosion
protection
phorus. networks.
Figure 5 He
compares
it with
a
require
a proactive
approach
to industrial
currently acts
as the
Tanscient
outperforms
conventional
traditional
phosphate
ester
and
a
cyber
security
that involves
ongo- North
American Technical Advisor to the
HTCIrisk
products
in bothwell-defined
passivation Industrial
thiophosphate
in a Center.
four-hour static
Cyber Security
ing
assessment,

Conclusion0

and maintenance dosages of phos-

test under severe conditions 11.3

T +49 2961 7405-0

T +44 1234 783366
T +65 6702 3707

Made
in
Germany

REMBE | All rights reserved

Your Specialist for

PROCESS SAFETY in the

OIL & GAS INDUSTRY

worldwide

Consulting. Engineering. Products. Service.

Ltd.

www.rembe.de

LI REM 178x125 Anz PS ENG PTQ.indd 1

www.eptq.com

www.rembe.co.uk

Asia Pacific Pte. Ltd.

www.rembe.sg

03.03.15 15:06

PTQ Q2
Q3 2015 85
69

TRI-CON

TRI-CHECK

TRI-BLOCK

WWW.ZWICK-ARMATUREN.DE

THE NEXT GENERATION

TRI-SHARK

TRI-CONTROL

TRI-JACK

Countering marine corrosion in

instrumentation tube
How super-duplex instrumentation tube overcame severe marine atmospheres
at a coastal site
ANDREAS FURUKRONA
Sandvik Materials Technology

orrosive marine atmospheres

at coastal refining and petrochemical sites present distinct
challenges for oil and gas operators, and frequently cause failures
in stainless steel tube. The next
generation of stainless steels are
helping
to
overcome
these
problems.
It has traditionally been the case
that the standard stainless steel
grade 316L is favoured for hydraulic and instrumentation tubing.
Although the material is frequently
used in different marine environments, like offshore platforms or
close to sea water in on-shore
plants, it is prone to pitting and
crevice corrosion on the external
surface of the tube and stress corrosion cracking (SCC).
These are among the main causes
of failure in 316L tube, resulting in
significant economic losses for
operators.
Such issues prompted an intensive study programme by Sandvik
Materials Technology. Based on the
realisation that it is sometimes
necessary to go to extremes in
order to fully understand the
effects of external environments on
materials, Sandvik set out to investigate the performances of various
grades of hydraulic and instrumentation tube in a tropical marine
environment.
A year-long atmospheric study
was conducted by researchers
Katarina Persson and Wenle He to
assess three different tube types,
each bent and attached to fittings,
and to examine the ability of each
grade to prevent crevice corrosion
and
pitting
corrosion
while
subjected to extreme tropical

www.eptq.com

environmental factors. Results of

the test were then subjected to
additional analysis in order to fully
understand how external environmental effects and situational
factors manifest themselves on a
material.
The results would identify a
superior stainless steel replacement
for standard materials like 316L,
and thereby aid development of the
next generation of material grades
capable of supporting the operational demands for tomorrows
demanding oil and gas applications.

Environmental factors

In marine environments, high

concentrations of chlorides can
gather in the deposit on the tube
which has a strong effect on corrosion damage. Additional factors
like high air temperature, high
average relative humidity and
generally
high
precipitation
volumes influence how fast the
corrosion progresses. These factors
in combination are behind the
severity
of
tropical
marine
environments.
Based
on
this
knowledge,
Sandvik determined that one way
to improve the life of tubing in
tropical environments is with
higher alloyed materials, like austenitic
904L.
However,
tests
subsequently found that 904L

exhibits signs of discolouration,

pitting and crevice corrosion after
exposure to an outdoor tropical
environment.
It was therefore necessary to
identify an improvement over 904L,
in the form of high alloy Sandvik
SAF 2507 seamless super duplex
tubing.
Both
materials
meet
European Federation of Corrosion
and NACE International Joint
Publication (EFC) Number 63
industry standards.

Atmospheric testing in Florida

Sandvik performed atmospheric

corrosion tests in a marine outdoor
atmosphere on the east coast of
Florida in the US, directly at the
Atlantic Ocean. The testing would
subject 316L, 904L and Sandvik
SAF 2507 hydraulic and instrumentation tubing to exposure in a
tropical marine environment for a
period of one year specifically
Sandvik 3R60 (UNS S31603),
Sandvik 2RK65 (UNS N08904) and
Sandvik SAF 2507 (UNS S32750).
The location was well suited for a
climate that is considered tropical
and all tests were performed at a
site located 150 feet from the mean
high tide line, close enough to the
sea for the test samples to be
affected by salt spray and mist.
With high salt, high humidity, high
ultraviolet (UV) light levels and

Dimensions of the tubes used for atmospheric corrosion testing

UNS number
UNS S31603
UNS N08904
UNS S32750

ASTM
316L
904L
2507

Outer diameter, mm
12.7
12.7
12.7

Wall thickness, mm
1.24
1.24
1.24

Table 1

PTQ Q3 2015 71

Figure 2 Tube samples 316L, 904L and 2507 showing the side facing
the sea before cleaning

Figure 1 Bent tube samples with fittings

connected to a test rack pictured after one
year of exposure in the tropical marine
environment in Florida

high temperature characteristic of

Florida sea coast conditions the
test site provided an aggressive and
very corrosive environment for the
test samples.
Each steel grade was represented
with a tube sample of length
250 mm and bent by 90, with ends
terminated with attached commercial fittings made of UNS S31603.
The dimensions of the tubes corresponded with ASTM standards and
are shown in Table 1. The tubes
were each examined in the as-

received
surface
finish
produced
from
bright
annealing.
The
three
hydraulic
and
instrumentation tubing grades were
studied in an alumina (or aluminium oxide) test rack that was
designed according to ASTM G-5
(1997)
standard
Practice
for
Conducting Atmospheric Corrosion
Tests on Metals. Each tube was
secured onto the rack using an
ultra-high molecular weight polyethylene material, and the rack was
placed at a 30 angle facing the
ocean where it remained for the
duration of the study.
After a one year period of
outdoor exposure to the coastal
atmospheric
environment,
the
samples were removed from the
rack and returned to the laboratory
for detailed evaluation. The results

would give insight into how the

different grades withstood crevice
corrosion and pitting corrosion.

Results and analysis

After removal from the test environment and the rack, the tubes
were subjected to visual inspection
and examination in a stereo microscope to assess any corrosion
damage on the tubes and their
attached fittings. The inspection
was performed both before and
after cleaning the tube sample
surfaces. Rust was removed from
the tube surfaces with saturated
oxide for one to two hours,
followed by soft brushing in rinsing tap water, and drying in
compressed air.
Figure 1 shows the bent tubes
after one year of exposure.
Examination of the tube samples
revealed notable differences in
corrosion resistance between each

Visit DigitalRefin
ing.com to see
what content is
relevant to you
DigitalRefining.com is the most extensive
source of freely available information on
all aspects of the refining, gas and
petrochemical processing industries,

72 PTQ Q3 2015

providing a constantly growing database

of technical articles, company literature,
product brochures, videos, industry news,
industry events and company information.

www.eptq.com

grade. In each case, the side of the

tube facing the sea showed more
evidence of corrosion than the side
facing land. Red corrosion products
were found on all of the fitting
screwing nuts made of UNS S31603
(see Figure 2).
It was immediately clear that the
samples most affected by corrosion
were the 316L tubes. These showed
corrosion damage with discolouration and some pitting. The 904L
tubes were in comparatively better
condition yet still exhibited discolouration and pitting, although to a
lesser extent than 316L. Pits on
904L were only observed around
the ink mark of the tube.
The Sandvik SAF 2507 superduplex tubing was found to be in
good condition after one year of
exposure. Only some evidence of
discolouration was identified and
this was easily removed after cleaning in water.
Overall, no signs of SCC were
found on any of the tube samples,
which is of note given that each
tube had a 90 bend. Crevices that
exist between tubing, tube supports
or adjoining tube runs are prone to
contamination
and
deposits,
particularly tight crevices in which
cracks often initiate and act as
stress raisers.

Removal of fittings

Because the area immediately

beneath the fitting is particularly
vulnerable to corrosion, Sandviks
researchers removed the fittings
and ferules from each tube sample
to determine whether crevice corrosion had occurred in the space
between the tubing and the fitting.
Corrosion
products
were
observed on the treading of the
fitting and the tube underneath,
with a lot of red corrosion products
at points where the treading had
been in contact with the fitting.
These findings made it clear that
316L material will corrode if
exposed to sea water in severe
tropical atmospheres. Although a
very tight crevice was present, the
corrosion damage did not seem to
affect the integrity of the fitting for
the duration of the test.
The tube parts that were in
contact with the fitting were then

www.eptq.com

Figure 3 Optical microscope images of tube samples after removing the fittings and
ferrules. The images to the left show uncleaned samples, whereas images to the right
show tube samples cleaned in oxalic acid at the position where the ferrule has been: a)
316L tube, b) 316L tube after ferrule removed, c) 904L tube, d) 904L tube after ferrule
removed, e) 2507 tube, f) 2507 tube after ferrule removed

examined with an optical microscope after rust on the treading was

removed. This stage of the examination found crevice corrosion on
the 316L and 904L tubes, but not on

These findings
made it clear that
316L material will
corrode if exposed to
sea water in severe
tropical atmospheres
SAF 2507. This is visible in the right
hand images in Figure 3 and the
impressions after the ferrule was
removed are also clearly visible.

Pitting corrosion resistance

Pitting corrosion refers to the

formation of individual shallow
pits and, in later stages, deep and
sometimes connected pits that are
visible to the naked eye. Pitting can
penetrate deep into tubing walls,
resulting in leaks and tube failures.
Signs of both pitting and crevice
corrosion were found on the 316L
tube, while pitting was only found
around the ink mark of 904L but
crevicing
was
observed.
Meanwhile, neither pitting corrosion nor crevice corrosion was
found on SAF 2507, and from this it
can be concluded that super-duplex
withstood the tropical temperatures
better than the other grades.

Chemical composition

The ability of UNS S31603, UNS

PTQ Q3 2015 73

Chemical analysis, wt%

Grade
UNS S31603
UNS N08904
UNS S32750
....

C max
0.030
0.020
0.030

S max Mn max
0.4
1.7
0.5
1.8
0.8
1.2

Cr
17.5
20
25

Ni
13
25
7

Mo
2.6
4.5
4

Other
Cu=1.5
-

N
0.3

PRE*
26
34
42.5

* The minimum pitting resistance equivalent PRE =%Cr + 3.3%Mo + 16%N

Table 2

N08904 and UNS S32750 to withstand corrosion attacks in the

aggressive tropical environment
correlates well with the pitting
resistance equivalent (PRE) number
of each grade. The value is determined by the content of chromium
(Cr), molybdenum (Mo) and nitrogen (N) within the materials
chemical makeup and calculated as:
PRE=%Cr + 3.3x%Mo + 16x%N.
Standard 904L stainless steel has
a PRE number of 34 which
indicates that it has good resistance
to a range of corrosive environments such as warm sea water and
other high chloride environments.
High carbon (C) content of typically 19-23% is among the elements
that determine the materials PRE
value and enhance its resistance to
corrosion in certain acids. The
element helps initiate a protective
surface film that forms on the tube
surface over a period of time.
The element C can also react with
sulphides
or
ammonia
from
polluted waters, leading to higher
corrosion rates. This negative factor

imbues 904L with corrosion resistance that is intermediate between

the standard austenitic grade 316L
and very high alloyed 6Mo stainless steels; the latter are so-called
because they contain around 6%
molybdenum (Mo) and are consequently more expensive.
The chemical makeup of SAF
2507 is informed by prior experiences with standard 904L and 316L
grades. It has enhanced levels of
Cr, the most common element for
improving a stainless steels resistance to general and localised
corrosion, reduced C content and
increased amounts of Mo and N.
Due to its superior chemistry,
super-duplex has exhibited superior
properties
in
corrosive
environments since the 1980s.
Table 2 shows the chemical
composition of Sandvik SAF 2507
(UNS 32750) alongside Sandvik
3R60 (UNS S31603) and Sandvik
2RK65 (UNS N08904).
The nominal PRE value of SAF
2507 is 42.5, significantly higher
than the equivalent values of 26 for
UNS S31603 and 34 for UNS
N08904. The super-duplex materi-

100
90

Temperature, C

80

CPT
CCT

Conclusions

The year-long atmospheric corrosion test in Florida demonstrated

that Sandvik SAF 2507 has considerably better corrosion resistance
than 904L and 316L in sea water
applications. Overall, the superduplex material was the only grade
in the study that showed acceptable
corrosion resistance in a tropical
marine environment with no pits or
crevice corrosion observed.
The findings are in line with
other reports, and confirm the suitability of Sandvik SAF 2507 as a
superior stainless steel grade for
the kinds of severe marine atmospheres encountered regularly at
coastal refining and petrochemical
sites.
Sandvik, Sandvik 3R60, Sandvik 2RK65 and
Sandvik SAF 2507 are marks of Sandvik
Intellectual Property AB.

70
60
50
40
30
20
10
0

904L

Sandvik 25 Cr duplex 6Mo+N

SAF 2205 (25Cr3Mo.2N) austenitic

Sandvik
SAF 2507

Figure 4 Critical pitting and crevice temperatures in 6% FeCl3 tested for periods of 24
hours (similar to ASTM G48)

74 PTQ Q3 2015

als PRE value has been verified

with ASTM G48, one of the most
severe pitting and corrosion tests
applied to stainless steels in which
a test specimen is exposed to 6%
ferric chloride (FeCl3) solution with
and without crevices.
In a modified version of the test,
ASTM G48 A, a SAF 2507 test specimen, was exposed to test solution
for periods of 24 hours alongside
6Mo and 904L steels. The results
are illustrated in Figure 4.
Despite having a lower content of
costly elements like Ni and Mo,
super-duplex 2507 is capable of
similar performance in sea water to
highly alloyed austenitic stainless
steels like 6Mo. Sandvik SAF 2507
can be considered a more costefficient material than 6Mo grades
and is also better equipped to
handle more severe environments
than 316L and 904L.

Andreas Furukrona is Portfolio Manager for

Hydraulic & Instrumentation, Heat Exchanger
and Automotive-tubing with Sandvik Materials
Technology (SMT) in Sandviken, Sweden.
After graduating with an MSc in mechanical
engineering from Lule Technical University,
Sweden, he joined Sandvik Coromant, which
supplies cutting tools and services, then
became Product Manager for SMT.

www.eptq.com

3rd

REFINING
INDIA

PTQ's international conference for India

held in partnership with Industrial Development Services

Lalit Hotel, New Delhi

12-13 October 2015

Opening Address
B Ashok, Chairman, Indian Oil Corporation

Adding value to hydrogen production, the BOO concept

Air Products

Geo-politics of oil and energy security

Talmiz Ahmad, Former Indian Ambassador to Saudi
Arabia, Oman and UAE

More hydrogen in the refinery through revamping options

KT Kinetics Technology

Achieving operational excellence

Engineers India
Petrochemicals demand scenario
Indian Oil
Value addition through refinery and petrochemical
integration: case study
Engineers India
Maximising bottom-of-the barrel conversion with
commercially proven technologies
Axens
FCC additives for increased residue processing with
high metals feeds
Johnson Matthey Process Technologies
Coke selective bottoms upgrading
W R Grace
Coking and integrated steam air gasification produces
clean flexigas
ExxonMobil Research and Engineering Company
Low cost revamp for processing heavy sour crudes
Technip India

Maximising value generation across a diesel hydrotreater

CRI/Criterion
Improved energy and process efficiency in the naphtha
splitter: a case study
Indian Oil
Demystifying the refinery energy index
Centre for High Technology
Cost effective solution for refinery heat exchangers
Sandvik Materials Technology
Continuous tightening specifications for SO2 emissions
for the refining industry
Jacobs Comprimo
The Impact of SRU plant operation, in particular the SRU
burner, on tubesheet refractory
Industrial Ceramics
Design and materials for improved flue gas tunnel
reliability
Blasch Ceramics
Fired heater modularisation concept, concerns and
benefits
Esteem Projects

www.RefiningIndia.com

SPONSORS:

Real Research

Real Results
Ariel R&D rigorously tests real world
conditions that you face everyday.
Ariel Corporation is the worlds largest manufacturer of reciprocating
compressors with over 50,000 units shipped since 1966.

Every Ariel

compressor is designed, built, and rigorously tested to maximize run-life

and minimize downtime. Ask for Ariel compressors by name and youll find
out why Ariel is the world standard in compression.

For more information about Ariel Compressors in real world applications visit www.arielcorp.com/applications

Energy savings from electric

capacity control
Updated stepless capacity control of reciprocating compressors saved energy
and improved process stability at a petrochemicals site
KLAUS STACHEL
Hoerbiger

he largest refining, petrochemical and agrochemical complex

in the Czech Republic lies just
outside the town of Litvnov, 80 km
north west of Prague and close to
the German border. The plant is
owned by Unipetrol, the leading oil
and petrochemical company in the
Czech Republic and a member of
the Polish PKN Orlen group. The
Litvnov site is also home to a
complex producing polymers and
other chemical products, including
ammonia for use as a fertilizer.
Ammonia manufacture requires
nitrogen, which in this case comes
from a dedicated plant on the site
operated by industrial gases
company Air Products.
Recently, reciprocating compressor technology company Hoerbiger
had the opportunity to help Air
Products and Unipetrol save money
and improve process stability by
installing its new eHydroCOM
capacity control system on a nitrogen compressor (see Figure 1). Both
Air Products and Unipetrol say that
the new control system saves an
estimated annual 330 000 in electricity costs and allows the
ammonia plant to operate more
smoothly.

Looking for power savings

The Air Products plant at Litvnov

liquefies and distills air, splitting it
into oxygen and nitrogen. It is one
of Air Products larger facilities in
Europe.
The refinery uses the oxygen to
produce hydrogen from heavy oil
residues. Some of the hydrogen is
then used in the refinerys hydroprocessing facilities to make clean
fuels. The rest, with nitrogen

www.eptq.com

Figure 1 Installing an eHydroCOM capacity control system for a reciprocating

compressor at Unipetrol Litvinov

added, is converted into ammonia,

an important fertilizer for crops.
In the Air Products plant, nitrogen leaving the air separation unit
at 60 bar is compressed to 324 bar

Electricity accounts
for 70% of the
plants variable
operating costs, so
saving on power is a
key driver
by a Peter Brotherhood (DresserRand) reciprocating compressor
driven by an electric motor. The
nitrogen is piped to Unipetrols

ammonia plant, where hydrogen is

added, and the resulting mixture is
sent to an ammonia reactor.
Electricity accounts for 70% of the
plants variable operating costs, so
saving on power is a key driver.
With a power rating of 2300 kW,
the nitrogen compressor costs
Unipetrol close to 2 million each
year in electricity alone. That made
it an obvious target when Air
Products and Unipetrol were looking for energy savings.
In this case, savings were possible because the original method of
varying the compressors throughput to suit the needs of the
ammonia process was far from
optimal. To run at 75% of rated
capacity, clearance pockets were
opened on the cylinder heads. For
50% capacity, the suction valves

PTQ Q3 2015 77

were held open to reduce the

amount
of
nitrogen
to
be
compressed.
At either 75% or 50% capacity,
this stepwise arrangement worked
well. At any other throughput,
however, nitrogen had to be throttled from the discharge back to the
suction side of the compressor via a
recycle valve, wasting energy
consumed in compressing it. As
well as using unnecessary electricity,
stepwise
control
had
undesirable effects on the downstream process.

Advantages of stepless control

At just the time when Air Products

and Unipetrol were seeking to cut
their operating costs at Litvnov,
Hoerbiger was looking for customers willing to host pilot installations
of its new eHydroCOM capacity
control system. In 2013, Hoerbiger
brought the new system to the
attention of the two companies.
Straight away, they spotted an
opportunity to save money.
Five
years
ago,
Unipetrol
installed Hoerbigers established
HydroCOM
stepless
capacity
control system on five hydrogen
compressors, and was pleased with
the resulting energy savings and
reliability. The prospect of an
all-electric version of the control
system, equally capable yet with
lower project costs, seemed a good
opportunity.
For its part, Hoerbiger believed
the Air Products compressor would
be a suitable candidate for eHydroCOM because of its size and
variable flow rate.
Hoerbigers
original
system
provides stepless capacity control
for
reciprocating
compressors.
Around 1000 HydroCOM systems
are in operation worldwide. The
working principle is known as
reverse flow control. At part load,
the system delays the closure of the
suction valves at the start of the
compression stroke. As a result,
some of the gas in the cylinder is
pushed back out through the
suction valves before the actual
compression starts. This means that
only the amount of gas required by
the process is compressed, so the
compressor uses less energy.

78 PTQ Q3 2015

HydroCOM uses hydraulic actuators to move the unloaders that

hold the suction valves open at the
appropriate points in the compression cycle. A central hydraulic unit
supplies oil to the actuators. The
timing of the actuators is controlled
electronically from a unit located in
the plants switch room, which
communicates with the main
distributed control system.
Old-style capacity control by
recycling discharge gas back to the
compressor inlet, in contrast, is
wasteful because it requires the
compressor to run at full power
regardless of throughput. Stepwise
capacity control systems such as
clearance pockets fare a little better,
but they waste energy unless the
process demand exactly matches
one of the control steps typically
25%, 50%, 75%, and 100%.

Maintenance
is simple, with
comprehensive selftest features, and no
adjustments required
after valve servicing
All-electric stepless capacity control

The new eHydroCOM capacity

control system does everything that
the original system can do, but
with lower project and maintenance
costs. Because it uses electric actuators for the suction valves, instead
of the hydraulic actuators used by
HydroCOM, it is simpler to install
and easier to service. As a result, it
is well suited to small compressors
of a few hundred kW though, as
the Air Products project demonstrated, it is equally suitable for
MW-class compressors.
The system architecture gives
high-level control dynamics, so the
compressors
output
can
be
changed within just three revolutions of the crankshaft. Avoiding
trips caused by incorrect suction,
interstage or discharge pressures is
key to the smooth operation of any
reciprocating compressor. With

control performance no longer

limited by a slow recycle valve,
problems with upstream or downstream equipment are much less
likely to result in a compressor trip.
eHydroCOM can be installed or
retrofitted quickly, because of its
plug and play connections, and
requires no cooling water or
hydraulics. Maintenance is simple,
with
comprehensive
self-test
features, and no adjustments
required after valve servicing.
The system has been extensively
field tested at customer sites including
RAG
Rohl-Aufsuchungs
Aktiengesellschaft
in
Haidach,
Austria, and Air Products Litvnov
plant.

Forecasting cost savings

From information provided by Air

Products, the Hoerbiger team calculated that, on average, 12% of the
compressors nominal flow rate
was recycled through the spillback
valve. With a full-load power rating
of 2300 kW, an electricity cost of
0.11/kWh and a run time of 8000
hours/y, they estimated that stepless control via eHydroCOM would
save Unipetrol at least 228 000
annually.
On top of that figure could be
added savings in the cooling water
supply system, thanks to a corresponding reduction in the heat load
on the interstage coolers. Also not
included in the calculated savings
was an expected gain in the efficiency of a turbocompressor
downstream of the reciprocating
compressor. Thanks to the more
stable suction flow expected with
eHydroCOM, the turbocompressor
would be able to keep its inlet
guide vanes at their optimum position, and hence save energy.
Air Products engineers knew
that demonstrating the new technology would bring some element
of risk. But given Hoerbigers experience following the previous
installation for Unipetrol, and the
large potential savings, they
decided that the benefits were
worth the small degree of
uncertainty.

Meticulous planning

Once Air Products had decided to

www.eptq.com

go ahead, things started well.

According to the companys engineers, the Hoerbiger team members
set up channels for the transfer of
project information and both sides
agreed the scope of work and the
deliverables.
All three companies ran detailed
risk assessments and mitigated the
risks through meticulous planning
of commissioning work, loop
checking, and step-by-step testing
both ofine and online. They also
took care to set up back-up solutions for individual crisis scenarios.
In due course, a project team
from Hoerbiger arrived at Litvnov.
The installation had to be accomplished in a very short time frame,
and the project was completed
successfully.

Savings on target

The eHydroCOM system was

commissioned in May 2014, and
since then the power savings have
been on target. When the Air
Products engineers measured the
specic power before and after the

installation, the gures showed

22% energy savings.
This gure met original expectations for the upgrade, giving
annual savings of about 3000 MWh
plus a safety margin. From this
gross saving, Air Products has had
to subtract the project installation
costs, and probably a small increase
in maintenance costs for the
compressor. On the other hand, the
price of electricity always tends to
rise in the long term.
Improved control response has
also brought worthwhile benets.
This has been especially useful
since Unipetrol has recently been
varying the throughput of the
ammonia plant to a greater extent
than before. With stepwise control,
the varying demand was a challenge, the Air Products engineers
admit, but the eHydroCOM copes
well with the new regime.
Now the compressors output
pressure is more stable, without the
bumps experienced with the stepwise capacity control system. In the
future, this should allow the

downstream turbocompressor to be
operated closer to its surge limit,
which will increase efciency.
The faster control response available is also valuable. Under the
previous contract, Air Products just
had to deliver an agreed amount of
gas at a set pressure. Now the
arrangement is more exible, so the
plant needs to react to follow
production demand. Because the
system is fast-acting, the response
is very sensitive.
Having proved the initial performance, the partners next step is to
show that eHydroCOM is reliable
in the long term. The rst, critical,
priority was to get the power
savings and the right control
response. Now it is important to
maintain that efciency, and get the
benet of it through reliability. Air
Products is monitoring the reliability of the system, and plans to
discuss maintenance programmes
and spare parts with Hoerbiger.
Klaus Stachel is Global Product Manager
Monitoring & Controls with Hoerbiger.

A KEY COMPONENT IN YOUR PROCESS EFFICIENCY:

PROCESS GAS SOLUTIONS FROM AERZEN.

The chemical industry is completely dependent on reliable supplies of process gas. But each process is
different and requires specialized machines. The solution: the efficient process gas compressors and
blowers from AERZEN proven in numerous operating environments, suited for almost all media,
unrivalled in their reliability, and adaptable to individual client specifications. Can the solution really
be this easy? It can!
www.aerzen.com

PG 178x125.indd 1

www.eptq.com

29.05.15 14:05

PTQ Q3 2015 79

REE

und

Relia

t
en

En

so

iro
nmentally

Reliability?

e Effici
bl

The REE Assessment carried out on-site by HOERBIGER senior

compressor experts ranks your equipment using the REE index.
It supports you in choosing the technology needed and in
making business decisions.

www.hoerbiger.com
PTQ_REE_full_page.indd 1

03.06.2015 17:05:05

Revamp of a reciprocating
compressor unit
A step-by-step guide to the revamp of a reciprocating compressor unit to meet
increased hydrogen throughput
ANDREAS HAHN
Neuman & Esser

ew or expanded process
requirements lead to adaptations
for
compressors.
Reciprocating
compressors
are
always tailor made for the original
operating conditions and capacity.
Since processes and/or product
specifications have to meet changes
in operating conditions, it makes
sense to verify existing compressor
equipment to see if it is possible to
modify or revamp it accordingly.
A reciprocating compressor is
designed for a long lifetime.
Revamping offers the possibility to
use an original investment for
changing process requirements and
the means to do this is described
here.
Current European legislation
requires that significant compressor
modifications follow the appropriate directives as well as ATEX rules
for hazardous installations. The
procedures for an ATEX declaration are presented, to demonstrate
the investigations required for
compressors
and
accessories.
Depending on the kind of modification, the installation of a revamp
and the operation of such modified
equipment is allowed only with a
valid original equipment manufacturers declaration according to a
machine directive and ATEX.
In 2003 Neuman & Esser (NEA)
Group delivered a reciprocating
compressor, unit size 2 SZL 320H,
to a refinery in Eastern Europe for
a
desulphurisation
process
compressing hydrogen from 28 bar
suction to 85 bar discharge pressure. The reciprocating compressor
is a two-crank, horizontal, twocylinder, double acting, lubricated
service machine (see Figure 1). It

www.eptq.com

operating conditions and to find

the best solution to meet them. The
jobs definition was to reach the
required capacities even at worst
operating conditions.

Verification of compressor feasibility

As-built and pre-check

Speed
Piston speed
Stroke
Rod diameter
Power at compressor shaft

485 l/min
4.3 ms
265 mm
100 mm
1486 kW

Figure 1 Two-crank compressor

arrangement with two stages and flywheel

has a nominal allowable rod load

of 530 kN.
The compressor is direct driven
by an electric motor rigidly
coupled, with a nominal drive
power of 1700 kW. The original
design capacity was 33 000 Nm/h
at a suction temperature of -5C.
NEAs scope for the compressor
unit included pulsation vessels
(suction and discharge side for each
stage), interstage cooler, and interstage separator up to the check
valve of the last stage.
After five years of compressor
operation, demand for hydrogen
gas increased due to clean fuels
requirements.
An engineering contractor was
assigned to prepare detailed specifications based on the new

First of all, the existing reciprocating compressor was recalculated

according to the original specification, using a compressor design
tool. This verified that the compressor fulfilled the designated process
conditions without deviations, to
avoid general mechanical or performance
problems.
The
main
characteristics of the compressor
dimensions, materials weights and
loads were checked as the basis
for further calculations.
Next, a calculation was carried
out to check the new operating
conditions against the existing
compressor design. This showed
that the capacity required for new
operating conditions could not be
reached with the existing compressor. This led to a pre-check in order
to see if an adaptation of the
compressor for the new operating
conditions was possible. The
compressor itself would have to be
modified, as well as the accessory
equipment.
A positive pre-check result meant
that there were one or more technically feasible opportunities to
match the requirements. There
were
various
possibilities
to
increase the capacity: by compressor speed, by compressor stroke,
and by cylinder bore.

Preparation of bid
The

detailed

verification

of

PTQ Q3 2015 81

Revamp evaluation matrix

Quantifier
(1 ---- 10)
(1 ---- 10)
(1 ---- 10)
(1 ---- 10)
(1 ---- 10)
(1 ---- 10)
(1 ---- 10)
(1 ---- 10)

Project features
Costs
Efficiency
Reliability
Availability
Safety
Vibrations
Emissions

SCORE

Revamp possib. 1
()
(okbetterbest)
()
()
()
()

Revamp possib. 2
()
()
(times
quantifier)
()
()
()
()

Table 1

compressor revamp is done in the

form of an engineering study. In
addition to the job specifications,
compressor
characteristics
and
details have to be checked based on
existing documentation. The accessories also need to be checked for
the new application. To optimise
the revamp solution and to prepare
a revamp proposal, different possibilities have to be evaluated that
will finally result in a tailor made,
technically safe and economically
responsible solution.

Feasibility study and revamp

possibilities
Evaluation matrix

To fulfill requirements, the best

modification option has to be evaluated. All sorts of technically
feasible possibilities are selected
and prepared in an evaluation
matrix (see Table 1). The matrix
shows on one side the different
technical solutions for the compressor revamp and on the other side
different project features. According
to the customers ideas, the project
features are benchmarked with
different quantifiers on a scale 1-10.
Position by position, the revamp
possibilities are evaluated by multiplying the quantifier of each project
feature by the evaluation factor
(ok-better-best) of the technical
revamp possibility. So in the end,
by scoring the most points, the best
solution for a revamp is obtained.
According to the scoring, the
optimum revamp solution can be
selected and used for the following
detailed verification.

Thermodynamics and compressor

calculations
Detailed

82 PTQ Q3 2015

verification

starts

by

implementing the specific compressor

details
regarding
loads,
dimensions and weights into NEAs
compressor
design
tool
KO
(Kompressor-Optimierung Version
3). According to internal upgrades
of different compressors, the year
of construction is essential for the
layout and the allowable loads. For
non-NEA compressors, a similar
NEA compressor size is selected.
The load limits for the existing
compressor have to be adjusted
and the main characters for the
existing compressor have to be
cross-checked and adapted.
The new case data for future
operation have to be added to the
calculation software. To run the
thermodynamic calculation, the gas
analysis, suction pressure, suction
temperature, discharge pressure
and required capacity must be
known for each process or case.
To match the required capacity, it
is necessary to vary the compressor
design figures according to the
selected revamp option. To increase
the capacity of a reciprocating
compressor, in principle the following possibilities are generally
possible, but have to be evaluated
according to actual compressor
feasibilities.
Possibility 1: increase suction pressure Suction pressure could not be
increased
due
to
process
conditions.
Possibility 2: increase speed
Because of the direct coupled drive,
the next asynchronous motor speed
was not feasible, since the allowable medium piston speed was
exceeded. Also, a speed increase
always means a risk of vibration
and
additional
loads
on
foundations.

Possibility 3: increase cylinder

diameter Due to the operating
conditions and allowable rod loads,
it was not possible to increase the
cylinder diameter because the rod
loads would be exceeded.
Possibility 4: increase compressor
stroke Stroke increase is possible
due to the design of the driving
mechanism, without modifying the
cylinder diameters.
Possibility 5: a combination of
alternatives In view of the required
capacity, it is enough simply to
increase the compressor stroke. A
combination of various alternatives
would simply increase costs.
Possibility 6: installation of an
additional compressor in parallel
There was no additional space to
install another compressor.
In this case, the conclusion was to
install a new compressor crankshaft
with enlarged stroke to fulfill the
requirements
within
allowable
compressor limits. The piston rods
and piston had to be replaced but
the cylinders do not need to
change, hence the general compressor arrangement can remain.

Verification of compressor valves

Each variation in process conditions or compressor characteristics

has an impact on the compressor
valves. After detailed calculation of
the compressor layout, the valve
design needs to be confirmed. Since
the valve dynamics have a major
influence on compressor performance, a valve check is mandatory.
Typically, a slight change in the
stroke of the compressor does not
influence the valve dynamics drastically. However, it was reported
that the compressor suffered from
valve problems and lack of delivery
rate. The assumption was that oil
sticking effects could be responsible
for this problem. Therefore, a valve
dynamics calculation was done.
The valve dynamics tool can also
be responsible for oil sticking
effects. It was immediately obvious
that a late closing of the suction
valve due to sticking at the valve
guard takes place. This effect leads
to a reduction of flow and to higher
impact velocities. The effect can be
remedied by stronger spring design
in the existing valve.

www.eptq.com

Intelligence on Duty

PROGNOST-Predictor Early fault detection with Confidence

PROGNOST-Predictor is a true diagnostic system for bearings and gearboxes.
Even inexperienced users benefit from Confidence health status information for each individual
component and their failure modes. Gear and bearing analyses are based on both traditional and
high frequency enveloped spectral analyses, while gear tooth defects are further defined using
Time Synchronous Averaging. The integrated fault frequency database works in conjunction with
the unique Confidence Factor plot providing users with essential lead time when decisions are
due for planning maintenance.
With 30 years of continuous development and installations on multiple machine types,
including planetary gearboxes, PROGNOST-Predictor delivers the early fault detection
reliability managers need.
predictor@prognost.com www.prognost.com

required operating conditions.

Piston-rod neck

Mechanical properties, rod loads,

bearing calculation

Connecting-rod
big end
Connecting-rod
small end
Connecting-rod
shaft
Connecting-rod
bolts
Crosshead
Crosshead bolts
Crosshead pin
Crosshead bearing
contact-pressure
Crosshead bearing
gap
Crosshead bearingrod reversal
Crankpin bearing
Crankcase distancepiece strength
Distance-piece
bolts strength
Distance-piece
bolts gapping
0

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Force
Figure 2 Bar chart for loads on sensitive compressor construction groups (Rod 1)

A new spring design was carried

out with the valve manufacturer. In
this design, the suction valve starts
to close earlier due to the stronger
springs. The valve is closed at

bottom dead centre and the delivery rate is correct.

In the end the valve size could
remain; only the internals had to be
adapted according to the new

Piston-rod neck
Connecting-rod
big end
Connecting-rod
small end
Connecting-rod
shaft
Connecting-rod
bolts
Crosshead
Crosshead bolts
Crosshead pin
Crosshead bearing
contact-pressure
Crosshead bearing
gap
Crosshead bearingrod reversal
Crankpin bearing
Crankcase distancepiece strength
Distance-piece
bolts strength
Distance-piece
bolts gapping
0

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Force
Figure 3 Bar chart for loads on sensitive compressor construction groups (Rod 2)

84 PTQ Q3 2015

After the thermodynamics and

valves are confirmed, the mechanical properties of the compressor
must be verified. For verification,
the safety relief valve settings of
each stage are an important factor
in the compressors layout because
the set pressures determine the
maximum rod forces and static
design pressures.
For compressor layout and verification, the mass and gas forces and
the combined forces (rod forces)
have to be considered. All kinds of
forces have to be verified as being
inside the allowable loading limits.
Typically, this kind of revamp
increases the rod loading of the
compressor. Therefore, the standard procedure is to examine the
compressor by means of a compressor design program. All compressor
parts loaded by the rod load need
to be reviewed in terms of possible
overloading. The design program
KO gives a direct impression of
the load situation for a given case
by generating a bar chart diagram
for each rod.
For reciprocating compressors
there are sensitive construction
groups that have to be checked
separately according to operating
conditions. The most sensitive
compressor construction groups are
shown in the form of a bar chart
diagram for each rod (see Figures 2
and 3). Figure 2 indicates the different loads for each calculation for
Rod 1 and Figure 3 the corresponding loads for Rod 2.
The maximum allowable loads
for NEA compressors are already
available for the KO calculation.
To find a reliable result for
non-NEA compressors, the analogies and specific values for
dimensions, weights and materials
need to be checked and the rod
load limits need to be adapted.
In Figures 2 and 3, the yellow
bars represent the load utilisation
of the individual components in
terms of their loading at idle run,
meaning the compression chambers
are loaded by the suction pressure
of the stage. This is more or less the

www.eptq.com

VISIT US
at the

K O B E L C O

C E N T R I F U G A L

C O M P R E S S O R S

K O B E L C O

C E N T R I F U G A L

C O M P R E S S O R S

Turbo Sym

posiu

m
14th-17th
Booth # 25 , 2015
37!

September

going
where
others
dont

multiple centrifugal compressor solutions for

your most challenging environments
multiple centrifugal compressor solutions for
High-efficiency performance in the largest refineries
yourpower
mostapplications
challengingaround
environments
and
the world, Kobelcos

API compliant Integrally Geared and Single Shaft

Overhang
(DH) compressors
work
in the
harshest
High-efficiency
performance
in the
largest
refineries
flammable,
explosive, corrosive
andworld,
toxic Kobelcos
and power applications
around the
environments.
API compliant Integrally Geared and Single Shaft
Overhang (DH) compressors work in the harshest
Thousands
Kobelco centrifugal
compressors
flammable,ofexplosive,
corrosive and
toxic
have
been in service worldwide in air separation,
environments.
refining, petrochemical, energy services and many
other
applications.
Kobelcos
integrally
geared
Thousands
of Kobelco
centrifugal
compressors
compressor
a perfect
fit for cryogenic
services and
have been inisservice
worldwide
in air separation,
air
separation
services while
keeping
a small
footprint.
refining,
petrochemical,
energy
services
and many
Kobelcos
DH compressor
offers
an opengeared
back-end
other applications.
Kobelcos
integrally
design
to eliminate
powder
build
up and simplify
compressor
is a perfect
fit for
cryogenic
services and
maintenance.
air separation services while keeping a small footprint.
Kobelcos DH compressor offers an open back-end
Let
Kobelco
take you where
others
design
to eliminate
powder
builddont.
up and simplify
Call
today
and
ask
how.
maintenance.
Let Kobelco take you where others dont.
RECIPROCATING
SCREW
Call today and ask how.
SCREW

RECIPROCATING

CENTRIFUGAL

CENTRIFUGAL

Kobelco Compressors America, Inc.

Kobelco DH compressor allows for easy
Houston Office: sales@kobelco-kca.com
maintenance because of its simple design.
713.655.0015
www.kobelcocompressors.com
Kobelco Compressors America, Inc.
Kobelco DH compressor allows for easy
Houston Office: sales@kobelco-kca.com
maintenance because of its simple design.
T713.655.0015
o k y o , J a p a n H o u s t o n , Te x a s M u n i c h , G e r m a n y J u r o n g , S i n g a p o r e D u b a i , U . A . E .
www.kobelcocompressors.com
To k y o , J a p a n H o u s t o n , Te x a s M u n i c h , G e r m a n y J u r o n g , S i n g a p o r e D u b a i , U . A . E .

inertia load acting on the components. Typically, this load is not

critical for low to mean speed
compressors.
The blue bars represent the load
utilisation of the individual components in terms of the loading at
design condition. The compressor
parts are loaded by the gas load
due to design suction and discharge
pressure and by the inertia load.
The red bars represent the load
utilisation of individual components in terms of the loading at
pressure safety valve (PSV) condition on the discharge side.
Another aspect concerns the
different situations for crosshead-pin bearings at full load and
part load operation. It is required
that there is always sufcient rod
reversal available. Typically, the
crosshead bearing is one of the most
critical components in a reciprocating compressor. This bearing can
fail due to excessive hydrodynamic
oil pressure, too small oil lm
thickness, or too little rod load
reversal. The design program
accommodates this by checking
these three failure scenarios individually. The calculation showed that
the crosshead bearing seems to be
slightly overloaded at PSV condition in terms of the minimum oil
lm thickness for this revamp. The
equations behind the load utilisation of the crosshead bearing are
derived from a parameter study
done with the hydrodynamic bearing calculation tool developed some
years ago. The safety margins
included in these equations are
conservatively chosen. As long as
the load utilisation of the crosshead
bearing is below 100%, the design is
denitely on the safe side. On the
other hand, this does not mean that
the bearing is overloaded when the
load utilisation is above 100%. It
just
requires
more
detailed
investigation.
For this, an elasto-hydrodynamic
(EHD) tool is available to assess the
bearing hydrodynamics. This tool is
an integral part of the compressor
design program. The EHD calculation time for the bearing takes only
a few seconds. Assessment of the
results is also very simple since one
only has to compare the maximum

86 PTQ Q3 2015

210.3
196.2
182.2
168.2
154.2
140.2
126.2
112.2
98.14
84.13
70.11
56.10
42.08
28.07
14.05
0.0379

Figure 4 FEA stress plot of the crankshaft

hydrodynamic pressure and the

minimum
hydrodynamic
lm
thickness with the critical reference
values.
A simple representation of the
rod loading by a single load would
simplify the calculation process
since the deformation would be
independent
of
hydrodynamic
pressure. Unfortunately, this simple
representation is not possible
because it would yield such a
dramatic deformation that interference between bearing and pin
would occur. The calculation
process therefore becomes more
complex since the resilience of the
structure needs to be implicitly
introduced in the hydrodynamic
solver.
Deformation reduces the hydrodynamic peak pressure compared
to a rigid structure since a more
equally distributed hydrodynamic
pressure takes place along the
loaded side of the bearing.
Consequently, the minimum oil
lm thickness also becomes less
critical.
According to the calculation, a
slight deformation of the bearing
takes place. Consequently, the
hydrodynamic pressure build-up is
concentrated on the central load
area between the corresponding oil
supply grooves. This is the reason
why, even when there is identical
compression and tension rod load,
the compression side of the crosshead pin bearing is much more
critical than the tension side.
Unfortunately, the compression
load of a reciprocating compressor
is typically higher than the tension
load.

EHD analysis of the crosshead

bearing in terms of hydrodynamic
pressure and minimum oil lm
thickness revealed no critical bearing load for this revamp.

Crankshaft strength

Increasing the stroke of a crankshaft produces higher stress levels

which need to be checked in terms
of fatigue strength.
The crankshaft load is generally
dominated by bending and/or
torsion. Their quantication can be
carried out by making use of
analytical bar models yielding
nominal stress levels. Prospective
critical locations are the llets at the
crank webs. The evaluation of the
stress concentration there is the
decisive task in this context. Due to
the permanently varying rod load
during a crank shaft revolution, the
maximum stress in these llets
changes its circumferential position
and magnitude all the time.
This effect can be best quantied
by utilising nite element analysis
(FEA) models. They are used to
adjust the local stress situation of
the FEA with the analytical bar
model by means of stress concentration factors (SCF). Figure 4 shows
an example of such a model.
Once a sufcient number of FEA
simulations have been performed,
their results can be used to identify
and adjust analytical approaches
which produce approximately the
same results as the FEA. That way,
the individual crankshaft strength
for a given job can be veried most
accurately and quickly in the
compressor design program without the need for intensive FEA
studies.
The compressor crankshaft in
question was modied with an
increased stroke. The torsional load
at the most critical location had
been very low before the revamp
and hardly increased afterwards.
The bending load, on the other
hand, was more dominant in this
application and rose approximately
proportionally with the increase in
stroke. Nevertheless, it remained
moderate and left a robust shaft
that raised no necessity for any
further design adjustment such as
using a higher grade material, or

www.eptq.com

Indicated power, %

stages
into astripping
process
that
typically
increased
delta
at aexchangers.
given
steam
rate or
contents.
Blending
is also
used
reduce
the naphthenic
heat
Based
ondeveloputilities
increasing
the fillet
radii
or to
even
a large
flywheel
is rules
clamped
by
twocoke
of the seven
ofBoth
sedimentation
Parameters
affecting
the
and
microbiological
aspects
of Summary
biofilm-metal
interface.
aero100
References
3
Respini,
Jones,
Spanu,
Sesselego,
Avoiding
foul
seal
drum
off-gas
contains
10-40%

Multiply
the
volume
of
air
by
its
result
in
better
handling
of
fouling,
which
provided
refinery
personnel
runs
at
70-90%
load
of
the
rated
a
higher
hydrogen
in
coke
measurement
through
heat
acid
content
of
the
feed
and
reduce
the
corrosion
rate.
and
design
limits,
the
equipment
is
the
pin
diameter.
rigid
flanges
which
makes
the
drive
microbial corrosion. Step control
bic and anaerobic organisms play ment of biofilms include:
1play,
Troubleshooting
Vacuum
Systems,
Wiley
Hydrocarbon
Engineering,
Novoperating
2006.
S.
The
fatal
concentration
of
H
S
H
density,
to
obtain
pounds
of
air
and
at
the
same
time
increase
with
the
right
information
to
be
capacity.
When
considering
pneu90
balance
calculations.
This
delta
coke
is
especially
detriBlending
two
different
feedstocks
may
lead
to
incomrecalculated
for
the
new
train
susceptible
to
torsional
vibra2
2
Temperature of the system or
Microbial
activitiesReverse
act more
asflow
a drivimportant
role
in the initiation,
control an Energy
Rule
Parameter
Higher
Mitigation
methods
savings
Publications.
is
0.1%.
Fresh
airinhibition
equipment
flow
per
second
HVGO
directly
(less
gasoil
recycled
able
to
constantly
keep
the
unit
at
suction
valve
unloaders
and
mental
due
the matic
high
hydrogen
content
that results
patibility
of80
theaccessories
heavy
crude
constituents
(asphaltene)
in
Verification
of
conditions.
tions.
The
large
flywheel
had
totobe
ambient
temperature
ing
force
for
biocorrosion.
and
of
2
Process
Equipment
Malfunctions,
sediment
Operationpropagation,
Catalyst
Design
Ideal
control
system
Matteo
Virzi
is Senior
TechnologyaMcGrawManager
as
wash
oil).
Figure
17
shows
the
the
best
process
severity
for
any
content
has
increased
when
obtaining
these
samples
is

As
the
N
spillback
valve
control,
fourin
increased
steam
production
in
the
regenerator.
the
more
paraffinic
light
crude
oil;
for
this
reason,
there
Main
motor
For
a
new
flow
or
new
operating
1
Temperature
Yes
Lower
High
asphaltene
conversion
Mix
temperature,
plenum,
cup
installed
with
the
original
machin2
Hill.
Water
past
Microbiologically
induced corro- corrosion in different ways by their
70
with
ISAB flow
Priolo rate
refinery
in the
Italy.surface
He is an
trend
of
wash
oilclaimed
rate.
processed
feed.
by
(25-5%),
you
may
assume
advisable.
2 20%
Incompatible
Yesto requires
Minimise
None
cyclones,include
aromatic
cylinder
with
two
Possible
causes
plugged
or
damaged
steam
are
many
models
predict
3
Linesfor
J quench
R, compressor
Understanding
ejector
systems
Each
compressor
upgrade
conditions,
the sizing
and
ratings
of
ery
in order
to
protect
theHPS
electric

Nutrient
availability
sion
causing
organisms
areincompatibility.
different
energy
deriving
expert
in
distillation
and
thermal
conversion
60
3
Asphaltene
conversion
No
Maximise
High
asphaltene
conversion
Mix
temperature,
plenum,
cup
The
decrease
in
wash
oil
rate
that
the
air
flow
calculated
above
cylinders
per
compression
stage
necessary
to
troubleshoot
vacuum
distillation,
lines/nozzles,
fallen
refractory
restricting
the
steam
High
asphaltene
content,
together
with
incompatibilverification
of
the
accessories.
The
piping
and
vessels
need
to
be
power
supply
grid
against
current
Surface
of the
substratum
sulphate
bacteria
processes.
more
than 20 years of
4
H50
ppreducing
No Limit
ofThe
~1800pathways.
psia
None
Mix temperature,
plenum,
cupWith
2
was
continuously
optimised.
Conclusions
is
about
20%
of sludge
the
non-condensiGraham
Corp.
Technology
Bulletin,
Batavia,
would
be
required
tooperations
enable
at
flow,
orHPS
damage
to
the of
stripper
Damaged
ity,
lead
precipitation
storage
tanks.
new
compressor
power
mustLow
betempin
confirmed.
To
avoid
excessively
fluctuations
which
are
inherent
in mix

pH
water
ininternals.
thearomatic
system
(Desulphovibrio,
experience
inplenum,
technology,
and
Fungi
and
algae
may
also
be
5 will
Coketo Desulphotomaculum,
Yes
High asphaltene
conversion
cyclones,
temperature,
cup,
for quench
40 the
HVGO
was
analysed
on
a
daily
NY,
US.25%
The
examples
cited
here
are
a
tiny
ble
gas
flow.
Therefore,
multiply
least
load
increments
to
run,
steam
lines
maybe
identified
by
measuring
the
flow
ITW
can
help
in
solving
the
above
problem
by
using
automation,
he
holds
a
degree
in
chemical
covered
by
existing
motor
6
Solubility
Yes
None
High
asphaltene
conversion
HPS
cyclones,
mix
temperature,
plenum,
cup,
aromatic
for
quench
high
pulsation,
a
damper
check
is
every
motor
driving
a
reciprocating
and Desulphomonas sp.), iron reduc- involved in metal deterioration. In Effectiveness of biofouling reme4Damage
Putman to
Rfrom
E,stripping
Steam
Surface
Condensers,
basis
with
respect
to
the
level
of
segment
of
the
vast
number
of
the
pounds
of
air
per
second
7
Aromatics
No
Maximise
None
HPS
cyclones,
mix
temperature,
plenum,
cup,
aromatic
for
quench
relatively
efficiently,
at
loads
30
engineering
the
University
of
Palermo.
through
each
line.
internals
may
patented
asphaltene
stabilisers
in
the
crude
blend.
power,
or
the
existing
motor
must
performed
based
on
the
existing
compressor.
So
the
system
is
basiing bacteria (Gallionellea ferrugine fuel and oil storage tanks, fungal dial measures.
ASME
Press,
2001. and 75%. However,
contaminants,
using
the
VCI
tech- problems
that
steam
ejector catalyst
calculated
above
byAPI
approximately
Email:
mvirzi@isab.com
between
70%
cause
a restricted
flow
resulting
theresult
need
to
Additionally,
sludge
accumulates
intoa the
tanks,
be
adapted
too.
618
Rev.5
vessel
design
and according
to the
cally
two-mass
oscillator
capable
20 when
Biofilm
formation
isinfor
the
of
and
Ferrobacillus
sp.),
acid
producspecies
such
asbeset
aspergillus,
penicil5
Applied
process
design76%
chemical
&
nique
that
measures
the
coke
Table
1
and
surface
condenser
perforfour
between
100%
and
load,
the
open
up
the
spent
catalyst
slide
valve
for
the
same
patented
ITW
Online
Cleaning
can
be
used
to
remove
requires
sufficient
motor
power
for
limits
in
API
618.
The
damper
of
torsion
within
the
motor
shaft
not
ing
bacteria
(Pseudomonas, lium and
fusarium may grow on an
Marcoaccumulation
Respini isplants,
a Seniorprocess
Technology
Expert
10
petrochemical
Ejector
& Vacuum
3,4,5
particles
entrained
within
the
isandalways
anIf this

If the
molecular
weight
of and
the
dispensable
process
gas
still
needs
historical
flow.
there
a high
likelihood
precipitated
sludge,
without
the
need mance.
to open
the There
tank flywheel
each
stage
against
safety
check
can
be
runis
with
a KO
calcusection
between
and
rotor
withoccurs,
Baker
Hughes
Downstream
Chemicals,
necessarily
uniform
in
time
or
Aerobacter,
and the
Bacillus),
components
produce
Systems,
218-222.
Selection
and
Compatibility,
NPRA
Ann.relief
Mtg. fuel
Correlations
from a Commercial
Unit,
1988
Robert Ohmes
is a Principal
Consultant
0 isnot
HVGO
(thus
removed
by
wash
understandable
temptation
to
cracked
gas
different
from
the
to
be
recycled
through
the
spillspecialising
in
refinery
and
petrochemical
that
there
is
increased
flow
on
one
side
of
the
stripper
and
require
manpower
to
manually
clean
it.
This
will
valve
set
pressure,
plus
a
5%
safety
lation
and
considers
the
new
core.
The
natural
frequency
of
this
Advanced
that starts
immediately
sulphur
acids
0
10
20 bacteria
30
40 carboxylic
50 Ann
60 Mtg703/20-22/88,
80 which
90
100corrode space
2011 AM 11-46. oxidising
NPRA
AM-88-59.
with
KBC
Technologies,
Houston,
oil). or weight
consult
the equipment
vendor
for
molecular
of Unit
air,reduce
multiply
back
In
this
up
processvalve.
improvements
fouling
control.
less
flow
on another
side.
The
mass
fluxcase,
rate
in
downtime
and
the
margin.
compressor
layouts
overall
is not%affected
by aand
changed
crank
after
immersion
ofinworst
metal
in
the
(Thiobacillus
sp.). H-Oil
iron.
2avoid
Sayles et al, dramatically
Texaco
Commercial
Capacity,
Texas. His
primary
responsibilities
are
profit
Also,
the
rate
ofto
antifoulant
on advice
2 engineer
He
has
15
years
of
refining
experience
and
and
guidance,
when
trying
the
pounds
of
flow
by
the
ratio
of
to
24%
of
the
rated
compressor
Norman
Lieberman
is
a
chemical
.
If
the
conventional
strippers
is
about
1650
lb/min/ft
concerns
related
conventional
tank
cleaning
processes
and
operating
cases.
The
shaft
stroke.
It
is
only
the
torsional
Operation,
Japan Pet. are
Inst.ubiquitous
Pet. Ref. Conf.
environment.
The growth
Microorganisms
in Scott Sayles was a Principal Consultant aqueous
improvement
initiatives, hydroprocessing
unit
the wash
oil was optimised
and to
is currently
in are
developing
new
troubleshoot
steam
ejector are
the
molecular
of rates
the
who
specialises
inthere
troubleshooting
power
isandinvolved
Additionally,
stripper
damaged,
temporary
techniques.
(10/19-20/88).
Torsion
analysis
result
provides
first
information
excitation
that rises
to
some nozzles
extent.
with KBC
Advanced
Technologies,
Houston,
Role
of
biofilms
consulting,
organisational
consulting
of
biofilm
iswasted.
considered
to refinery
be for
a
nature
and
grow atweight
very rapid
technologies fixed
forprocess
improving
refinery conversion
when
needed.
non-catalytic
equipment
and
in
performance
problems.
Typically,
non-condensible
gas,
divided
by
Figure
6sludge
Comparison
of
the
compressor
power
consumption
between
step
control
andhelp,
installed
clearance
3increased
Kunnas
et al,
Mitigate
fouling
in ebullatednozzles
that
opening
up
the
reactor/stripper
The
isand
transformed
into
aitfully
reusable
prodTexas,
with
than
35
years
of
For
direct
coupled
compressors,
domestic
and
international
clients. pockets
He
holds
whether
the
existing
vessels
are
still
Therefore
the
torsional
vibration
The
termmore
biofilm
refers
torefinery
the but
result
of
complex
processes
involvin
soil,
water
air.
They
show
processes.
With design.
extensive
experience
in
retrofit
process
He
has
taught
his
reverse
flow the
control
as
aProc.,
function
oftorsion
the
compressor
capacity
this
does had
not
help.
Only
field
the
molecular
weight
of10/10.
air
(29).
would
situation
bed
hydrocrackers,
Hydro.
and
ranging
from
degrees
from
Kansas
State
University
(BSChE)
andrechecked
performing
istransport
theimprove
best
solution.
uctmandatory
and
recovered
sludge
can development
be petrochemical
reprocessed
is
to
run
avarying
new
sufficient
or
tooproblems
small.
simulation
tomicrobial
be
forrepairs
ofexperience,
commuing
of the
organic
and
extreme
tolerance
to
enviasphaltene
relatedSeminar
in oil production
Troubleshooting
since
1983
to
Conclusions
4without
Sayles,
Romero,
Understand
theset
differences
investigation,
experience,
careful
See
Table
for
an
of
test
but
the
control
logic
refinery
plant
manager
to
research
engineer.
andThe
Tulane
University
(MBA)and
and KO
is
a licensed
any
concern.
analysis
so1conditions
that
theactual
components
of nities
damper
check
by
softthe
revamp
and
everything
was
on submerged
surfaces
in slightly,
inorganic
molecules
microbial
ronmental
such
as
acidic
and refining,
he
isengineers
an inventor
of
fiveexperienced
US patents
over
18 become
000
and
VisTec
is a markareas
of and
Baker
Hughes
Incorporated.
This
article
describes
annumber
example
of of
between
thermal
and
hydrocracking,
Hydro.
A
His
technical
of
expertise
include
professional
engineer
in
Louisiana.
measurements
experimentation
results
from
a
low
sulphur
refinery
would
more
complicated
automation
allows
for
precise
of
the
compressor,
the
of
Charles
Radcliffe,
Senior
Technical
Service
Engineer,
the
drive
train
can
be
verified
to
ware
only
gives
preliminary
results
found
to
be
within
acceptable
to published
the surface,
of
and alkaline pH, low and higher aqueous environments. Biofilm cells
and has
10
technical
papersbooks,
and
plant
operators.
first
ofadsorption
his eight
high9/11.
conversion
visbreaking
with plant
Proc.,
operationcontrol
and
ebullated
bed
residual
Rick Manner
is The
a pulsations
Principal
Consultant
with
appear
to design,
offer
hope
of ofa Matthey
vacuum
and
access
the
compressor
for
fine-tuning
cylinders
per stage
can be
Europe,
Johnson
Process
Technologies,
Charles.
avoid tower.
torsion
vibrations
for
after
the
levels.
theandany
physic-chemical
molecules
totothe
surface
and
initial
temperatures
as shown
in reduced.
Tableand
1. influences
sevenexpected
conference
papers
onOperations,
visbreakers
and
Troubleshooting
Process
has
increased
runa calibration
length
when
5Q
Sayles,
Bailor,
Catalyst
addition
in ebullated
hydrocracking,
hydrotreating,
FCC,
and maintenance
KBC Advanced Technologies,
Parsippany,
New
successful
outcome.
The
ejectors
If
using
as
gas
is process
impeded.
However,
parameters,
such
as and
gas
Hence,
theairhad
capital
costs
footRadcliffe@matthey.com
We
have
a fall-off
in and
steam
stripping
efficiency
in
our
compressor
damage.
heavycontinuously
fuel oilThe
stability
problems.
Agives
graduate
revamp.
a
interactions
between
metal
attachment
of
microbial
cells
Microorganisms
and
their
metabeen
insoftware
print
for 35
years.
He
bed
units,
PTQ,
Q2
2005.
practical
understanding
of
most
processes.
Jersey.
He
has
been
with
KBC
for
over
27
compared
to
typical
visbreaker
run
should
and
will
perform
on
the
unacceptable,
argon
or
helium
may
two
cylinders
per
stage
are
still
pressure,
hydrogen/feed
ratio
and
print
of
the
compressor
are
of
Milan
University
with
a
degree
in
industrial
References
FCC
unit
with
resulting
carryover
of
hydrocarbons
into
the
The
most
likely
cause
for
a
decline
in
stripping
effiGas
coolers,
vessels,
piping
In
this
revamp
case,
because
of
its
proposal
for
pulsation
damper
environment,
frequently
enhances
by their
irreversible
adhebolic
products,
including
previously
worked
for Amoco
in Whiting,
6lengths.
Sayles,
Ebullation Factor,
Hydro.enzymes,
3/06.
He is a member of
the American
Fuel and followed
years, primarily
specialising
in Oil
hydrocracking,
manufacturers
performance
curves.
be
employed,
but
isEng.,
acranktime
chemistry,
he
has
been
Research
Fellowand
in
required
asTexas
compared
with
stepless
flow is
rate.
minimised.
Furthermore,
fewer
1 Petralito
G,and
Respini
M,
Achieving
optimal
regenerator.
Causes
andthis
solutions,
please.
ciency
can
be
divided
into
two
areas:
If
there
an
increase
capacity
short
length,
the
compressor
design.
Indiana
and
City,
through
1980,
corrosion,
leads
toinholds
deteriorasion
facilitated
byaprocesses,
production
of
exopolysaccharides,
organic
and gas
7 Wisdom,
Characterization
of
Asphaltenes
Petrochemical
Manufacturers,
a
BS
hydrotreating,
hydrogen
hydrogen
This successful
result
was
made
theFinally,
field
ofwas
organometallic
catalysts
andGood
is a
visbreaking
severity,
PTQ,
Q1,
2010.
The
main
ingredient
needed
to
consuming
option.
But
take
care
to
control.
The
following
example
explains
cylinders
mean
reduced
mainte1.
Process
related:
until
1983
Refinery
Manager
for
the
and
power,
the
coolers
have
to
be
shaft
is
barely
sensitive
to
torsional
for
detailed
verification
the metal.
Biofilm
inorganic
acids, Mogas
and 3rd
volatile
During
HydrobyConversion,
Eb Bed tion
degreeof
in chemical
engineering
frominduces
Michigan extracellular
plants, reforming,polymeric
isomerisation substances
and alternate
possible
coupling
Baker
Hughes
registered
professional
in Italy.
He
is
2 A Phase
Separation
Model
for Coke
A
achieve
this
result
is
the
dedication
select
screwed
fitting
size
that
The
same
application
with
a effect
stepthe
impact
on
the
compressor
nance
costs,
shorter
shutdown
Hope
refinery
in
Louisiana.
He
from
CliffaConf,
Avery,
Global
FCC
Process
Advisor,
Albemarle,
a.
An
increase
in
pressure
will
have
achemist
significant
verified
for
theinKinetic
new
operating
vibrations.
This
does
not
signifiand
for
providing
agraduated
warranty
many
changes
the
type
and
(EPS).
Once
attached,
the
organisms
compounds,
such
as
ammonia
or
Hvy
Oils
July
2007.
Technological
University
and
an
MS
degree
in
fuels
technology.
He
holds
BS
and
MS
degrees
in
VisTec
anticoke/antifoulant
treatalso
a
member
of
ACS
and
NACE.
Formation,
Preprints
ACS,
Div.
Pet
Chem,
38,
and
theengineering
determination
of
field
will
notchange
greatly
impact
the
vacuum
Cooper
Union
in New
York
City
in overhead
1964.
less
full-range
flow
control
system
based
thethe
chosen
periods
forsulphide,
and
fewer
cliff.avery@albemarle.com
onon
stripping
efficiency.
Possible
causes
are
conditions.
These
detail
checks
cantly
after
increasing
the arrangement
regarding
vibrations
according
to
8
McDaniel
etoverhauls,
al,
AMOCOs
LC-Fining
chemical
Lamar
University,
chemical
from
the
Massachusetts
concentration
of from
ions,
pH
values,
begin
toengineering
produce
material
termed
hydrogen
can
alter
elecEmail: norm@lieberman-eng.com
marco.respini@bakerhughes.com
428-433, 1993.
mentsparts
with
monitoring
technology,
Email:
troubleshooter.
tower
operating
pressure.
Also,
the
would
require
only
one
cylinder
capacity
control
system.
A
reciprospare
to
be
kept
in
stock.
The
Poorer
stripping
efficiency
is
generally
measured
as
line
coking,
fractionator
flooding,
and
wet
gas
Resid
Hydrocracking
Yield
and
Performance
Texas.
need
to
be
done
by
the
original
Institute
of
Technology.
stroke.
Between
the
compressor
API
618
and
NEA
workshop
standextracellular biopolymer, or slime
trochemical
processes
at
the and oxidation reduction potential.
stage (two
cylin-a
compressor
shall compress
higher
reduceshowever,
electric cating
equipment
manufacturer
of the per
and the efficiency
induction motor,
ards,compression
it is necessary
to run
power costs, and the high degree process gas in two compression ders in total for the compressor)

A controlled decrease
in wash oil can result
in better handling of
fouling, and at the
same time increase
HVGO directly

alves
Best V 67
8
since 1

www.eptq.com
38 PTQ
PTQ
Q2 2014
2014
www.eptq.com
www.eptq.com
www.eptq.com
98
Q3
www.eptq.com

hoerbiger.indd
5
baker hughes.indd
8
bpcl6.indd
2

PTQ Q2 2015 97
www.eptq.com
PTQQ1
Q12014
2015 131
27
PTQ
Q3
2015
87
PTQ
www.eptq.com
PTQ Q4 2014 129

10/03/2014 13:40
13/12/2013
10/06/2014 11:38
14:26

Evaluation case by case

of a risk assessment

Substantial
modification of
compressor unit.
Change of
function

Yes

Substantial
modification in
safety concept
Additional or new
protective device
New substantial
safety measures
Yes

Change of
application range
Change of
intended use
Change of
safety related
components

New hazard?
Yes
New safety concept
Is there an increase
required?
in risk?

Change of
performance

No

No
No substantial
modification of
compressor unit

Other
safety related
modifications

Figure 5 Evaluation of substantial modifications

pulsation study and mechanical

response study.

Pulsation study

A
pulsation
and
mechanical
response study was also carried out
for the original compressor design
in this revamp case. The modifications to compressor design and
thermodynamic operating conditions were communicated to the
supplier to reinvestigate and validate the system.
For the pulsation study, the cases
to be investigated are single and
parallel operation, the different
operating conditions at normal situation (100%) capacity and also for
the capacity control range between
100-15% by stepless suction valve
unloading. According to experience
with synchronised suction valve
unloading, the critical point for
pulsation volume design is at
approximately 70% capacity.
If the study requires it, the modifications to the system to satisfy
acceptable operation, by order of
preference, are:
Modification of restriction orifices
Modification of existing piping
supports
Modification of pulsation dampers.
In this case, the calculation
showed excessive pulsation levels

88 PTQ Q3 2015

as well as unbalanced forces on the

suction side first stage for the existing vessel volume. Therefore a new
acoustic filter had to be considered
with increased volume and dimensions. Also, a steel construction to
support the new vessel and new
orifices for the line side had to be
considered.

Machine directive/ATEX

According to machine directive

2006/42/EC and directive 94/9/EC
(ATEX) it is essential, in the event
of a substantial modification, to
provide a manufacturers declaration for the modified equipment.
The operating company has to
ensure
conformity
with
the
machine directive and ATEX for
the whole operation. This is also
mandatory for used machines
which are substantially modified in
a revamp. ATEX does not apply to
repairs without new features or
other modifications, or for spare
parts intended to replace defective
or worn out parts.
In this revamp case, there is a
capacity increase within the original operating conditions through
the installation of a new crankshaft
with an enlarged stroke. The crankcase and other internals of the
driving mechanism remain. Also,

the first and second stage cylinders

of the compressor remain. To
achieve the enlarged stroke, a new
piston and piston rods were
designed. Other compressor characteristics such as speed and cylinder
bore remain identical to the original
design.
To run the compressor under
high suction pressure and at low
suction temperature, the driver had
to be renewed.
The oil system, the control system
and other equipment was already
certified according to ATEX when
the compressor unit was installed
and did not need to be modified.
The distributed control system
(DCS) and the compressor instrumentation for safety control could
remain; only the set points in the
instrument list had to be adapted.
Does a substantial modification
need to be considered? The following questions have to be carefully
checked:
Are there changes in function or
operating
conditions
for
the
compressor unit?
Are all hazards covered by risk
analysis?
Or do new risks or new hazardous conditions exist after the
revamp?
What about the effects or rate of
incidents arising from existing
hazards?
Figure 5 shows an evaluation for
the original equipment manufacturer and operating company to
check whether a substantial modification is occurring.

Evaluation by risk analysis

NEA Group has generated a spark

hazard analysis and risk assessment for reciprocating compressor
units. If there is substantial modification in a revamp, the assessment
is performed and actions are indicated.
When
there
is
full
documentation
for
compressor
data, materials and design available, it is possible to prepare a
declaration
according
to
the
machine directive and ATEX. The
reciprocating compressors can be
confirmed for CE Ex II 2G T3.

Results of analysis

For this revamp there are:

www.eptq.com

 No new hazardous events

compared with the previous
situation
None of the existing hazardous
events increased; the new stroke
was recalculated and the loads
checked by the original equipment
manufacturer. This means that
existing safety protection would
have been sufcient even after the
revamp
Besides compressor layout, all
hazardous events were covered by
the existing risk analysis:
o incident rates and effects of
pressure and temperature shift
o valve failures
o pressure packings
o rod-reversals
o crosshead and bearings
o drive components
o oil pump/oil system
o compressor function and safety
system is identical as before the
revamp
o boundaries and interaction
with bordering systems
o compressor remains within the
same operational environment;
classication and conguration of
hazardous areas do not change.

Risk evaluation is mandatory for

each revamp project and must be
performed during the revamp
project phase. The results of the
evaluation must be documented,
for proof.
For revamps and modernisations,
special management of product
safety is necessary. The original
equipment manufacturer or an
expert authority must state that the
revamp measures are a safe solution. How the revamp is selected
and designed, as well as the associated risk analysis, must be
documented in detail.

Conclusion

Revamp and modernisation of

reciprocating compressors requires
special handling. Because a lot of
reciprocating compressors have
been running for decades, and may
even run for decades more, a
revamp is a good opportunity to
ensure that they t current technical specications and process
conditions. With the right technical
support by a compressor OEM, a
reciprocating compressor can be
prepared for long term operation,

meeting the demands of the operator company.

A revamp proposal will demonstrate a technically safe and
economically reasonable solution
using a major component of existing
resources.
Beside
the
compressor and the compressor
equipment itself, infrastructure
such as foundations, compressor
housing, cable routing, pipe racks
and so on may be further used.
A major revamp does not automatically
require
substantial
modication, but extensive modication can arise in smaller revamps
if there is a change in risk assessments. Only an OEM or an expert
authority with sophisticated engineering knowledge of reciprocating
compressors can provide the necessary manufacturers declaration or
documentation so that the operating
company
can
receive
permission to put the revamped
unit into operation.
Andreas Hahn is Head of Revamp and
Modernization, Neuman & Esser Deutschland
GmbH & Co KG. He holds a degree in mechanical
engineering from Aachen University.

Get Real

SOUR WATER STRIPPING IS CHALLENGING

Dont Guess!

Accurately simulate sour water strippers;

Predict ammonia distribution in amine systems
ProTreat 5.0 mass transfer rate-model calculates ammonia
absorption and stripping, with heat stable salts and caustic
injection, including pH prediction.

Turn your conceptual design into a virtual plant

with the industrys most advanced gas-treating
simulator.
ProTreat uses a real, mass transfer rate-based tower model.

Accurate, reliable simulation of refinery & syngas applications.

No ideal stages, absolutely no translation from theory

to reality.

Next-generation science for design, optimization,

troubleshooting of all plant configurations.

Completely predictive, with outstanding

reliability to significantly reduce risk.

ProTreat 5.0 is the only simulator that can accurately

predict sour water stripper performance and the
operation of amine units containing ammonia.

10 years of proven, highly successful

customer performance.

ProTreat Taking the Guesswork Out of Sour Water Stripping

Learn how state-of-the-art simulation science can benefit you.
www.ogtrt.com

+1 281 970 2700

OGT-0109 Ammonia HE Half Horizontal Ad.indd 1

www.eptq.com

info@ogtrt.com

11/13/11 5:38 AM

PTQ Q3 2015 89

bach-Palenberg

THINK GERMAN,
ACT LOCAL.
DO YOU WANT TO GET MORE
OUT OF YOUR RECIP?

BLUESTROKE
COMPRESSOR
SYSTEMS

REVAMP: THE BEST ECONOMICAL

CHOICE TO GET MORE.
Process adaptation or optimization? Bringing
your recip compressor back to state of the art?
Whichever parameter needs to be changed, trust
NEA when it comes to revamp or modernization
of your machine. We have references on hand in
all kind of applications with tailor-made solutions.
NEA can apply its revamp experience to its 11 legacy
and any other brands including risk assessment
acc. to CE and ATEX directives. Revamp with
as-new guarantee.
Authorized OEM supplier for
reciprocating compressor lines:

NEUMAN & ESSER

Deutschland GmbH & Co KG

www.neuman-esser.com
Contact me:
Andreas Hahn
Product Manager Revamp & Modernization
Andreas.hahn@neuman-esser.de
Direct Phone: +49 2451 481-182

NEA GROUP Headquarters in Germany

PTQ_NEA-Andreas Hahn-Compressor-2015-4_rev2.indd 1

18.05.15 11:40

Advanced turbomachinery controls

improve ethylene plant yield
Advanced integrated turbomachinery controls can significantly alleviate the suction
pressure dip of a cracked gas compressor, leading to considerable yield improvements
MEDHAT ZAGHLOUL
Compressor Controls Corporation

thylene plant yield is a function

of pyrolysis furnace back pressure, among other factors. The
mechanism to keep the furnace
back pressure at a certain value is
the suction pressure to speed the
cascade control loop on the cracked
gas compressor. A schematic
arrangement of the gas path from
the furnaces to the compressor is
shown in Figure 1.
In Figure 1, the design pressures,
temperatures and flow rates of the
gas path are depicted in a simplified manner for a 300 000 t/y
ethylene plant. Any particular plant
may show slight differences.
In every plant, the selected
set-point value for cracked gas
compressor suction pressure is
determined to ensure that, in the
event of single or multiple furnace
trip events, the suction pressure
does not dip low enough to subatmospheric values, thereby creating the risk of sucking in air with
the process gas, which could create
an explosive mixture.
The cracked gas compressor is a
multi-section centrifugal compressor, with four or five stages of
compression required to bring the
cracked gas pressure up to the
desired value of approximately
40 bara. The cracked gas compressor is usually driven by an
extraction type steam turbine using
high pressure live steam and
extracting steam to the plants
medium pressure steam network,
as well as condensing the remaining steam that is used in the turbine
low pressure compartment.
Caustic wash is usually installed
between the third and fourth (or
fourth and fifth) stages to remove

www.eptq.com

Cracked gas compressor

Stage
5

Stages
1, 2, 3, 4

Furnace No. 1
Quench tower
and steam
stripper

Acid gas
removal

Furnace No. 6
96 T/h
1.85 bara
850C

96 T/h
1.50 bara
310C

70 T/h
1.35 bara
45C

66.6 T/h
39.2 bara
120C

Figure 1 Schematic arrangement of pyrolysis furnaces and cracked gas compressor

any trace quantities of acid gas

(H2S) present in the cracked gas.
As the cracked gas that enters the
compressor is saturated with steam,
condensate removal is required after
each stages discharge cooler and
prior to introducing the gas to the
succeeding stage of compression.
Typically, the first group of

In the event of
a furnace trip, a
significant portion
of the cracked gas
available from the
furnaces drops almost
instantaneously
compressor stages (upstream of the
caustic wash) is protected against
surging by a single anti-surge
valve, while the stage (or stages)
downstream of the caustic wash
section is (are) protected by another
anti-surge valve.

A suction pressure controller uses

the cracked gas compressors first
stage suction drum pressure as its
process variable (PV), compares it
to the configured set-point (SP),
and the resulting error used to
generate a PID response to modulate the speed set-point of the steam
turbine driver, between its control
range of minimum governor to
maximum governor values.
The actual modulation of the live
steam valve (V1 valve) is done by a
speed governor, cascaded with the
cracked gas suction pressure
controller. The modulation of the
live steam valve, for those plants
that have an extraction turbine
driver, needs to be closely coordinated with the turbines extraction
(V2) valve.
In the event of a furnace trip, a
significant portion of the cracked
gas available from the furnaces
drops almost instantaneously. In
this example, we will consider
16.7% of the plants rated capacity
(assuming that normally there are
six furnaces in service).
The cracked gas pressure at the
compressors suction drum will

PTQ Q3 2015 91

145
140

Set point
Conventional control
Low limit

Furnace trips

135
130

Pressure dip of 25 kPa

125
120
115
110
105
100
5
0
0
50
10
150
0
20
250
0
30
350
0
40
450
0
50
550
0
60
650
0
70
750
0
80
850
0
90
950
10 0
1000
5
11 0
11 00
5
12 0
0
12 0
50

Cracked gas suction pressure,

kPa a

150

Time, seconds

Figure 2 Conventional controls dynamic behaviour for a furnace trip event

also start to drop steeply after the

time delay caused by the process
volume of the process equipment
between the furnace outlets and the
compressors suction drum.
Several factors will influence the
value of the minimum suction pressure value reached during the
furnace trip event. These include:
The volume (capacity) of the
process equipment between the
furnaces outlet and the cracked gas
compressor suction drum
The tuning of the compressor
suction pressure controller, and
The speed of response (quality)
of turbine speed control.
The volume of the process equipment in question is not a factor that
can be influenced by better control
strategies. However, the other two
factors are influenced by better
control strategies. Of those, the
quality of the turbine speed control
is more important, since it is the
inner loop of the pressure-tospeed control cascade arrangement.
The turbine speed control could
be based on a mechanical-hydraulic
governor, or a more modern digital
governor. In either case, the extraction control mechanism usually
presents control challenges to the
turbine speed control loop.
Take the example of an older
control system where the speed
governor is a mechanical-hydraulic
device using the flywheel principle
and is linked to the extraction valve
actuator with a mechanical linkage.
This type of older system would

92 PTQ Q3 2015

introduce significant oscillations in

the speed control whenever there is
a change in speed or extraction
demand which, in order to produce
a smoother response, would require
considerable dampening of the
control responses, rendering the
overall speed control response
quite sluggish.
While the most sophisticated
dedicated digital control hardware
and software algorithms available
in the market greatly improve the
speed control response, there is still
a minimum time constant achieva-

Advanced and
integrated
turbomachinery
controls can
significantly alleviate
the suction pressure dip
ble in terms of the speed control of
an extraction type turbine.
One of the most important rules
of controls implementation is that
the outer loop of a cascade control
arrangement (in this case the pressure controller) must be tuned at
least five times slower than the
time constant of the inner loop.
Thus, it is fair to conclude that even
with the best dedicated digital
speed control hardware, and the
best speed-extraction control algo-

rithms, there is still going to be a

somewhat sluggish to very sluggish
suction pressure control time
response in any ethylene plant.
It was perceived that, in order to
deal with the suction pressure
collapse that would accompany a
furnace trip (or multiple furnaces
tripping) and prevent the pressure
dip from becoming sub-atmospheric, a low suction pressure
limiting loop needs to be implemented, in tandem with the
normally sluggish suction pressure
control loop. This limiting loop
would be tuned to be faster than
the principle pressure control loop,
but would require a set-point value
that was offset from the main
suction pressure controllers.
For many years this was considered state-of-the-art for setting up
the cracked gas suction pressure
control set-point value. When a
time recording is made of the
behaviour of such a control
arrangement, the result could look
something like Figure 2.
With advanced control algorithms that incorporate the full
integration of all the compressors
turbo-machinery controls functions,
such as anti-surge, performance,
speed and extraction controls,
into one dedicated high-speed
controls platform, it is possible to
improve on the pressure dynamic
behaviour time trend shown in
Figure 2.
Modern surge control systems,
needed and implemented to protect
the charge gas compressor from
surging and its associated mechanical damage to the machine, require
that the installed anti-surge valves
have sufficient capacity to handle
full recycle operation of the
compressor train. In addition, these
anti-surge valve actuators are
required to have exemplary stroking characteristics, in terms of
stroking speed and freedom of
stroking from overshoot in both
directions of actuator travel. Thus it
is possible to utilise recycle gas
flow from the low pressure antisurge valve back into the cracked
gas compressor suction drum to
help recover the suction pressure
for a furnace trip event.
When the suction pressure drops

www.eptq.com

Discover
our comprehensive
portfolio

Turning sulfur into profit

Your reliable partner for efficient sulfur processes

Sulfuric acid production and sulfur recovery demand for

technologies that ensure highest availability and energy
sensitive operation. Reliable and efficient, Siemens
supplies the perfect solution for both with pre-designed
turbocompressors. Outstanding expertise and comprehensive R&D work at Siemens have led to technology that
guarantees mechanical durability and wear resistance,
especially under complex operation conditions.

The compressor series STC-SOF, STC-GO and STC-SOR

meet and even exceed these demanding requirements.
Pre-designed turbocompressors from Siemens have
proven to be more efficient than conventional solutions
in the long run. This is how Siemens is already meeting
the future demands of the sulfur industry.

siemens.com/singlestageturbocompressors

145

Furnace trips

140
135
130

Pressure dip of only 12.5 kPa

125
120
Set point
Conventional control
Low limit
CCC integrated control

115
110
105
100

5
0
0
50
10
150
0
20
250
0
30
350
0
40
450
0
50
550
0
60
650
0
70
750
0
80
850
0
90
950
10 0
1000
5
11 0
11 00
5
12 0
0
12 0
50

Cracked gas suction pressure,

kPa a

150

Time, seconds

150
145
140

Furnace trips

135

New set point

Conventional control
New low limit
CCC integrated control

130
125
120
115
110
105
100
5
0
0
50
10
150
0
20
250
0
30
350
0
40
450
0
50
550
0
60
650
0
70
750
0
80
850
0
90
950
10 0
1000
5
11 0
11 00
5
12 0
0
12 0
50

Cracked gas suction pressure,

kPa a

Figure 3 Conventional controls vs advanced integrated controls dynamic behaviour for a

furnace trip event

Time, seconds

Figure 4 The suction pressure set-point may be reduced with advanced integrated
controls without increasing the risk of a compressor trip during a furnace trip event

32

Ethylene yield, %

31
30
29
28
27
26
25

80
90
10
0
11
0
12
0
13
0
14
0
15
0
16
0
17
0
18
0
19
0
20
0
21
0
22
0

60
70

50

24

Furnace outlet pressure, kPa a

Figure 5 Furnace outlet pressure vs ethylene yield

Source: Petrochemical Processes by A Chauvel & G Lefebvre

94 PTQ Q3 2015

below the low suction pressure

limit, a fast limiting control
response can be used to open
temporarily the low pressure antisurge valve, in addition to lowering
the speed of the turbine.
It should be noted that, while the
low pressure anti-surge valve is
providing recycle flow back to the
cracked gas compressor suction
drum for the express purpose of
preventing the suction pressure
collapse in the event of a furnace
trip the high pressure stages
of the cracked gas compressor
could be starved of flow, sufficient
to require partial recycle as well.
It would thus be helpful if the
high pressure anti-surge controller
could open the high pressure antisurge valve in an anticipatory
manner, using loop decoupling
techniques between the two antisurge controllers of the cracked gas
compressor.
As may be seen in a comparison
of
conventional
controls
vs
advanced integrated controls in
dealing with a furnace trip event,
the pressure dip recorded was only
12.5 kPa (see Figure 3). There are
two possible control algorithm
strategies
to
implement
this
enhanced control function of using
the anti-surge valve to prevent
suction pressure collapse in the
charge gas compressor during a
furnace trip event:
Use a conventional limiting function entirely within the anti-surge
controller, that is when the first
stage suction pressure of the
cracked gas compressor drops
below a preconfigured fixed value,
then open the anti-surge valve
appropriately, or
Set up a pressure override function within the suction pressure
controller that transmits a control
response to open the LP anti-surge
valve of the cracked gas compressor when the first stage suction
pressure drops below a configurable offset of the suction pressure
set-point assigned value.
The latter approach may provide
an ever better overall recovery of
the suction pressure due to its
limiting threshold being an offset
to the suction pressure set-point,
rather than an absolute pressure

www.eptq.com

value. In either case, it should be

Changing perceptions
possible
to drop the assigned
The incremental
phosphoset-point
value of risk
the of
cracked
gas
rous fouling from
the inef
ciency
compressor
suction
pressure
of conventional,
thermally amount.
unstable
controller
by a significant
acidic
has beenina
For
thephosphate
purposes esters
of illustration
major
factor let
in us
limiting
the that
use of
this
article,
assume
it
chemical
corrosion
inhibitors.
may
be reduced
by approximately
Tanscient
new to
chemistry
that
12.5
kPa, isdown
122.5 kPa,
achievesof the
corrosion
instead
the required
135 kPa that
was
protectionwith
at treatthe
ratesconventional
of 0.1 to 0.5
possible
ppm phosphorus,
which 4).
is up to
control
system (see Figure
80%
phosphorus
than gas
in
This less
reduction
in the cracked
conventional
phosphate
esters.value
This
suction
pressure
set-point
high translate
phosphorus
ciency is
should
into aef
proportional
providing inrethe
nersfurnaces
with a outlet
safer
reduction
option for which
corrosion
and
pressure,
forprotection
our typical
is changing
reners perceptions
process
illustrated
in Figure of
1
HTCIs.be a drop from 185 kPa to
would
172.5 kPa. Based on the data
Conclusionin Figure 5, the correprovided
Integratingimprovement
opportunity in
crudes
into
sponding
ethylene
a recould
nerys becrude
slateas is
an
yield
as high
0.55%
accepted
strategy
to drive
(from
25.6%
to 26.15%).
Forincrethe
mental
improvement.
96
t/h ofmargin
feedstock
processed inHigh
the
naphthenic
furnaces
(see acid
Figurecrudes
1), this present
would
calcium removal
and corrosion
correspond
to an increase
in ethylprotection
processing
risks which
ene
production
of approximately
has t/h
limited
incremental
prot
0.53
of ethylene
or 4560 t/year,

with a market value of approxiimprovement

from these crudes.
mately
$5.7 million.
Dorf
Ketals be
Calcium
Removal
It should
pointed
out,
Additives that
and Tanscient
HTCI
are
however,
the slightly
lower
reducing
processset-point
risks andvalue
providsuction
pressure
for
ing cracked
reners with
safer, more reliable
the
gas compressor
would
and less2.86-3.00%
complex more
chemistry
to
require
power
increase high
naphthenic
acid
generated
by the
compressors
crudes turbine
in their crude
steam
driver,slate.
which does
impose a slight energy penalty (less
than $300 000-400 000/y) compared
TANSCIENT
is a trademark
Dorf Ketal. due
to
the potential
extraofrevenue,
to the improved ethylene yield.
Reference

1 Study on oil refining and oil market prepared

Conclusion

by Purvin and Gertz Inc. for the European

For
many years, the suction presCommission.
sure set-point of the cracked gas
compressor has represented a
constraint that plant operations
Parag Shah is Assistant Technical Director
consider
inviolate.
However
with Dorf Ketal, instrumental in software
advanced and integrated turbomadevelopment for desalter adequacy testing and
chinery
controls can significantly
monitoring fouling in preheat exchanger trains,
alleviate
the
suction
dip as
and Technical Manager
for pressure
Tansient technology.
aHeconsequence
of
single
or
multiple
holds a BEng in chemical engineering from
furnace
trips, thus allowing the
Mumbai University.
charge
gas compressor normal
Email: paragshah@dorfketal.com.
Mahesh Subramaniyam
is Director
of
suction
pressure setting
to be
Research & Development
Ketal.
significantly
lower with
thanDorf with
At the forefront control
of all Dorfsystems.
Ketals chemical
conventional
This

in turn can lead to considerable

developments,
he holds a number
yield
improvements
in theof patents
plant
for oil treatment
and revenue
engineers
concerned,
with chemicals
a potential
products to address emerging issues in the
increase
in the many millions of
hydrocarbon industry. He holds a PhD in
dollars.

chemistry from Indian Institute of Technology,

Bombay. Email: drmaheshs@dorfketal.com
James Ondyak is Executive Vice President,
Further reading
Marketing & Business Development with Dorf
1 Narayanan K, Ramakrishnan S, Maximizing
Ketal. He is responsible for global marketing
ethylene plant production with CCCs
of Dorf Ketal technologies and holds a MBA
Advanced Turbomachinery Constraint Control
from Harvard Business School and a MS in
System, Compressor Controls Corporation,
chemical engineering from Illinois Institute of
2005.
Technology.
2 Chauvel A, Lefebvre G, Petrochemical
Email: james.ondyak@dorfketal.com
Processes.
James Noland is Senior Director of the Process
Chemicals Division with Dorf Ketal USA, LLC.
He has been involved in the refining and
petrochemical industry for over 17 year and
Medhat Zaghloul is Regional Technology
now leads the process chemical division for
Manager for the Europe, Middle East and Africa
North America. He holds a BEng (Chemical)
Regions with Compressor Controls Corporation
from Mississippi State University.
(CCC), based in the Abu Dhabi Office. He has
Email:
over
38james.noland@dorfketal.com
years of experience in the oil and gas
India
Nagi-Hanspal
is Lead
nery Engineer:
industry
with 15 years
in Refi
instrumentation
Technical
Services
Dorf
Ketal.
She
primarily
and control for the petrochemicalis industry,
responsible
for
providing
refi
nery
process
and is responsible for the development
of
improvement
strategies
through
unit
technical
solutions
and control
applications
monitoring
provides
chemical
and
providingand
technical
guidance
and solutions
support.
to
Dorf
Ketal
site
teams
globally.
She from
holds
He holds a BSc in electrical engineering
a
MEng
(Chemical)
from
Imperial
College
the Cairo Institute of Technology (now Helwan
London. Email:
indianagi@dorfketal.com
University)
in Egypt.

THE RVP PROFIT ANALYZER

MINIVAP ON-LINE

Vapor Pressure of Gasoline, Crude Oil and LPG

RVPE, DVPE, TVP, V/L Analysis
According to Latest ASTM & ISO Standards
Unparalleled Measuring Accuracy
Up to 2 Sample Streams
Adjustable Temperature: 20-60C
Automatic Calibration
Fast and Easy Maintenance

Grabner Instruments Messtechnik GmbH | A-1220 Vienna, Austria | Dr. Otto-Neurath-Gasse 1

Phone +43-1-282 16 27-0 | Fax +43-1-280 73 34 | info.grabner-instruments@ametek.at
ON-LINE half page.indd 1

06.12.2013 10:17:28

www.eptq.com

PTQ Q3 2015 95

www.eptq.com

PTQ Q2 2015 87

Taking the Next Step

The new 520mm oil-injected screw compressor

Marine Engines & Systems Power Plants Turbomachinery After Sales

Continuing the on-going success story of our screw compressor technology, we are proudly
introducing the all-new CPO400 with a rotor diameter of 520mm. Based on more than 60 years
of screw compressor experience, we are ready to take the next step and boost your compression business. With a maximum discharge pressure of 40 bar (580 psi) and volume ows of up
to 17,000 m/h (10,000 SCFM) we are now able to handle your oil-injected screw compressor
application with our extended CPO series. Find out more at www.mandieselturbo.com

Typ 02 - Taking the Next Step - 210x297.indd 1

2015-05-11 10:35:05

Mixing in large crude tanks

Computational fluid dynamics was used to analyse the performance of side entry
mixers in large crude tanks
JANAKIRAMULU ADEPU, RAHUL C PATIL, AJAY GUPTA and ASIT KUMAR DAS Reliance Technology Group
PRAVIN POTDAR, JISNA RAGHAVAN and KANNAN SRINIVASAN Central Technical Services, Reliance Industries Limited

n a refinery, crude oil and product storage tanks are very

important factors in smooth and
continuous production. Crudes are
generally received in large storage
tanks from ships via pipelines for
further processing. The storage
tanks are often equipped with side
entry mixers, primarily to avoid
settling/layering of sludge inside
the tank. These mixers are operated
to obtain a homogeneous blend of
different crudes with respect to
basic sediment and water (BS&W)
and density of the crudes. A homogeneous crude blend is desirable
for consistent operation of downstream units. The crude blends
homogeneity is determined based
on the density variation between
three samples of crude taken from
the top, middle and bottom of the
tank and a BS&W value which is
specific to an individual refinery.
The mixer can be installed in the
storage tank in different configurations. The configurations can be
differentiated based on number of
mixers, which can vary from one to
seven, and the swivel angle of the
mixer which can vary about the
mixers axis of rotation. The swivel
angles of mixers in tanks are
changed based on the desired
results. In this study, two different
configurations of mixer were simulated and the time for achieving
homogeneity was computed using
commercial computational fluid
dynamic (CFD) software.

The CFD approach

The primary objective of the study

was to understand the effect of the
number of mixers in a crude storage tank on the time required to

www.eptq.com

Figure 1 Crude storage tank geometry

achieve homogeneity with respect

to the density of the crude. CFD
studies were carried out for two
layouts of the tanks:
Case 1 Tank with three mixers
Case 2 Tank with six mixers.
Initially, steady state simulations
were carried out to visualise the

During the steady

state simulations, the
entire tank was filled
with heavy crude
and the mixers were
rotated
flow profiles inside the tank. The
velocity profiles inside the storage
tank
were
established
and
compared for different cases in this
study. Later, unsteady state simulations were carried out to study the
effect of the number of mixers on
the time required for complete
homogenisation. The criterion for

complete
homogenisation
was
considered as a density difference
of less than 1 kg/m3 at different
cross-sections of the tank.
The tank geometries for the two
cases with mixers are shown in
Figure 1. The mixer locations with
respect to 0 location are also
shown in Figure 1. The mixers are
oriented towards one side by a
fixed angle around their axis of
rotation in both cases. If the diameter of the tank is of the order of
100m, then the tank is filled to a
height of D/7.5. Analysis of flow in
the domain was carried out using
commercial CFD software. During
the steady state simulations, the
entire tank was filled with heavy
crude and the mixers were rotated.
The mixers were modelled in CFD
using a multiple reference frame
(MRF) model. The simulations were
continued until the residues fell
below 10-5, indicating convergence
of the flow problem.
For the unsteady state simulations, equal volumes of heavy and
light crudes were considered to be
in the tank. Heavy crude (density

PTQ Q3 2015 97

the homogeneity criterion of a

density difference at different
cross-sections of less than 1 kg/m3
was met.

Results
Steady state simulations

Figure 2 Density contours at the centre (x = 0) of the tank; time t=0 hrs
Comparison of maximum and average velocity at different cross-sections
m/s
Mixers
Plane
1
2
3
4

Max velocity

Average velocity

VM
0.85 VM
0.68 VM
0.57 VM

1.14 VM
0.91 VM
0.87 VM
0.71 VM

VA
0.94 VA
0.92 VA
0.89 VA

1.89 VA
1.85 VA
1.80 VA
1.64 VA

VM = Maximum velocity, VA = Average velocity

Table 1

= 926 kg/m3) filled the bottom half

of the tank and the remaining top
portion of the tank was filled with
light crude (density = 859.5 kg/
m3). The initial condition of the
unsteady state simulation is
shown in Figure 2, the red colour
indicating
heavy
crude
and
the blue colour indicating light
crude.
With this initial condition, the
mixers were rotated at fixed speed
and variations of density at different cross-sections of the tank were
monitored to estimate the time
required for complete homogenisation. An average density of 893 kg/

For better
understanding, the
velocity profiles
were examined at six
cross-sections along
the vertical axis of
the tank
m3 is expected after complete
homogenisation in the tank. The
simulations were terminated when

Figure 3 Velocity magnitude vectors at the level of the impeller shaft

98 PTQ Q3 2015

The velocity distribution inside the

crude tank for the two cases was
examined. For better understanding, the velocity profiles were
examined at four different crosssections along the vertical axis of
the tank.
The velocity profiles were examined at planes 1 to 4 along the
vertical axis of the tank, starting
from bottom to top. The maximum
and average velocities for the two
cases are compared in Table 1. It
can be seen from Table 1 that, as
expected, the maximum and average velocities at all cross-sections
increases as the number of mixers
increases. It can be also seen that
as the height increases the maximum
and
average
velocity
decreases,
showing
that
the
effect of the jet developed due to
mixer
action
reduces
along
the vertical axis of the tank.
The velocity vectors at the impeller level for the different cases are
shown in Figure 3. The velocity is
scaled in m/s with the red colour
showing highest velocity magnitude and the blue colour showing
lowest velocity magnitude. It is
apparent from Figure 3 that the jets
of the impeller intersect with each
other, forming a big jet swiping the
periphery of the tank. The velocity
of this resultant jet increases as the
number of mixers increases. The
path lines of velocity for the two
cases are also shown in Figure 4 for
reference. It is again clear from
Figure 4 that, with an increase in
the number of mixers, the swiping
action at the periphery of the tank
increases.
Although it is not very apparent
from the steady state simulations
which layout, three mixers or six
mixers, is better than the other, it
can be seen that the number of
mixers has a marked effect on the
velocity profiles in the tank. An
increase in the number of mixers
increases the amount of swiping at
the periphery of the tank

www.eptq.com

Process Focus:

Gas Processing

With broad capabilities, versatility, and flexibility in a single powerful process simulation tool,
ProMax is used worldwide to design and optimize gas processing, refining, and chemical facilities.
Fully integrated with Microsoft Visio, Excel, and Word, ProMax is the premier simulation
resource for designing gas plants and predicting performance under varying process conditions.

Discover why engineers around the world depend on ProMax!

Model any NGL, LNG, LPG recovery and fractionation process
Rate heat exchangers including brazed aluminum
Calculate air emissions for condensate, oil and produced water tanks, well sites, flares and more
Determine hydrate formation and CO2 freeze-out temperatures using best-in-industry predictions
Use Excel to perform parametric studies and economic analyses
Calculate hydrate inhibitor requirements using industry-best MeOH/H2O/HC predictions
Model amine treating, dehydration, cryogenic, and fractionation facilities in the same project

Process Solutions

Engineering Support

Worldwide Training

ProMax

Process Simulation Software

Engineering Solutions for the Oil and Gas, Refining, and Chemical Industries
sales@bre.com
www.bre.com
979-776-5220 US 800-776-5220

Case 1

Case 2

Figure 4 Path lines of velocity magnitude

Average density, kg/m3

930
920
910
900
890
880

Plane 1
Plane 2
Plane 3
Plane 4

870
860
850

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Time, h

achieve homogenisation was estimated for both layouts. The

criterion for complete homogenisation was considered as a density
difference of less than 1 kg/m3 at
different cross-sections of the tank
and simulations were stopped
when this criterion was met. The
densities of the mixture in the tank
were compared at different planes,
plane 1 to 4 from the bottom to the
top of the tank.
The variation of densities at
different planes of the tank with
time for Case 1 is represented
graphically in Figure 5. The time
estimated for complete homogenisation for Case 1 is 18 hours.
The variation of densities for
Case 2 at different planes of the
tank with time is represented
graphically in Figure 6.
For Case 2, the time estimated for
complete homogenisation is less,
compared to Case 1. Around 11
hours are required to achieve
complete homogenisation for Case
2; that is, seven hours less than for
Case 1.

Conclusions

The effect of the number of mixers

on the time required for homogenisation was studied using CFD. Two
layouts were studied in steady as
well as unsteady state simulations:

Figure 5 Variation of density with time for Case 1

Average density, kg/m3

930
920

Case 1 Tank with three mixers

Case 2 Tank with six mixers.

910
900
890
880

Plane 1
Plane 2
Plane 3
Plane 4

870
860
850

10

11

12

Time, h

Figure 6 Variation of density with time for Case 2

and helps to achieve more velocities at each cross-section of the

tank along its vertical axis. For
more conclusive results, the
unsteady state simulations were
carried out for two layouts. The
time required for achieving homogeneity was estimated from these
simulations.

100 PTQ Q3 2015

Unsteady state simulation

As was explained earlier, the

unsteady state simulations were
carried out in a tank initially filled
with two crudes of equal volumes,
the bottom layer heavy crude and
the top layer light crude. The
mixers were then rotated at fixed
speed and the time required to

Although it is not very apparent

from the steady state simulations
which layout, three mixers or six
mixers, is better than the other, it
was established that the number of
mixers makes a marked difference
to the velocity profiles in the tank.
Increasing the number of mixers
increases the swiping effect at the
periphery of the tank and helps to
achieve more velocities at each
cross-section of the tank along its
vertical axis. For more conclusive
results, unsteady state simulations
were carried out.
In the unsteady state simulations,
two layers of equal volumes of two
different crudes (heavy at the
bottom and light at the top) were
considered as the initial condition,
and the time required to achieve

www.eptq.com

homogeneity was estimated. The density variation

along with time was monitored and the time required
to achieve density differences of less than 1 kg/m3 at
different cross-sections of the tank along its vertical
axis was calculated. The time required for homogenisation for a layout of three mixers was found to be
around 18 hours and that for a layout of six mixers
the time required for homogenisation was found to
be around 11 hours.

The time required for

homogenisation for a layout of
three mixers was found to be
around 18 hours
Further reading
1 Dakhel A A, Rahimi M, CFD simulation of homogenization in
large-scale crude oil storage tanks, Journal of Petroleum Science and
Engineering, 43, 151-161, 2004.
3 Rahimi M, The effect of impellers layout on mixing time in a
large-scale crude oil storage tank, Journal of Petroleum Science and
Engineering, 46, 161-170, 2005.

Janakiramulu Adepu is a Manager in the Refining R&D division

of Reliance Industries (RIL). He has eight years experience in
mathematical modelling, computational fluid dynamics and
troubleshooting refinery units, and holds degrees in chemical
engineering (BTech and MTech) from NIT Warangal and IIT, Kanpur
respectively.
Rahul C Patil is a Senior Manager in the Refining R&D division of
RIL. He has about eight years experience in process modelling,
simulation and computational fluid dynamics in refining units, and
holds chemical engineering degrees (BTech and MTech) from UICT,
Mumbai and IIT Kanpur respectively.
Ajay Gupta is Assistant Vice President in the Refining R&D division
of RIL. He has 25 years experience in process modelling, simulation
and control, and computational fluid dynamics, and holds chemical
engineering degrees, (BTech, MTech and PhD) from IIT, Delhi.
Asit Kumar Das heads the Refining R&D division at RIL, Jamnagar. He
has 30 years experience in refining research with several publications
and patents to his credit. He holds degrees in chemical engineering
(BTech, MTech and PhD) from Jadhavpur University, IIT, Kanpur, and
University of Gent, Belgium, respectively.
Pravin Potdar is a Senior Manager in the Central Technical Services
department of Reliance refinery, Jamnagar. His experience includes
technical support, energy/blending optimisation and troubleshooting
in refinery/marine tank farm environments. He holds a bachelors
degree in chemical engineering from Shivaji University.
Jisna Raghavan is a Manager in the Central Technical Services
department of Reliance refinery Jamnagar. She has about eight
years experience in process modelling, process optimisation, product
blending optimisation, energy optimisation and troubleshooting
in various refinery process units, and holds a BTech in chemical
engineering from Calicut University.
Kannan Srinivasan is a Senior Vice President in the Central Technical
Services department of Reliance refinery, Jamnagar. He has 30 years
experience in the chemical industry and holds chemical engineering
degrees (BTech and MTech) from A.C. Tech, Chennai and IIT, Chennai
respectively.

DAMPF TRIX
The Best for the
transport of steam
Steam hoses from ContiTech have been designed to meet
the highest requirements on hose material for the safe
transport of steam and hot water in the chemical and
petrochemical industry, in refineries, the building industry,
in shipyards. Three types were designed according to DIN
EN ISO 6134 for specific applications:
q DAMPF TRIX 5000 according to DIN EN ISO 6134-1A
q DAMPF TRIX 6000 according to DIN EN ISO 6134-2A
q DAMPF TRIX 6000 OIL according to DIN EN ISO 6134-2B

Fluid Technology
info@fluid.contitech.de/ih

www.contitech.de

www.eptq.com

PTQ Q3 2015 101

ContiTechAnzeige DampfTrix+OL 270x86 GB 10-2014.indd 1

07.05.15 09:10

Improved feedstocks for ethane crackers

Cryogenic recovery technology enables wet shale gas to become a valuable
source of liquids for petrochemicals production
RON KEY
Linde Process Plants

he petrochemical industry is
booming in the US, largely
fuelled by the low cost of
shale gas and shale gas liquids as
feedstocks. New drilling and
exploration
techniques
enable
natural gas and liquids to be
extracted cost-effectively from
shale rock and, when employing
state-of-the-art cryogenic gases
technology, even wet shale gas
basins can now be developed efficiently to deliver valuable natural
gas liquids for various industries.
While natural gas reserves occur
in many other parts of the world,
it is in the US where industry is
most determined to develop
unconventional sources of natural
gas in order to reduce dependence
on energy imports. But shale gas is
a lot more than just a source of
energy. Not only is it an ideal fuel
for power plants and district heating, it also delivers valuable
feedstock such as ethane for the
petrochemical industry. Natural
gas can be easily converted into
base chemicals such as ethylene,
which is a hydrocarbon compound
and precursor of many common
chemicals. The best known are
probably the plastics found in
everything from packaging and
cable insulation through car seats
to toys, according to Thomas
Rings, A.T. Kearney partner and
expert for energy and utilities. The
American
Chemistry
Council
(ACC) estimates that, by 2020, the
shale gas boom will fuel investments to the tune of almost $70
billion in the American petrochemical industry alone. Yet gas fields
that will be of real interest to
ethylene producers are still at the

www.eptq.com

very early stages of development.

Ethylene produced from shale gas
has the potential to unleash a new
industrial revolution in the US.
Before that can happen, however,
there are hurdles to overcome.
Processes for the recovery of shale
gas liquids typically used thus far
are not the most efficient. Efficient
processing of these valuable raw
materials calls for highly specialised know-how as the quality of
shale gas reserves can vary significantly from one shale basin to
another.

IHS Chemical
predicts that global
demand for ethylene
will rise to around
160 million tonnes a
year by 2017
Wet shale gas

Until now, extraction efforts have

mainly focused on dry gas, which
is an excellent source of energy
containing almost pure methane
with a small share of longer chain
hydrocarbons. Wet shale gas is a
much more interesting prospect
for industry, as these reserves
contain extremely valuable raw
materials. Their compositions can
vary greatly. Wet shale gas
contains less methane, but higher
concentrations of ethane (C2) and
longer-chain hydrocarbons such as
propane (C3) and butane (C4) and a
variety of higher molecular weight
hydrocarbons. Referred to as natu-

ral gas liquids (NGL), they are the

perfect feed for gas crackers,
which produce ethylene. This
translates into a huge market
opportunity. Chemical insight and
forecasting company IHS Chemical
predicts that global demand for
ethylene will rise to around 160
million tonnes a year by 2017.
Unconventional reserves present
a number of challenges to chemical
engineers, however. Wet shale gas
is occasionally contaminated with
impurities such as trace amounts
of mercury and sometimes hydrogen sulphide (H2S) as well as
carbon dioxide (CO2). These
substances
must
be
almost
completely removed before the
ethane and propane in the shale
gas can be fed to an olefin plant.
Cryogenic condensation technology that was originally designed
for recovery of ethylene, propylene
and heavier hydrocarbons for
refinery and petrochemical plants
has evolved for the recovery of
ethylene,
ethane,
propylene,
propane as well as other C2s and
C3+ hydrocarbons from natural
gas. Plants that
previously
compressed
and
cryogenically
cooled conventional raw natural
gas only, in order to crack it into
its constituent parts, can now also
easily handle wet shale gas.
With so many refineries either
being fully integrated or located
next to petrochemical plants, existing cryogenic C2+ and C3+
recovery technology had to be
adapted to meet evolving market
demands in other words,
increase plant flexibility and
recover more liquid hydrocarbons.
This involved quite a few technical

PTQ Q3 2015 103

95
90
85
80
75
70
65
60
55
50

980

780

Inlet pressure,
psia

580 10.905

5.076

90-95
85-90
80-85
75-80
70-75
65-70
60-65
55-60
50-55
2.042

Inlet gas,
gpm

Cryo-Plus

C2 recovery, %

C2 recovery, %

Typical NGL plant

95
90
85
80
75
70
65
60
55
50

980

780

Inlet pressure,
psia

580 10.905

5.076

2.042

Inlet gas,
gpm

Figure 1 Cryo-Plus provides higher recovery over a wider range of pressures and compositions

refinements. The higher concentration of hydrocarbons with a high

molecular weight in wet shale gas
made it also necessary to modify
the feed treatment process.

The process

Both the original technology and

its enhanced sister system are
cryogenic recovery technologies
utilising a turbo-expander to
recover energy while cooling the
feed gas. These cryogenic technologies are unique in their ability to
process low-pressure hydrogen
bearing refinery fuel gas streams
and obtain high recoveries with
less compressor and/or refrigeration horsepower than conventional
or competing cryogenic processes.
To protect the unit against upset
conditions, feeds may first pass
through a coalescing filter/separator designed to remove solid
particles and liquid droplets that
may carry over from upstream
processes. Although the cryogenic
systems can tolerate small quantities of H2S and CO2 these
compounds are not desirable. The
use of an amine treating unit for
removal of acid gas components
removes these compounds in an
absorption process as a feed conditioning step. The next phase is to
compress the feed streams unless
they are already at elevated pressures. Air coolers or cooling water
reduces the temperature of the gas
downstream of the compressor to
remove the heat of compression.
To avoid ice and hydrate formation in the cryogenic section of the
process, the water content of the

104 PTQ Q3 2015

gas is reduced to an acceptable

level through adsorption in molecular sieve desiccant beds. The
regeneration of the beds is a batch
process, where two or even multiple adsorption beds are used. One
or more of the adsorption beds is
being regenerated to restore their
capacity while the others are
on-line and drying the feed gas. A
recycle portion of the dry gas can
be heated and used for regeneration of the beds to drive off the
adsorbed water. Cooling of this
stream condenses the removed

Although the
cryogenic systems
can tolerate small
quantities of H2S and
CO2 these compounds
are not desirable
water before it recycles and
combines with the feed gas. A
portion of the residue gas may
also be used for regeneration on a
once-through basis. Downstream
of the adsorption beds, the gas
passes through a dust filter to
remove any particulate carry-over
before subsequent processing.
After dehydration, the feed gas
flows into the cold section of the
process,
where
cooling
by
exchange of heat with the residue
gas and cold separator liquids
takes place using a brazed aluminium plate fin heat exchanger.

Although it is not always a

requirement, the gas may be
further cooled using external
refrigeration before it goes to the
cryogenic portion of the process.
Following cooling, the feed gas
is partially condensed and delivered to a vapour/liquid separator.
The liquid then flows through the
inlet exchanger to cool the feed
gas before entering the deethaniser or demethaniser for C2
recovery for fractionation. The
vapour flows to the inlet of the
expander/compressor. As the gas
expands, it provides the work/
energy for the compression. The
expansion and removal of energy
cools the gas further and causes
additional
condensation.
The
expander discharges into the first
tower of a two-stage fractionation
process. The configuration and the
combination of fractionation and
heat transfer between these two
columns
is
the
proprietary,
patented technology that gives
this cryogenic technology its
advantages namely higher
recovery at reduced horsepower
over competing technologies (see
Figure 1).
A residue gas and a deethanised
or demethanised liquid product
are produced from this two tower
scheme. The residue gas is at a
slightly
reduced
pressure.
Following exchange with the feed
gas in the inlet cooling step, it
arrives at the pipeline or pipeline
compressor system as a dry, stable
heating value fuel. The liquid
product from the fractionation
system is the recovered C2+ or C3+
liquid hydrocarbons. The liquid
often
undergoes
additional
processing, such as additional fractionation in downstream columns.
For C3+ recovery, the liquid stream
is normally debutanised. The C3s
and C4s may then be fed to a
cracker process or split, with the
C3s going to LPG for fuel and only
the C4s going to isomerisation or
other purpose. For C2+ recovery, a
deethaniser normally precedes the
debutaniser. The overhead from
the deethaniser, ethane, becomes
an excellent feedstock for the
cracker (see Figure 2).
Also, the hydraulics of the facil-

www.eptq.com

ity required a thorough review

and were adapted to better
support turndown cases of the
design. Competing recovery methods can provide high levels of
combined ethane/propane recovery. However, these processes fall
short if demand changes, for
instance, to a product with high
levels of propane recovery with
limited or no demand for ethane
recovery.
When refining the cryogenic
system design to cope with the
additional wet shale gas components, Linde engineers were able
to build on the experience they
had already gained with their
Cryo-Plus technology in the field.
The original technology has
already been successfully deployed
in over 26 refineries and petrochemical plants, where it is used to
recover ethane, propane and
hydrocarbons with a higher molecular weight from refinery and
petrochemical
off-gases.
And
demand for this technology is
rising around the world, driven in
particular by growing concerns
over climate protection.
In Iraq, for instance, almost 60%
of off-gases from oil production
much of which is ethane-rich is
flared. The region around Basra
alone flares enough ethane to
produce as much as two million
tonnes of ethylene every year. That
ethane could be easily captured
with Cryo-Plus.
Successful as a result of its
production synergies, the US natural gas boom is providing strong
momentum for the petrochemical
industry. Many North American
refineries are already extracting so
much feedstock from shale gas
that US oil companies are thinking
about exporting crude oil for the
first time since the oil crises at the
start of the 1970s. Integrated crude
oil refineries and petrochemical
plants have the strongest prospects
of increased profits due to production synergies, coupled with
integrated, more cost-effective utility systems.
In addition, integrated designs
have a greater ability to control the
risk of fluctuations in demand for
either fuel or petrochemical

www.eptq.com

production. The naphtha cracker

run by BASF-Total Petrochemicals
in Port Arthur, Texas, is a good
example. Linde engineers will
soon be finished upgrading the
facility so that it can process more
ethane. Since April 2013, the
cracker has been meeting 40% of
its ethylene needs from ethane and
another 40% from butane and
propane. Supported by Linde,
BASF-Total
Petrochemicals
announced the startup of operations of a further ethane furnace in
March 2014. The petrochemicals
company says that the revamped
steam cracker has improved its
feedstock flexibility, enhanced
plant reliability, increased the
sites annual production capacity
of ethylene to more than 1 million
tonnes and investment in the
steam cracker allows the plant to
produce more than 85% of its

In just a few years,

shale gas has turned
the US from one of
the worlds most
expensive chemical
hubs into one of the
worlds cheapest
ethylene based on cost advantaged feedstocks. According to
Patrick Pouyann, President of
Total Refining and Chemicals, his
company is also looking at a
project to build a new ethane
cracking furnace, which would be
attached to the original Port
Arthur cracker. So it seems that
many operators in the US are
already in the process of upgrading their cracker systems.

Shale gas production still to rise

In just a few years, shale gas has

turned the US from one of the
worlds most expensive chemical
hubs into one of the worlds cheapest.
In
its
2014
report,
the
US
Energy
Information
Administration (EIA) projects that
shale gas production will continue

Feed conditioning

Particulate removal
Liquids separation
Acid gas removal
Other treatment

Compression
and dehydration

Cooling
Heat exchange
Refrigeration

Recovery
Turboexpansion
Fractionation

Additional
fractionation

Deethanising
Depropanising
Debutanising
Deisobutanising
C2
C3
C4s
C5s

Upgrading
processes

Chemical feedstock
Isomerisation
Polymerisation
Alkylation
Gasoline blending
Dry conditioned
fuel gas

Figure 2 Cryo-Plus process flow

to rise in the country until 2040.

However, Linde sees applications
in many global locations and
particularly in areas where natural
gas liquid-rich shale is available.
Ron Key is Vice President of Technology
and Innovation with Linde Process Plants,
Inc., and is an experienced business leader
in engineering, procurement, fabrication,
and construction. He holds multiple process
related patents and graduated from the
University of Tulsa, Oklahoma, with Bachelor
and Master of Science degrees in chemical
engineering.
Email: Ron.Key@LPPUSA.com

PTQ Q3 2015 105

Leading-edge technologies for solid feedstocks

With a growing shortage of resources, sustainable solutions for
energy conversion are becoming increasingly important. Whether
coal, petroleum coke, biomass or municipal solid waste: Thanks
to our advanced and robust Uhde Gasication Technologies,
PRENFLO and HTW, we are in a position to offer the right
process for every feedstock and end-to-end solutions for every
application from a single source.

Over

70

years of experience
in solids gasication
Over 100 gasiers worldwide

ThyssenKrupp
Industrial Solutions
www.thyssenkrupp-industrial-solutions.com

Controlling the dew point of natural gas

Meeting water and hydrocarbon dew point specifications for natural gas without
acid gas components
TEK SUTIKNO
Fluor Enterprises

atural
gas
transported
through pipelines to users
will generally need to meet
both water and hydrocarbon dew
point specifications. Several proven
processing options are well known
for meeting these specifications, but
the cost effectiveness and suitability
of these available options depend
on a particular application including the required dew point targets.
To evaluate these options for a
particular project, the design team
typically needs to perform process
simulation with proper thermodynamics packages, size the items of
equipment required, and estimate
the capital and operating costs for
each of the evaluated options.
However, there are application
areas where one of these options
generally becomes favourable or
the most cost effective. This article
highlights these application areas
and
summarises
options
for
controlling both water and hydrocarbon dew points of natural gas.
Gas dehydration is necessary as
hydrate or ice may form in nondehydrated gas at low temperatures; this will be encountered in
cold climates or when gas reduces
in pressure through the delivery
system to the users. Condensation
of water in the natural gas stream
also leads to corrosion problems,
especially from acid gas components such as CO2 and H2S. In
addition to corrosion, the acid gas
content affects the equilibrium
water saturation of natural gas and
the design and selection of the
required
dehydration
process.
However, this article excludes
topics on acid gas removal and
focuses on controlling both the

www.eptq.com

Dehydrated
gas

TEG
contactor

Flash gas

Water
vapour

Regeneration
still

Lean TEG

Flash
drum

Reboiler
Surge drum

Rich TEG

Feed gas

Filter

Figure 1 TEG contactor and TEG regeneration system

water dew point (WDP) and hydrocarbon dew point (HDP) of natural
gas with no or insignificant
contents of acid gas.
To avoid condensation of heavy
components when the gas pressure
reduces in the delivery system, the
HDP of natural gas also needs to be
reduced as necessary. Condensation
of heavy components increases
pipeline pressure drop and necessitates higher frequencies of pigging
operation to maintain the targeted
deliver capacity.
Controlling or reducing both the
water and hydrocarbon dew points
can be accomplished via a number
of processing options. References 1
and 2 are examples of numerous
publications covering WDP and
HDP control units. Most of the

HDP control options involve lowering gas temperatures to condense

heavy components, and the feed
gas is typically dehydrated first to
prevent gas from freezing and
equipment from plugging. This
article discusses typical dehydration schemes first, followed by
comparative evaluation of the HDP
options where minimisation of both
capital and operating costs typically
depends on the feed gas composition, available feed supply pressure,
and the targeted delivery pressure,
in addition to the gas specification
required by the user.

Dehydration

Natural gas from reservoirs typically contains water, and treated

natural gas from an acid gas

PTQ Q3 2015 107

Feed gas typically discharged

from a compressor or separator

Sales
gas

Sales gas
compressor
Gas-gas exchanger
Refrigerant
chiller

Liquid
hydrocarbon

Refrigeration
system

Inhibitor
recovery
system

Liquid
hydrocarbon

Inhibitor injection

Figure 2 Inhibitor injection schematic for WDP control and refrigeration for HDP control

removal unit is typically water-saturated. Water-saturated gas is

commonly dehydrated through
adsorption
or
an
absorption
process. Adsorption utilises solid
desiccants such as silica gels,
alumina, and zeolite molecular
sieves. These desiccants vary in
adsorption isotherms and dehydration performance characteristics.
Among these sorbents, the molecular sieve is the only option which
can dehydrate the feed gas to less
than 0.1 ppmv of water or with a
WDP as low as -150F (-100C) or
less.
A
molecular
sieve
is
commonly used upstream of cryogenic gas processing such as
liquefied natural gas (LNG) plants
and demethaniser units for ethane
recoveries. In addition to water,
other polar compounds such as
acid gases like CO2 and H2S are
also adsorbed in certain types of
molecular sieves.
Water content specifications for
pipeline-transported natural gas
typically range from 1-7 lb/
MMSCF (20 ppmv to 140 ppmv),
and glycol absorption is commonly
utilised for gas dehydration with
this specification target range. A
glycol dehydration unit (see Figure
1) consists of a contactor absorbing
water from the feed stream, and a
glycol regeneration system strip-

108 PTQ Q3 2015

ping off absorbed water in the rich

glycol and generating lean glycol.
Glycol absorption via a contactor to
meet the 1-7 lb/MMSCF target
range requires less capital and
fewer operating costs relative to
those of the molecular sieve option.
When the WDP target is very low
and requires molecular sieve or

The molecular sieve

is the only option
which can dehydrate
the feed gas to less
than 0.1 ppmv of
water or with a WDP
as low as -150F
another solid sorbent, a combination of glycol absorption and
molecular sieve adsorption possibly
improves the economics of the
dehydration system.
Commonly used glycols are
ethylene glycol (EG) and triethylene
glycol
(TEG).
A
TEG
dehydration unit with a conventional regeneration scheme can
reach about 99 wt% lean TEG. The
weight per cent of the regenerated

TEG increases at lower regenerator

pressures and higher reboiler
temperatures.
At
temperatures
higher than 404F (207C) however,
TEG starts to degrade. The glycol
regeneration system can become
more costly when the regenerator
off-gas stream contains compounds
for which emission is regulated.
About 99.96 wt% TEG can be
attained with stripping gas in the
regenerator. Licensed regeneration
schemes can reach 99.99 wt% TEG
and higher.
Regenerating TEG to a higher
weight per cent enables further
reduction of the water content (or
concentration) of the dehydrated
gas and the associated WDP. Water
absorption in the TEG contactor is
somewhat less sensitive to pressures up to 1500 psig, and the
required minimum (or equilibrium)
weight per cent of lean TEG entering the contactor can be estimated
from published correlations.3 For a
given lean TEG wt% and flow rate
to the contactor, the water content
in lb/MMSCF of the dehydrated
gas from the contactor reduces to a
lower level when the contactor
operates at a lower temperature.
Most contactors, however, operate
at 60-110F (15-43C) as excessively
low operating temperatures sharply
increase the viscosity of TEG and
condensation of heavy components
in the feed gas.
Moreover, the dehydrated gas
WDP will also depend on the
number of equilibrium stages in the
contactor. Generally, 99.96 wt%
lean TEG can dehydrate gas down
to -40F (-40C) WDP at 80F (27C)
contactor temperature. However,
for a given water content or
concentration of the dehydrated
gas, the WDP increases proportionally with pressure. To meet a given
WDP specification target, higher
operating pressures require lower
water content in the dehydrated
gas, and lower water contents
require higher concentrations of
lean TEG.
The water content of a gas in
equilibrium with a hydrate can be
lower than the water content of the
same gas in equilibrium with
supercooled or meta-stable water.
Mainly depending upon the gas

www.eptq.com

Merichem Company MERICAT J.pdf 1 5/21/2015 9:02:01 AM

THE FUTURE OF

JET FUEL TREATING

HAS ARRIVED

MERICAT J is an advanced liquid treating technology that removes

heavy mercaptans in jet fuel and middle distillate streams without using
caustic or carbon beds. MERICAT J utilizes Merichems FIBER FILM
Contactor as the mass transfer mixing device in combination with the
proprietary JeSOL-9 treating reagent to oxidize heavy mercaptans.
Since there is no fixed carbon bed, operators experience no downtime for
water or caustic wash maintenance or carbon bed change outs. This
significantly increases the on-stream performance and eliminates a
refinery waste stream. Additionally, the non-dispersive FIBER FILM
Contactor reduces capital expenditures and overall plant space
requirements, making MERICAT J the technology of choice.

CM

MY

CY

CMY

Follow Merichem on Social Media

@MerichemCompany

www.merichem.com/mericat-j

Pressure, psig

1200
Dew points
1100
Bubble points
1000
Critical point
900
800
700
600
500
400
300
200
100
0
250
200
150
100

Cricondentherm

50

Temperature, F

0
HDP

Figure 3 Plots of bubble and dew points

composition and water saturation

levels of natural gas, hydrate
formation temperatures for a
specific gas composition and water
content may be different from the
WDP temperatures in equilibrium
with meta-stable water. This difference is especially common for gases
at very low WDP temperature
regions where thermodynamically
unstable (meta-stable) water is
formed, and the targeted water
content may need to be reduced to
avoid hydrate formation.
Other
common
dehydration
schemes include injection with a
hydrate inhibitor (see Figure 2).
Methanol or ethylene glycol (EG)
are the most commonly used inhibitors and are typically applied in
sufficient quantities upstream of
process equipment cooling the gas
temperature. The injected inhibitor
absorbs water in the gas to prevent
freezing or hydrate formation, and
a regeneration still recovers the
inhibitor at a targeted concentration
from the water-rich inhibitor.
Generally, EG can be considered
when
downstream
processing
temperatures are not excessively
low (above -40F), and methanol is
adequate for much lower downstream processing temperatures (as
low as below -140F) and relatively
more susceptible to vapour losses
at high operating temperatures.
Applicable regulations should also
be reviewed before finalising the
selection of an inhibitor.
Relative to glycol absorption
using a contactor, an inhibitor injection scheme could require less

110 PTQ Q3 2015

capital cost. However, the capital

cost savings may become less when
the presence of water in process
equipment downstream of the
injection requires non-compact
design or materials with improved
corrosion resistance. In Figure 2, for
example, a brazed aluminum type
exchanger offers high overall heat
transfer coefficients and compact
design with low approach temperatures. Low approach temperatures
are essential for enhancing the efficiency of the refrigeration scheme

Relative to glycol
absorption using
a contactor, an
inhibitor injection
scheme could require
less capital cost
for controlling the HDP, and
brazed aluminum exchangers are
suitable for the gas-gas exchanger
servicing gas streams which have
been dehydrated. As the feed gas in
Figure 2 has not been dehydrated in
an upstream contactor or another
scheme, brazed aluminum exchangers may not be viable for the
gas-gas exchanger, and shell and
tube exchangers with acceptable
metallurgy will need to be installed
in series. Relative to brazed aluminum exchangers, the shell and tube
option requires higher capital costs
and larger plot spaces. Fluid veloci-

ties and effective mixing of the

inhibitor in the exchanger are also
essential for successful operation of
the dehydration scheme using
inhibitor injection.
Cooling the water-saturated feed
gas to lower, viable operating
temperatures for condensing water
out of the gas phase should also be
considered when designing a dehydration system. For example, air
cooled exchangers can be used
upstream of a TEG contactor to
cool the water-saturated feed to a
lower temperature and reduce the
water content in the gas by condensation. Reduced water content in
the gas feeding the TEG contactor
requires less TEG circulation and
minimises the operating and capital
cost of the TEG regeneration
system. While cooling the gas to
lower temperatures increases the
capital cost for the air cooled
exchangers which may need to be
designed with lower approach
temperatures, this gas cooling
option can be economically attractive for sites with low ambient
temperatures.

Hydrocarbon dew point control

A HDP control scheme is typically

preceded by, or integrated with, a
dehydration scheme, and the
targeted HDP is normally defined
as the highest hydrocarbon dew
point (cricondentherm, see Figure 3)
temperature at any pressure. In
addition to the targeted HDP, the
following main factors affect the
selection of the most cost effective
HDP control scheme:
Feed gas composition
Available feed supply pressure
and export delivery pressure.
Compositions
or
forecasted
compositions of the feed gas from
reservoirs determine process steps
necessary to meet HDP specifications
and
other
product
specifications which may include
maximum limits on H2S, CO2, and
heating value. Removal of acid gas
or other contaminants such as
mercury is necessary for environmental, safety, corrosion, or freeze
prevention reason, and integration
of the acid gas removal process and
the HDP control scheme will need
to be evaluated as necessary.

www.eptq.com

However, this article focuses on

controlling the HDP of sweet gas
with low or negligible percentages
of acid gases.
Assessment of the feed gas
composition for the lifetime of
reservoirs is essential to identify the
most economical scheme for
controlling HDP to meet the specified target. For raw gas containing
high percentages of ethane, for
example, the amount of ethane in
the natural gas export needs to be
adequately reduced to meet the
heating value and other specifications such as Wobbe number. When
a refrigeration scheme is considered
as an option, high percentages of
ethane and heavier compounds in
the raw gas also affect the horsepower requirement of refrigeration
compressors to meet a given
HDP specification. For locations
where ethane can be marketed
economically, a cryogenic processing unit with high ethane recoveries
can be an attractive option to
consider.
The feed gas supply pressure
available from the reservoir is an
important parameter for selecting
the most economical scheme for
controlling HDP. When the reservoir supply pressure is sufficiently
higher than the required delivery
pressure of the export sales gas to
the pipeline, an HDP control
scheme using a Joule Thompson
(JT) valve or a swirling supersonic
separator can be a relatively low
capital cost option.
A JT valve reduces the gas pressure in an isenthalpic condition to a
lower pressure where the gas
expands and becomes cooler as
some of the inlet enthalpy converts
to kinetic energy. Equation 1
defines the JT coefficient describing
a change of temperature at an isenthalpic change of pressure:
JT = ( T /P) H

(1)

JT is positive for natural gas in

normal or typical operating pressure and temperature region, and a
JT valve can be used to control
HDP in the operating region where
JT is positive. JT is theoretically
zero for an ideal gas and can

www.eptq.com

Dehydrated gas
from a glycol
contactor

Gas-gas
exchanger

Sales
gas

Sales gas
compressor

Gas-liquid
exchanger

Turboexpander

Recompressor

To
recovery
Liquid
hydrocarbon

Figure 4 Turbo-expander HDP control

become negative for real gases at

specific regions.
At the outlet of a JT valve, the
heavy components of the gas
condense at low temperatures, and
the condensate can be separated in
a downstream separator which
generates a cool gas stream with
lower HDP. This cool gas stream is
typically routed to a heat
exchanger for cooling the feed gas
stream to the JT valve. Lower
expander inlet temperatures reduce
the pressure drop across the JT
valves necessary to meet the specified HDP. As the HDP specification
becomes lower for a given gas
composition, the JT valve needs to
operate with a larger pressure drop
to reach a lower outlet temperature
for meeting the lower HDP specification. Recompression of the gas
stream from the separator becomes
necessary when the required
export sales gas delivery pressure
is higher than the JT valve outlet
pressure. Consequently, the JT
valve option generally becomes
favourable when the available feed
gas supply pressure is adequately
higher than the required delivery
pressure of the product gas, and
recompression is not necessary.
However, available feed gas supply
pressures throughout the lifetime
of the reservoir may need to be
considered.
Somewhat similar to a JT valve, a

supersonic separator device such as

Twister reduces HDP through pressure reduction of the gas stream.
However, a supersonic separator
involves a converging and diverging nozzle with a design feature
causing cyclonic motion or swirling
of the cooled gas stream to enhance
vapour-liquid
separation.
The
converging and diverging nozzle
enables the gas stream to reach
supersonic velocities. Compared to
a JT valve which typically cannot
reach a velocity higher than sonic
velocity, a supersonic separator
reportedly requires less pressure
drop to reach a given HDP target.
Generally, relative to a JT valve, a
supersonic separator achieves a
larger dew point suppression at a
given pressure drop or requires less
pressure drop for the same
HDP target.
For cases with a delivery pressure
higher than the feed gas supply
pressure, HDP control schemes
typically involve a refrigeration
system or a turbo expander. An oil
absorption scheme is another
option where lean oil absorbs the
heavy components in the feed gas,
and removal of heavy components
reduces the HDP of the product gas
leaving the absorber. Heavy
components absorbed in the rich oil
are separated in a stripping
column, regenerating lean oil at the
bottom. The oil absorption option

PTQ Q3 2015 111

requires large items of equipment

and high energy consumption and
seems to be less popular for remote
project sites.
When the available feed gas
supply pressure is lower than the
required export sales gas delivery
pressure, HDP control options
using refrigeration (see Figure 2) or
a turbo expander (see Figure 4) will
be the likely cost effective candidates. The refrigeration option
shown in Figure 2 is a typical single
stage vapour compression cycle
generating
a
refrigerant
(for
instance, propane) stream at a low
temperature adequate for cooling
the feed gas stream in the refrigerant chiller. The fraction of the feed
gas is condensed in the chiller to
meet the HDP target. The power
consumption of the refrigerant
compressor
can
be
typically
reduced by 20% or more if optimisation schemes such as heat
integration, multi-component refrigerant, and increased compression
stages are implemented. However,
these schemes generally increase
the complexity and capital cost of
the refrigeration system. Relative to
the turbo expander option for
controlling HDP, the refrigeration
option can be more cost effective.
The turbo expander option for
HDP control expands the feed gas
to a lower pressure and generates
power. Compared to a JT valve
which expands to a low pressure
under isenthalpic condition, the
pressure reduction through an
expander is somewhat close to isentropic condition, and the enthalpy
of the gas at the expander outlet is
lower than that at the expander
inlet. As such, for a given pressure
drop across a JT valve or an
expander, the gas outlet temperature will be lower at the expander
outlet relative to that of a JT valve.
In other words, an expander
requires less pressure drop to reach
an outlet temperature target for
HDP control. The enthalpy difference across an expander is
essentially all converted to the
power which can be calculated
typically in a simulation model or
may be estimated by the isentropic
efficiency of the expander (see
Equation 2):

112 PTQ Q3 2015

H.P. Expander = 1.4145 m ( hs)

(2)

where:
m
= gas flow rate in lb/sec
hs = isentropic enthalpy difference in btu/lb between expander
inlet and outlet

= isentropic efficiency
H.P. = expander power output,
horsepower
The efficiency generally reaches
around 80% but varies with specific
expanders. The expander commonly
shares a common shaft with the
recompressor which operates using
power generated by the expander.
About 98% of expander power is
available for the recompressor. As
the adiabatic or polytropic efficiency of the recompressor will be
much less than 100% (or isentropic),
the recompressor discharge pressure is significantly less than the
inlet pressure to the expander. For
cases where a downstream sales gas
compressor is needed to reach the
required sales gas delivery pressure,
the pressure reduction across the

The turbo expander

option can be
attractive for remote
project sites where
modular construction
offers significant cost
savings
turbo expander-recompressor is the
primary reason that turbo expander
option requires a larger sales gas
compression power than the refrigeration option. As such, relative to
the refrigeration option the turbo
expander option generally is more
energy intensive but more compact
with less and smaller equipment.
The turbo expander option can be
attractive for remote project sites
where modular construction offers
significant cost savings.

Conclusion

To meet water and hydrocarbon

dew points specifications for natural
gas
without
acid
gas
components, the following may be

considered as general guidelines

for selecting the most cost effective
processing scheme for a particular
application.
TEG absorption with a contactor
or injection with a hydrate inhibitor
are typically the most cost effective
options for meeting the WDP
targets of pipeline natural gas.
While adsorption with molecular
sieve or other suitable sorbents is
required for very low WDP targets
meeting the requirements of cryogenic gas processing or LNG plant,
this adsorption option requires
relatively higher capital and operating costs. Combination of both
absorption and adsorption could
result in cost reduction.
For a given gas composition, the
most cost effective option for HDP
control mainly depends on the
available feed supply pressures and
product delivery pressure requirement. When the feed supply
pressure is sufficiently higher than
the delivery pressure, or when
product gas compression to the
delivery pressure is not necessary,
the JT valve option or the supersonic separator option will likely be
the most cost effective option. The
refrigeration or turbo expander
option becomes the likely cost
effective candidate when the feed
gas supply pressure is lower than
the delivery pressure. While the
refrigeration option may cost less
than the expander option, this
option is more compact and attractive to projects where less and
smaller equipment offers significant
savings in construction costs.
Twister is a mark of Twister BV.
References
1 Gas Processors Suppliers Association (GPSA),
Engineering Data Book, 12th Ed, 2004.
2 Hernandez-Valencia V N, Hlavinka M W,
Bullin J A, Proceedings of the Seventy-First
GPA Annual Convention, Tulsa, Oklahoma: Gas
Processors Association, 1992, 310-317.
3 Bahadori A, Vuthaluru H B, Journal of Natural
Gas Science and Engineering,1, 2009, 68-71.
Tek Sutikno is a Process Engineering Manager
with Fluor and a Professional Engineer registered
in 11 US states with over 25 years experience
in the process industries. He holds BSc, MSc,
and DEngr degrees in chemical engineering and
a MBA degree, all from the University of Kansas.
Email: tek.sutikno@fluor.com

www.eptq.com

Bringing your project

Full Circle
Concept through facility start-up

Our expertise in the refining, chemicals, petrochemicals, and polymers

industries ensures that we deliver Predictable Results
to your capital project from start to finish.

832.809.8909

marketing.woodgroupmustang.com/fullcircle

cattec.indd 1

26/02/2013 16:59

Testing and experience in FCC

catalyst selection
Resolving differences between test results and unit experience in FCC
catalyst selection
CAREL POUWELS Albemarle Catalysts Company
KEN BRUNO Albemarle Corporation

www.eptq.com

VGO testing
24%

VGO non-testing
19%

Resid testing
28%

Resid non-testing
29%

Figure 1 Half of the FCC market is based

on catalyst testing

unit is caused by insufficient simulation of catalyst deactivation as

well as the testing methodology.
These deficiencies penalise low
Z/M catalysts with excessively
high coke yields while at the same
time giving insufficient credit to
better bottoms conversion. When
the test results are plugged into
FCC models without any correction, these high coke yields from
the test lead to unrealistically

high regenerator temperatures.

Consequently, low Z/M catalysts
are considered high risk and are
almost never selected. Figure 3 illustrates the large coke penalty for
laboratory deactivated catalysts
(D-cat) compared to the FCC unit.
When equilibrium catalysts (E-cat)
are used, the shortcoming of the
testing method becomes apparent
and clearly shows the difference
with reality defined by the FCC
unit.
For decades Albemarle has been
supplying the broadest range of
catalysts from very high Z/M to
very low Z/M and has experienced
dozens of cases where refiners have
switched between such different
types of catalysts. These cases
confirm that low Z/M catalysts do
not show excessively high coke
selectivity or strongly increasing
regenerator temperatures. More
details of 19 side-by-side cases have
been reported.
Recently, several testing refiners

100

Proportion of catalyst type

selected by each type of
customer, %

or refiners, the choice of an FCC

catalyst is a challenging task. Is
the incumbent catalyst the best
option for the FCC unit and delivering the highest margin? Or should
the catalyst be optimised through a
reformulation? Or is there perhaps a
different and better technology?
About half of the industry relies on
the R&D department or a third
party laboratory to carry out a catalyst testing programme as their
main tool to determine the best catalyst. This is valid for both VGO (gas
oil) applications as well as for FCC
units processing resid feedstock (see
Figure 1).
We have reported about this
market segmentation on earlier
occasions.1 A more detailed analysis of the FCC market then revealed
that there is a remarkable difference between catalyst technologies
used by those refiners that test and
those that select their catalyst in
different ways.
In the market analysis that was
carried out, Albemarle found that
refiners that test predominantly use
catalysts which are high in zeolite
content and contain relatively low
levels of alumina matrices as active
ingredients. In this article we will
further refer to these catalysts as
high Z/M catalysts. Refiners that
do not test mostly use catalysts
which are relatively moderate to
low in zeolite and moderate to high
in alumina matrices and will be
referred to as low Z/M catalysts.
This observation was made for all
feed types, but is particularly
evident for resid applications (see
Figure 2).
This disconnect between laboratory testing and reality in the FCC

High Z/M
Low Z/M

90
80
70
60
50
40
30
20
10
0

Testing

Non-testing

Figure 2 Opposite catalyst choices in resid applications for refiners that test and
do not test

PTQ Q3 2015 115

Commercial case

Properties of catalysts used

Al2O3
52
44
59

Cat#1
Cat#2
Cat#3

RE2O3
1.9
2.5
0.6

SA
238
313
269

MSA
93
74
132

Z/M
1.6
3.3
1.0

Table 1

Slurry (low Z/M) Slurry (high Z/M),

wt%

took a novel approach to their catalyst selection. These refiners

previously chose their supplier by
thorough catalyst testing and selecting the catalyst that had the best
laboratory results. But this time,
analysis of test results either was
modified, or the refiner decided to
trial a catalyst that was not the
winner of the testing programme,
and/or the refiner decided on other
grounds. Examples include:
Resid FCC unit 1 in the AsiaPacific region trialled a low Z/M
catalyst that reduced bottoms and
increased LCO, contrary to test
results evaluated at constant coke
Resid FCC unit 2 in the AsiaPacific region trialled a low Z/M
catalyst that increased gasoline and
LPG olefins, contrary to test results
evaluated at constant coke
VGO FCC unit 1 selected the
winning maximum propylene catalyst based on testing. Unit yields
were not in line with the test results
and the refiner trialled a second
catalyst which was far superior and
exceeded expectations based on the
test
Major refiners 1 and 2 apply a
correction to lab coke yields for all
catalysts (compressing the coke
differences)
Major refiner 3 applies a correc2
1
0

tion to lab coke based on excessive

lab H2 yields
Major refiner 4 applies a correction to lab coke based on the matrix
content of the catalysts
Multiple refiners that start with
low levels of a catalyst of interest
and
increase
its
percentage
step-wise.
These refiners are happy with the
new technology they are applying.
It opened the door to more technology options and increased their
FCC margins.
Albemarle recognises the importance of catalyst testing. Catalyst
testing is an essential tool to
develop new catalysts with innovative ingredients or a novel assembly
technology and which brings the
performance that refiners are seeking. Catalyst testing is obviously
very important for those refiners
that choose to select their catalyst
by testing. At Albemarle we realise
that catalyst testing has its benefits.
The results we get in testing are not
identical to the changes obtained
in the unit. All test data need to
be corrected and translated. When
a new catalyst technology has
been used, a post evaluation is
required for further refinement of
the correction and translation
model.

FCC unit
E-cat
D-cat

1
2
3
4
-2

1
0
1
2
3
Coke (low Z/M) Coke (high Z/M), wt%

Figure 3 Catalyst lab study leads to unrealistic coke yields and penalises low Z/M catalysts

116 PTQ Q3 2015

Refineries that regularly switch

catalysts are interesting subjects for
learning purposes. This is particularly the case when these switches
include different suppliers and thus
different catalyst technologies. One
of those cases concerns a refiner
that is continuously seeking the
optimum catalyst for its operations
and switches regularly. Due to the
relatively small inventory and high
catalyst consumption of this resid
application, the change-out time is
short. Catalyst evaluations can thus
be done in a short time frame and
the incumbent catalyst supplier can
respond quickly for further fine
tuning if required.
This FCC unit processes a mix of
different feeds, including a substantial amount of residue. The feed
has a specific gravity of 0.91 and
Conradson carbon residue above
2 wt%. Metals on equilibrium catalyst vary over time, but are
typically of the order 3000 ppm
nickel and 6000 ppm vanadium.
The refiner operates this FCC unit
to maximise LCO and LPG olefins.
The main limitations are air blower,
wet gas compressor and regenerator temperature.
The commercial case which we
discuss concerns three different
catalysts from three different manufacturing technologies (see Table 1):
Cat #1: moderate Z/M catalyst
with moderate RE2O3
Cat #2: high Z/M catalyst with
high RE2O3
Cat #3: low Z/M catalyst with
low RE2O3.
In the study, yields of all three
catalysts are compared against each
other. As in almost all FCC units,
conditions are never exactly the
same. In order to make a sound
comparison, one needs to normalise
the cases. This means that products
need to be corrected to the same
cut-points, feed quality, and
process conditions through a FCC
process simulation model. At identical conditions, all three catalysts
can then be compared.
In addition to the comparison of
FCC unit data, a comparison
between the three catalysts was
made by laboratory testing. All
three catalysts were deactivated by

www.eptq.com

www.eptq.com

yields against base case,

wt%

6
4
2
0

2
4
6

Cat No. 2 FCCU

Cat No. 2 Test
DG

LPG

GLN

LCO

Slurry

Conv

Figure 4 Comparison of yields in the FCC unit and ACE test between Catalyst #2 and the
base catalyst

The dry gas yields are lower for

Catalyst #2 and Catalyst #3 in the
unit, but are found to be higher in
the test
A very strong reduction of LPG
yield was found in the test for
Catalyst #3, but the opposite was
found in the unit with a substantial
increase against the base case
The ACE test showed an
increased
gasoline
yield
for
Catalyst #3, while in the unit no
difference was seen against the
base case
LCO difference in the test
between Catalyst #2 and Catalyst
#3 is large, while in the unit it is
much closer. Slurry increases in the
test for Catalyst #3, while in the
unit it declines almost similarly to
Catalyst #2.
In summary, when we look at the
total conversion, Catalyst #2 and
Catalyst #3 have increased against
the base case in the unit. However,
the test results show an even rosier

yields against base case, wt%

so-called cyclic deactivation (CD)2

whereby metal levels were applied
similar to the typical metal levels
found on equilibrium catalyst in
this FCC unit. With CD, the metals
are added through cracking reactions followed by a number of
cycles under oxidising and reducing conditions at high temperatures
and in the presence of some steam.
In this very controlled method, an
identical amount of nickel and
vanadium is deposited on each of
the three catalysts. Subsequently,
the catalysts are subjected to ACE
testing at four different cat-to-oil
ratios. Hereby, a heavy feedstock is
used which is considered representative for the application. The yields
obtained can thus be plotted
against different parameters, such
as catalyst-to-oil ratio (CTO),
conversion, or coke and interpolated
at
preferred
constant
conditions. This is a very common
method that most R&D centres
follow. Often one compares the
catalysts at a conversion level similar to the conversion level in the
FCC unit and compares the yield
selectivities. Another method is a
comparison at constant coke yield.
The idea behind this method is that
the FCC unit also typically operates
at a constant coke yield (through
the heat balance).
This latter method is very often
applied in the industry and we
have thus chosen to do the same
for this case study. The three catalysts are compared against each
other whereby Catalyst #1 is
considered the base case. Yield
deltas of Catalysts #2 and #3
against this base case can thus be
obtained in the unit as well as from
the ACE unit.
We see a fair resemblance
between the FCC unit and the ACE
test yields for Catalyst #2, meaning
that the deltas between Catalysts
#2 and the base case are most times
in the same direction (higher or
lower) and, sometimes, approximately the same magnitude (see
Figure 4).
But for Catalyst #3 the results are
inconsistent between test and FCC
unit, as they are completely opposite (see Figure 5). The following is
observed from Figures 4 and 5:

picture for Catalyst #2 with a larger

conversion delta, and the simulation
of Catalyst #3 is awed. The test
results for Catalyst #3 suggest a
reduction in conversion, while in
reality a substantial conversion
increase is observed in the FCC unit.
As was said earlier, several yields
look consistent between FCC unit
and test. But there are quite a
number of yields where the differences between FCC unit and test are
large or even opposite. We do not
take any position on which catalyst
is the best. The point we want to
make is that test results can be very
different from reality. If this catalyst
testing programme would have been
used prior to commercial use to
select the catalyst, it is not difcult to
imagine that one catalyst could have
been rejected, while it would do well
in the unit. Or possibly a catalyst is
selected as the winner, but in the
FCC unit shows a worse picture and
leads to margin loss.

Cat No. 3 FCC unit

Cat No. 3 Test

4
2
0

2
4
6

DG

LPG

GLN

LCO

Slurry

Conv

Figure 5 Comparison of yields in the FCC unit and ACE test between Catalyst #3 and the
base catalyst

PTQ Q3 2015 117

Main differences in process conditions between lab and FCC unit

Lab
5 hours-2 days
788 or higher
50-100%
Ambient

Age deactivated catalyst

Temperature, C
Steam, %
Pressure, barg

FCCU
2-12 months
680-740
5-10%
1.5-3

Table 2

One of the reasons we see these

differences is related to the coke
artifact (a lab result that is not
equally observed in an FCC unit)
that is found with low Z/M catalysts. Catalyst #3 in the example
above is relatively low in Z/M as
reported in Table 1. When we make
a comparison at equal coke, the
high but unrealistic coke yield in
testing of such catalysts aggravates
and exaggerates the comparison at
equal coke. We recommend not
looking at iso-coke tables from a
lab test as coke differences are
amplified, but instead, as a first
step, use corrected iso-conversion
data. (Correction methods will be
discussed later in the article.)
One should not take testing data
as such and select a winner.
Corrections,
translations
and
commercial data should be added
to complement the catalyst selection and warrant a more sound
choice for the next catalyst in the
FCC unit. In the following, we will
zoom in on the pitfalls in deactivation and testing. What are the
reasons that testing and FCC units
differ, and what methods can be
more successful in selecting a better
or winning catalyst?
160

Surface area, m2/gm

150
140

Pitfalls in catalyst deactivation

and testing

One of the key properties that

refiners focus on is the surface area
(SA) of the equilibrium catalyst. It
is a crucial property to monitor as a
measure for the activity of the catalyst. The SA of the equilibrium
catalyst is measured and reported
routinely by the catalyst supplier
and the refiner uses this to set the
catalyst addition strategy for the
FCC unit. When refiners carry out a
catalyst selection by laboratory testing, the SA of the catalyst plays an
important role too. A common
approach is that R&D centres deactivate the incumbent fresh catalyst
and optimise the protocol to duplicate the SA of the incumbent
equilibrium catalyst. The deactivation method is tuned to mimic not
only the total surface area, but also
attempts to mimic the individual
zeolite SA and meso SA. This is the
basis of a proper simulation of the
E-cat. In addition, one tries to
mimic the activity level, the unit
cell size of the zeolite in the catalyst, and the level and activity
of metals like nickel and vanadium
deposited on the catalyst. Each
such property added increases

E-cat
D-cat

130
120
110
100
90
80
70
60

Cat No. 1

Cat No. 2

Cat No. 3

Figure 6 Total surface area of equilibrium catalysts compared to lab deactivated catalyst

118 PTQ Q3 2015

the complexity of this exercise.

When the matching deactivation
protocol is found, this method is
typically applied to all the candidate catalysts that are included in
the catalyst selection programme.
The assumption is that using the
same protocol for all the catalysts
will produce realistically simulated
E-cats of all candidate catalysts.
While on the surface this assumption might seem quite logical, one
can forget that the applied protocol
is a very severe simulation of reality applied in order to complete
deactivation within a practical
amount of time. The average age of
deactivated catalyst particles is
several months, but no R&D centres
take that time to simulate a real
equilibrium catalyst. Consequently,
deactivation temperature and steam
are increased to speed up deactivation (see Table 2). Thus a very
severe deactivation process takes
place and the catalyst deactivates at
a faster rate compared to reality.
The drawback is that under these
conditions surface areas can be
matched for one catalyst, but not
for other catalysts when they are
different in technology and particularly when the ratio in zeolite and
matrix is different. We will now
zoom in on the deactivation details
of the three catalysts applied in the
commercial case study (see Figures
4 and 5).
Figure 6 illustrates the (total)
surface area of the three catalysts
used in the deactivation and testing
programme. The fresh catalyst data
are in Table 1.
The surface area of laboratory
deactivated catalyst Catalyst #1
matches nicely with the equilibrium
catalyst (E-cat) #1. Catalyst #2
however deactivates to a higher SA
than its E-cat, while laboratory
deactivation of Catalyst #3 is very
different from its E-cat. When we
zoom in on the individual types of
surface area, we see that Catalyst
#1 has also a good match on zeolite
surface area, but that both Catalyst
#2 and Catalyst #3 show a substantially higher ZSA for the D-cats (see
Figure 7). While Catalyst #2 has
25% more ZSA, Catalyst #3 more
than doubles its ZSA when D-cat is
compared to E-cat.

www.eptq.com

FCC iron trap

Our INTERCATJM FCC additive, CAT-AID, not
only traps vanadium, it has also been found
to prevent iron nodules from forming on the
surface of the base catalyst particle.
When the base catalyst is cured from iron
poisoning, the inner pores once more become
available for cracking reactions. With CAT-Aid,
iron poisoning can be cured without reformulating
fresh catalyst or reverting to flushing with e-cat.

Base catalyst before CAT-AID

0.8 wt% Fe

UK

Base catalyst after 2 months

of CAT-AID 0.8 wt% Fe

Tel +44 (0)1642 553601

www.jmprotech.com

Fax +44 (0)1642 522542

Zeolite surface area, m2/gm

120
110
100
90
80
70
60
50
40
30
20
10
0

these D-cats will not properly simulate

the
behaviour
of
the
corresponding E-cats in the FCC
unit. We have demonstrated that in
the previous section of the article
(Figures 4 and 5). This case study
illustrates that comparing different
catalyst technologies and different
Z/M catalysts in one deactivation
and test protocol can lead to wrong
conclusions and thus to not selecting the best catalyst for the unit.

E-cat
D-cat

Cat No. 1

Cat No. 2

Cat No. 3

Figure 7 Zeolite surface area of equilibrium catalysts compared to lab deactivated catalyst
Catalyst properties after Deactivation Method X

SA, m2/g
ZSA/MSA
UCS, Angstrom
Brnsted acidity (relative)
Lewis acidity (relative)
AAI (relative diffusion)

E-cat A
139
2.95
24.29
1.00
1.46
Low

D-cat A
172
2.27
24.28
0.75
1.17
High

E-cat B
124
1.43
24.39
0.99
1.94
Moderate

D-cat B
127
0.58
24.51
0.77
2.70
High

Table 3

The simulation of the meso (or

matrix) surface area is again different. As Figure 8 shows, Catalyst #1
is simulated well with a close
match between E-cat and D-cat and
also the MSA of Catalyst #2 is
reasonably comparable for both
E-cat and D-cat. However, Catalyst
#3 is not simulated properly as the
D-cats MSA is about 25% higher
than for the E-cat.
In summary we have three different results:
Catalyst #1 matches total surface

Meso surface area, m2/gm

100
90
80

area, and also matches zeolite and

meso surface areas
Catalyst #2 matches well on
meso surface area, is a bit high on
total surface area, while zeolite
surface area is much too high
Catalyst #3 does not match at all
and deviates very strongly in all
types of surface area, D-cat SA
being much higher than E-cat SA.
As the magnitude and direction
of these deviations are different for
the three technologies, it is easy to
imagine that performance testing of

E-cat
D-cat

70
60
50
40
30
20
10
0

Cat No. 1

Cat No. 2

Cat No. 3

Figure 8 Meso surface area of equilibrium catalysts compared to lab deactivated catalyst

120 PTQ Q3 2015

Varying deactivation conditions to

replicate E-cat properties

Making D-cats in the laboratory

which are 100% identical to the
corresponding E-cats should be
possible when we apply truly identical conditions, as in the FCC unit.
This means the same (long) deactivation time would be needed. As
this is impractical for R&D centres
the question is: can one find an
accelerated deactivation protocol
with enhanced deactivation severity, whereby not only E-cat
properties but also performance is
matched?
To help answer this question, we
collected E-cats from Catalyst A
and Catalyst B, both used in the
same FCC unit under as similar as
possible feed, process, and catalyst
addition rate conditions. Catalysts
A and B were different technologies, evidenced by their very
different zeolite-to-matrix ratios
(see Table 3). We then deactivated
each of these fresh catalysts in the
laboratory under many different
conditions,
considering
time,
temperature,
partial
pressures,
metals, and age distributions. The
goal was to try to match the properties of the D-cats with their
companion E-cats.
Results for Deactivation Method
X are given in Table 3. These conditions resulted in the closest match
between the E-cat and D-cat SA for
Catalyst B. However, all other D-cat
properties for Catalyst B are far
from those of its E-cat. Consider
that the ZSA/MSA ratio is 1.43 for
E-cat B, while it is 0.58 for D-cat B.
It is quite common that accelerated,
severe deactivation methods preferentially deactivate the zeolite,
biasing
the
testing
yields.
Moreover, while we reasonably

www.eptq.com

sabin.indd 1

27/02/2013 15:01

Catalyst properties after Deactivation Method Y

SA, m2/g
ZSA/MSA
UCS, Angstrom
Brnsted acidity (relative)
Lewis acidity (relative)
AAI (relative diffusion)

E-cat A
139
2.95
24.29
1.00
1.46
Low

D-cat A
156
2.3
24.28
0.72
1.30
High

E-cat B
124
1.43
24.39
0.99
1.94
Moderate

D-cat B
110
0.43
24.38
0.67
2.58
High

Table 4
Catalyst properties after Deactivation Method Z

SA, m2/g
ZSA/MSA
UCS, Angstrom
Brnsted acidity (relative)
Lewis acidity (relative)
AAI (relative diffusion)

E-cat A
139
2.95
24.29
1.00
1.46
Low

D-cat A
179
2.16
24.29
0.94
1.38
High

E-cat B
124
1.43
24.39
0.99
1.94
Moderate

D-cat B
130
0.58
24.45
1.02
3.05
High

Table 5

matched the SA for Catalyst B, we

could not simultaneously match the
SA of D-cat A with E-cat A. We did
however match the unit cell size
(UCS) for Catalyst A, but not
Catalyst B.
Consider
the
results
for
Deactivation Method Y (see Table
4). Under these conditions, we were
able to reasonably match the UCS
for both D-cats with their respective E-cats. However, Deactivation
Method Y that enabled a match in
UCS for both Catalysts A and B
resulted in losing the SA match for
D-cat B and E-cat B obtained with
Deactivation Method X (see Table
3). Generalising, changing the deactivation method to better match
properties for one D-cat/E-cat pair
results is poorer agreement for a
different D-cat/E-cat pair. In our
experience the application of a
single laboratory protocol (other
than using conditions equal to
those in the FCC unit) to different
catalysts will not result in matching
the properties of all D-cats to their
E-cats.
Also, note that the ZSA/MSA in
Table 4 remains very different
between the D-cat and E-cat as the
zeolite continues to be deactivated
too severely in comparison with the
matrix. In fact the ZSA/MSA for
D-cat B is so low that very little
zeolite
remains.
Finally,
the

122 PTQ Q3 2015

Brnsted and Lewis acidities are

also quite dissimilar, and different
acidities portend different cracking
chemistry. With these differences,
there can be no expectation that
yields and selectivities from D-cat
A and D-cat B should match or be
close to their respective E-cats.
The final comparison is given in
Table 5. In this case, we were able
to reasonably match the Brnsted
acidities between both E-cats and
D-cats, suggesting that maybe the
cracking chemistry could be reasonably represented by the D-cats.
However, the Lewis acidities do
not match, especially for D-cat B
which is exorbitantly higher than
E-cat B. Lastly, the application of
Deactivation Method Z (see Table
5) resulted in losing the UCS match
for Catalyst B that we had with
Deactivation Method Y (see Table
4). Again, the inference is that the
application of a single (accelerated)
deactivation method to different
catalysts will not allow a match
between pairs of D-cat and E-cat.
Actions taken to solve one problem
undo previous successes. This
evidence suggests the deactivation
mechanism is a path function, not a
state function.
The
Albemarle
Accessibility
Index (AAI3) is a measure of a catalysts diffusional properties, both
surface and internal. Note that the

AAI for both D-cats is quantified as

high for all three deactivation
methods presented (see Tables 3, 4
and 5), while the E-cats had low
and
moderate
AAI
values.
Albemarle has developed a laboratory deactivation protocol that
generates the correct AAI (called
CD-ALFA4). This protocol requires
seven days and the application of
iron and/or calcium. However,
even after a seven-day deactivation, some subset of TSA, ZSA/
MSA, UCS, and acidities does not
match the E-cat.
The E-cats and the D-cats from
the many deactivation methods
applied to Catalysts A and B were
performance (yields) tested in the
laboratory. Not unexpectedly, the
yields for the D-cats were dependent on the deactivation protocol (or
method). First, the yields from each
D-cat did not match those from its
E-cat, the magnitude of the differences influenced by the deactivation
protocol. Second, when comparing
the results from E-cat A versus
E-cat B with those from D-cat A
versus D-cat B, the yields reversed
for some protocols.
In summary, we were not able to
find laboratory deactivation conditions that generate a D-cat that
replicates its E-cat. Moreover it
appears that the application of a
single lab protocol to multiple catalysts will not replicate the E-cat
properties of different catalysts
(especially catalysts made with
different technologies), meaning it
is not scientifically sound to apply
the same laboratory deactivation
method to multiple catalysts. Of
course one may argue that our
study was not exhaustive as all the
essentially infinite number of
combinations of deactivation conditions were not tried. This is true,
but the preponderance of evidence
leads us to believe that an accelerated and aggressive laboratory
deactivation does not replicate
E-cat and does not affect all catalyst
technologies in a manner consistent
with what occurs in an FCC unit.

Simulating zeolite and matrix

deactivation rates

Simulating the total SA of the equilibrium catalyst is only one of the

www.eptq.com

AFPM 2015
NAVIGATE NEW TRENDS
ATTEND AFPM FALL MEETINGS
The global energy landscape is rapidly changing. Growth in production
is leading to changes in technology, planning and management at all levels.
Gain a competitive advantage. Come to 2015 AFPM Meetings to hear
about the latest developments impacting our industry.
Q&A and
Technology Forum
New Orleans, LA
October 4 7

Environmental
Conference
Salt Lake City, UT
October 18 20

International
Lubricants &
Waxes Conference
Houston, TX
November 12 13

www.afpm.org/meetings/

AFPM_GeneralMeetingsAd_PTQ_v1.indd 1

5/20/15 4:09 PM

basic properties to match, and

adding more properties increases
the complexity. We will now focus
on techniques used to attempt to
match both the zeolite and matrix
surface area. In the following example we will demonstrate that the
typical laboratory deactivations are
too short and require conditions
too severe to achieve the proper
zeolite and matrix surface areas
(hence activity levels) present in the
FCC unit. Many R&D centres typically apply higher temperatures
(and too high steam rates) to speed
up the deactivation. When we
compare equilibrium catalysts with
laboratory deactivated catalysts, we
often see that, relative to E-cat, the
zeolite has been deactivated too
much in the laboratory, and that
the matrix is relatively too active.

Imbalance in zeolite and matrix

deactivation
E-cat
74
38
24.38

MSA
ZSA
UCS

D-cat
74
31
24.35

Table 6

Table 6 illustrates this scenario:

while obtaining equal matrix SA,
the zeolite SA for the lab is too low
and this is also reflected in the unit
cell size.
In order to get a good match in
both zeolite and matrix deactivation we look for a way to change
the relative rate of deactivation for
zeolite and matrix. We need to
have a method whereby we either
speed up the deactivation of matrix

Surface area retention, %

100
90
80
70
60
50

Accessibility after a laboratory

deactivation

40
30
20

Matrix
Zeolite

10
0
785

790

795

800

805

810

815

820

Deactivation temperature, C
Figure 9 The effect of temperature on the deactivation rates of zeolite and matrix
20

Low accessibility catalyst

High accessibility catalyst

18
16

Fresh

14

AAI

12
10
8
6
4

Fresh

AL
FA

C
D

ca
t
ea
m
in
g

E-

St

AL
FA

D
C

St

Eca
t
ea
m
in
g

Figure 10 The common laboratory deactivation methods are not able to simulate the
proper catalyst accessibility

124 PTQ Q3 2015

at equal rate for zeolite or to slow

down the rate of zeolite deactivation at equal rate for matrix. For
example, if a shorter deactivation is
applied, one can match the zeolite
SA, but then the matrix SA in the
lab deactivated catalyst will be too
high.
In addition to changing time,
another generic tool that laboratories use is the deactivation
temperature. We observe that many
laboratories are increasing the
severity by making it hotter, rather
than reducing the temperature.
Figure 9 shows that an increase in
temperature
further
enhances
zeolite deactivation, while matrix
deactivation is influenced very
little. Increasing the temperature
from 788C to 815C substantially
decreases zeolite retention, while
matrix retention is marginally
affected. The imbalance between
zeolite and matrix deactivation is
thus further exacerbated. In conclusion,
one
should
reduce
temperature. But at lower temperature, a longer deactivation will be
required to match the E-cat.

When
laboratories
focus
on
mimicking SA and activity, they
hardly ever take into account the
physical
degradation
of
the
morphology of the catalyst particle.
This process whereby the particles
get denser over time is complex
and slow. During this degradation,
the catalyst often loses its accessibility (expressed by AAI3) and this
matters particularly when resid is
being processed. As accessibility is
a crucial parameter of the catalyst,
failing a proper simulation of this
property will further aggravate the
situation and lead to wrong test
results. The differentiation between
good catalysts (with high AAI) and
mediocre catalysts (with low AAI)
is thus often neglected. Figure 10
shows that E-cat almost always has
lower AAI than the fresh catalyst.
But for the most common laboratory deactivation methods, the AAI
can even go up or remain relatively
high. This is valid for steaming and
cyclic deactivation, and also for
cyclic propene steaming. Only

www.eptq.com

The stripper plays an important

role in the FCC unit to remove
entrained and adsorbed hydrocarbons, which are carried over to the
regenerator with the spent catalyst.
Allowing the hydrogen-rich hydrocarbons into the regenerator leads
to losses in liquid products,
throughput and catalyst activity.
Though a proper stripper design
can enhance its performance, stripping is also influenced by the type
of catalyst. This effect is often
neglected in catalyst selections.
When FCC unit strippers operate at
sub-optimal
catalyst
residence
times, accounting for the catalyst
strippability in the lab test is very
important.
Most
laboratories
however do not take the effect of
catalyst stripping into account.
Instead, yield data are collected
only at catalyst strip times well in
excess of the commercial range,
meaning complete stripping is
simulated. This often leads to questionable catalyst selections.
Bench-scale units like FST or ACE
can measure strippability of different catalysts quite well. At constant
catalyst-to-oil (CTO) and temperature, coke yields are determined at
different stripping times. Figure 11
illustrates such an experiment, by
which the stripping behaviour is
measured
for
one
catalyst.
Typically stripping times in benchscale
units
are
long
and
consequently only the so called
hard coke is determined. In real
life FCC operations, stripping times
are in the order of about 90 seconds
and quite often the stripper design
is not optimal to remove all the
strippable hydrocarbons. The catalyst going to the regenerator thus
contains a substantial amount of
entrained or adsorbed hydrocarbons, which we call soft coke.

www.eptq.com

12

Carbon to regenerator,
wt% of feed

Strippability influences the coke

yield

13
11
10
9
8
7
6

Soft coke

5
4
3

Hard coke

100

200

300

400

500

600

700

Strip time, seconds

Figure 11 Stripping study determining soft coke versus time

A stripping study was carried out

on three catalysts of different technologies. Yields were determined at
typical bench scale conditions, 900
seconds. Also yields were achieved
at typical FCC unit conditions, 90
seconds. Due to the difference in
stripping times, some yields were
impacted quite strongly, leading to
a totally different picture. Most
important for now is to concentrate
on coke yields. Figure 12 shows
coke yields for the three different
catalysts at both stripping times
and how a shorter time can have a
significant impact. The labels above
the bars indicate the extra amount
of soft coke that should be added
to the hard coke measured in the
ACE (or any test unit with long
stripping times). So basically
Catalysts B and C have large and

unrealistic benefits of the extensive

laboratory stripping times. As
many R&D centres apply too long
stripping times in their testing rigs,
they neglect differences in stripping
performance. High AAI catalysts
are subject to this penalty. They are
known to be superior in strippability, but do not receive the full
credit in testing. This is one other
element that contributes to the artefact in coke yields.

Metals activity influences hydrogen

and coke

One of the mechanisms that

contributes to the testing artifact is
the activity of metals like nickel
and vanadium in dehydrogenation
reactions. Through these reactions
excessive hydrogen and coke are
formed. Whether E-cats or D-cats

9
8

Coke, wt%

when dedicated and time consuming methods are applied with more
realistic deactivation conditions can
the AAI decline be simulated. One
such method we named CD Alfa
and includes metals such as Fe and
Ca.4 This method is very time
consuming and therefore seldom
embraced by R&D centres.

Cat A
Cat B
Cat C

1.9

1.3

Large s for
Cats B and C

0.5

6
5
4

99

Stripping time, seconds

Figure 12 Using more realistic stripping times in the bench scale tests can have a strong
impact on coke yields for different catalyst technologies

PTQ Q3 2015 125

Coke yield, wt%

10

FCC
typical
H2 yields

8
6
4
2
0

ACE unit E-cats tested

FCC unit
0

0.2

0.4

0.6

0.8

1.0

1.2

H2 yield, wt%
Figure 13 Excessive dehydrogenation reactions in an ACE unit enhance hydrogen and
coke compared to the FCC unit

are used, this mechanism occurs in

all types of testing equipment, so
both bench-scale uid beds (like
ACE) as well as in circulating pilot
plants. Only the magnitude is
different. The higher (dehydrogenation) activity of the metals is caused
by several factors. In case of E-cat,
the deposition of metals is realistic,
but because the coke is (typically)
burned off prior to testing in an
ACE-like unit, the oxidation state of
the metals is different from the FCC
unit. Moreover, the reaction times
are far longer than in the FCC unit.
Consequently, the metals have
much more time for dehydrogenation reactions and keep generating
hydrogen and coke. Compared to
an E-cat, a laboratory deactivated
D-cat is even worse. The dehydrogenation
activity
is
further
amplied as metals are highly
dispersed over the matrix surface
area which is usually too high
because it is not deactivated sufciently. Also, the deactivation in the
laboratory method has been too
short for nickel to be encapsulated.
Therefore, insufcient formation of
concentrated, inactive nickel aluminate species has taken place.
Hydrogen yields in an FCC unit
are typically below 0.1 wt% in VGO
applications, whereas in resid FCC
units hydrogen yields can reach
about 0.2 wt% due to elevated
levels in nickel and vanadium.
Albemarle carries out a routine
E-cat testing programme for its
customers, where activity and
selectivity of products are measured in the so called Fluid bed

126 PTQ Q3 2015

Simulation Test (FST). (This test is

quite similar to the ACE testing
apparatus.) Figure 13 illustrates that
hydrogen levels can reach as high
as 1 wt% for these E-cats in this
bench scale unit, being ve times
more than in the real FCC units.
This strong difference in reaction
mechanism illustrates how fundamentally different a testing rig is
from reality. Hand-in-hand with
those excessive dehydrogenation
reactions, coke yield is boosted.

In catalytic cracking,
many types of
reaction play a role
of which the primary
cracking reactions are
most important
These excessive dehydrogenation
reactions are most extreme in FST
and ACE units, but also occur in
pilot plants where we have seen
unrealistically high hydrogen and
corresponding
coke
yields.
Knowing this, some laboratories
attempt to match the FCC hydrogen yield by reducing the metal
loadings of the D-cat on purpose.
We can have animated discussions
about the correctness of this solution. Albemarle believes that by
reducing the metals loadings the
real cause is not solved and recommends to apply the true metal
levels of the E-cat to properly

reect the activity and stability of

the catalyst, instead of using a
reduced metals band-aid.
Albemarle has investigated coke
and hydrogen yields of many R&D
laboratories and found identical
trends for all of them. The relation
between coke and hydrogen yields
(wt% on feed) is very strong and
typically has a slope of 7, meaning
that 0.1 wt% additional hydrogen
corresponds to a 0.7 wt% increase
in coke yield.
This relationship has been
conrmed and acknowledged by
several laboratories in the oil industry. Interestingly a number of R&D
centres are using this relationship
to bring down the hydrogen yields
in the test to levels observed in the
FCC unit and likewise the coke
yields are corrected through that
same method.
In catalytic cracking, many types
of reaction play a role of which the
primary cracking reactions (catalytic and thermal) are most
important. Subsequent secondary
reactions are essential too, as they
determine
the
right
balance
between desired and unwanted end
products. The most often discussed
ones include hydrogen transfer,
dehydrogenation, cyclisation, and
isomerisation reactions. The relative
rate of each of these reactions and
how they contribute to the performance of the FCC unit is
important.
While FCC units started as bed
crackers, process technology developed to modern FCC riser
congurations with relatively short
contact time. With the change to
short contact time riser reactors and
quick
disengagement
sections,
secondary reactions are minimised
thereby reducing unwanted dry gas
and (delta) coke yields. Testing rigs
in the laboratory try to simulate
this as well, and circulating pilot
riser units succeed to some extent
by the simulation of simultaneous
ow of hydrocarbons and catalyst
through the plug ow riser reactor.
Bench scale congurations like ACE
and FST are popular equipment
thanks to their ease of operation
and high output, but fail in this
simulation. With pumping times on
the order of about 1 to 2 minutes,

www.eptq.com

31st
Oil | Gas | Fertilizers | Metallurgy | Industrial

Sulphur 2015

50%

DISCOUNT
FOR
OPERATORS

International Conference & Exhibition

9-12 November 2015, Sheraton Centre, Toronto, Canada

The worlds leading event for the sulphur & sulphuric acid markets
NETWORK WITH

600+

PROFESSIONALS

70+ EXHIBITORS
showcasing the latest
technologies and services

50% DISCOUNT
for operators

ONE DAY operator pass

FREE ACCESS
to the exhibition

Gold sponsors:

GET THE CRU VIEW

on market
developments in
sulphur and acid, plus a
special focus on China

SHARE OPERATIONAL
EXPERIENCE
& INDENTIFY
SOLUTIONS
with your peers

SET UP BUSINESS
MEETINGS in advance
using the CRU Sulphur
app & portal

30+

TECHNICAL PAPERS
covering sulphur &
sulphuric acid

PRACTICAL
WORKSHOPS

50+

INDUSTRY EXPERTS
presenting over
3 days

Silver sponsors:

Official publications:

FOR MORE INFORMATION AND TO REGISTER VISIT www.sulphurconference.com

Sulphur 2015 A4 Advert June 2015.indd 1

12/06/2015 10:56

Propylene, wt%

17.5

FCC unit
ACE E-cat
ACE D-cat

15.0

12.5

10.0

7.5
65

70

75

80

85

Conversion, wt%
Figure 14 Low hydrogen transfer reactions in ACE units lead to high propylene yields
compared to FCC units

a signicant part of the feed gets

cracked by catalyst which is already
deactivated by nitrogen compounds
and coke formation. The relative
rates of the different reactions are
thereby signicantly different from
reality.
Another aspect worth mentioning
is the hydrocarbon partial pressure.
While the FCC units typically operate at pressures in the range of 1-3
barg, laboratory units commonly
run
at
ambient
pressure.
Furthermore,
relatively
high
amounts of inerts like nitrogen are
used in this small scale equipment
to ensure proper dispersion of
(resid) feed and avoid coking problems in the riser or bench scales
units. Consequently, the relative
rates of the different reactions are
inuenced. Clear evidence is found
in studies of max propylene cata-

lyst systems. Hydrogen transfer

reactions are suppressed and lead
to too high propylene yields in lab
test equipment.
Figure 14 compares propylene
from an FCC unit with yields in the
ACE. When E-cat is used in the
test, possible defects in laboratory
deactivation
are
excluded.
Nevertheless propylene yields from
the ACE are substantially higher
than in the FCC unit.

Correction and translation tools

In light of many artifacts previously

discussed, Albemarle has developed two coke correction models to
adjust laboratory data to better
reect a catalysts performance in
an FCC unit.
It is well established and illustrated above that metal effects on
laboratory deactivated catalysts

10

Coke yield, wt%

9
8
7
6

Uncorrected H2
and coke yield

Cat No. 1
Cat No. 2
Cat No. 3

Corrected H2
and coke yield

4
0

0.1

0.2

0.3

0.4

Cat No. 4
Cat No. 5
Cat No. 6
0.5

0.6

H2 yield, wt%
Figure 15 Correcting hydrogen and coke yields by the Metals Coke Correction Model
brings ACE yields closer to FCC unit yields

128 PTQ Q3 2015

exaggerate both the hydrogen and

coke yields compared to those
observed in a FCC unit. Therefore
Albemarle developed a Metals Coke
Correction Model5 which adjusts the
laboratory coke using a stoichiometric relationship between laboratory
coke and hydrogen yield. This
model is effective in accounting for
metals-driven
coke
differences
between catalysts in the laboratory
and catalysts in the FCC unit.
An example of such a correction
model is illustrated in Figure 15. In
this example, six catalysts have
been tested in an ACE conguration. Hydrogen (and coke) yields
from the ACE unit are too high.
Through
the
Metals
Coke
Correction Model, hydrogen yields
are corrected to typical yields as
obtained in the FCC unit. Coke
yields are reduced through the stoichiometric relationship. As can be
seen, the ranking in hydrogen
yields between the catalysts is not
affected. And the excessive differences in coke between the six
catalysts are tempered somewhat.
When hydrogen and coke are
reduced through this method, the
mass balance is reduced as well
and needs to be closed. The remaining
yields
thus
require
normalisation to close this gap.
Over the past few years we have
discussed this approach with R&D
centres that showed interest. Some
have looked into this model and
tested it on their own data, while
others have developed and applied
their own methods based on the
same principle.
This method is a step forward
compared to using as-such-testingdata. It offers reners a tool to
generate more realistic yield
patterns. It needs to be said though
that the differences in coke yields
between these six catalysts are still
relatively large. In this example the
relative difference between the
highest and lowest corrected coke
yield is more than 40%. This
number is still relatively high and
not sufcient to be in line with
the experience that we have
reported.1 Albemarle has therefore
developed another tool for reners
which we consider a further
improvement.

www.eptq.com

Conclusions

For reners, the choice of an FCC

catalyst is a challenging task. Is the
incumbent catalyst the best option
for the FCC unit or are there better
catalyst technologies available to
increase the FCC margin? Catalyst
testing in a laboratory unit is an
option to screen the best available
technologies. While this approach

www.eptq.com

10
9

Coke yield, wt%

which we consider a further

improvement.
This second method employs a
Lumped Coke Correction Model,
which accounts for all types of coke
contributions, such as catalytic,
stripping, metals, non-distillable,
and feed coke. It requires lab-toFCC unit correction factors and
ideally the rener determines these
factors for its own FCC unit. To do
this, a rener needs to conduct
commercial trials with different
catalyst technologies, that is a low
versus a high Z/M ratio catalyst, in
the FCC unit. If a rener does not
have such data, Albemarle offers a
robust estimate of this correction
factor based on 19 FCC unit sideby-side comparisons. The model
has high exibility to tune the
correction factors and calculate
different scenarios.
We have applied this model to
the six catalysts in the example
above. The result of this modelling
work is shown in Figure 16. The
graph illustrates more realistic coke
yields than the as-such-testing-data.
Consider, when we compare the
catalysts with the highest versus
the lowest lab coke yield against
their performance in the FCC unit,
a difference in regenerator temperature of approximately 10F is
estimated; 10F is quite realistic,
supported by published experience.
Process model simulation of the
corrected coke yields in Figure 16
would result in about a 10F
change in regenerator temperature.
Albemarle emphasises that it
encourages reners to build their
own experience with different catalyst technologies in their FCC unit
and determine the correction
factors themselves. The different
catalyst technologies to be trialled
should preferably include high and
low Z/M catalysts.

8
7
6

Uncorrected H2
and coke yield

Corrected H2
and coke yield

Cat No. 1
Cat No. 2
Cat No. 3

4
3

0.1

0.2

0.3

0.4

Cat No. 4
Cat No. 5
Cat No. 6
0.5

0.6

H2 yield, wt%
Figure 16 Correcting coke yields by the Lumped Coke Correction Model turns ACE yields
into FCC unit yields

gives a lot of valuable information,

lab data used as-is almost always
favours catalysts with a high Z/M
ratio. On the other hand, low Z/M
catalysts are overwhelmingly found
in FCC units of reners that do not
test. In general, these catalysts
bring greater FCC margins through
their enhanced bottoms cracking
ability and are not subject to the
high delta coke measured in
the lab.
Albemarle concludes that no

No single accelerated
deactivation method
and test unit will
enable a tester
to achieve yields
equivalent to those
from the FCC unit
single
accelerated
deactivation
method and test unit will enable a
tester to achieve yields, especially
coke and slurry, equivalent to those
from the FCC unit. It is therefore
very important for testing reners
to conduct commercial trials and
develop the knowledge and experience to relate test results to
commercial performance by means
of a translation or correction
method, thus enhancing the value
of their laboratory work.

This article is adapted from a paper given at the

AFPM Annual Meeting, 22-24 March 2015.
References
1 Pouwels C, Bruno K, Better catalyst
evaluation strategies for maximizing FCCU
margins, AFPM Annual Meeting 17-19 Mar
2013, AM-13-05.
2 Gerritsen L A, Wijngaards H N J, Verwoert
J, OConnor P, Cyclic deactivation: a novel
technique to simulate the deactivation of FCC
catalyst in commercial units, Catalysis Today,
11, 1991, 61.
3 Bruno K, Hakuli A, Imhof P, Fletcher R P, FCC
catalyst selection in diffusion limited operating
regimes, NPRA AM-03-58, 2003.
4 Rainer D R, Rautiainen E, Imhof P, Novel labscale deactivation method for FCC catalyst:
inducing realistic accessibility responses to
iron poisoning, Applied Catalysis A: General 249
2003, 69-80.
5 Arriaga R, Bruno K, Improving FCC catalyst
selection by adjusting laboratory data. Part
1: The need for and concepts behind coke
correction methods, Catalyst Courier #84.

Ken Bruno is the Global FCC Technology

Manager, Albemarle Corporation, Houston,
whose team has worldwide responsibility for
catalyst, additive and technology applications,
product development, modelling and training.
He has 27 years of refining experience and
holds a BS in chemical engineering from the
University of Akron, a Masters and PhD in
chemical engineering from the University of
Notre Dame.
Carel Pouwels is a Global FCC Specialist for
Resid at Albemarle Corporation. Over the course
of his career, he has been primarily responsible
for the development and application of novel
FCC catalyst technologies. He holds a degree
in chemical engineering from Delft University
of Technology.

PTQ Q3 2015 129

44 TH TURBOMACHINERY & 31 ST PUMP SYMPOSIA

HOUSTON, TEXAS | SEPTEMBER 14 - 17 2015 | GEORGE R. BROWN CONVENTION CENTER

REGISTER

44TH TURBOMACHINERY & 31ST PUMP SYMPOSIA

The premier conference for turbomachinery and pump professionals.

DEVELOPED FOR THE INDUSTRY, BY THE INDUSTRY.

TPS.TAMU.EDU

FOR MORE INFORMATION

EMAIL: info@turbo-lab.tamu.edu
OR VISIT: TPS.TAMU.EDU

Implementing real time optimisation

technology
Modelling implementation of real time optimisation and its potential for
increased profitability in refining processes
HAMID GANJI, SAEID SHOKRI, MAHDI AHMADI MARVAST, SOROOD ZAHEDI ABGHARI and BEHROUZ NONAHAL
Research Institute of Petroleum Industry

n this article, implementing real

time optimisation (RTO) technology in the oil and gas industries
is categorised worldwide, its potential execution is investigated, and
the rate of profitability offered is
obtained as far as possible. The
processes that contained the highest
level of RTO implementation
included FCC, distillation, olefins
production, utilities, hydrocracking,
alkylation, catalytic conversion and
polymerisation. By studying the oil
and gas industries in Iran and
comparing them to similar foreign
industries, the potential for implementing
RTO
technology
is
investigated.
By using the rate of reported
profitability
for
the
selected
processes and considering the
capacity of domestic units, the rate
of profitability for these processes
is calculated. It was found that, by
implementing RTO technology, the
potential annual profit was around
$114 million for distillation, catalytic reforming, FCC, hydrotreating,
hydrocracking, gas refining, olefin
and sulphur recovery units in Iran.
Also it is evident that, from the
economic point of view, the order
of priority for implementing RTO
is: gas refining, distillation, FCC,
hydrotreating, olefin production,
catalytic reforming, hydrocracking,
and sulphur recovery.

Optimising production

In the last two decades, due to the

increasing cost of energy and raw
materials, tough environmental
measures, safety issues and worldwide competition, considerable
changes in optimising production
and quality in the chemical indus-

www.eptq.com

tries have taken place. These

changes have been made in the
design of processes and operating
conditions to reduce costs and meet
the requirements of existing process
limitations. The performance of a
process may be influenced by
factors such as atmospheric conditions and reduced catalyst activity.
Developing smart processes to
react appropriately to modifications
by eliminating human interference
as much as possible is receiving
careful attention. Smart processes
provide better operating conditions
and should receive more attention.

Advances in network
communications
technology and the
enhanced power of
computing has made
RTO state of the art
in many industries
RTO is one of the most widely
used technologies to optimise a
chemical process online. By using
RTO technology, the operating
conditions of processes for certain
periods are calculated and optimised. In fact, RTO systems control
the mechanisms of closed loop
processes to maintain automatically
the operating conditions of the
process nearest to optimised conditions. By applying RTO, continuous
assessment of the system and maintaining the operating conditions of
the process can be achieved.
Nowadays, advances in network

communications technology and

the enhanced power of computing
has made RTO state of the art in
many industries. RTO offers greater
opportunities to chemical engineers
and technologists to achieve their
processing objectives. In a RTO
system, a high performance control
system must be used as the infrastructure employing complex and
accurate models of the processes as
well as economic data to determine
the optimised operating conditions.
These control systems mainly
include control algorithms and
calculation platforms. RTO takes
place in the management layer of
the integrated production structure
of the system and acts as a bridge
between unit planning (medium
term planning) and the control
system (short term planning).
In the hierarchy of process
control, disturbances with high
frequencies may be controlled by
using process controllers at the RTO
level. Variations in optimised operating conditions are chased due to
changes at low frequencies. An RTO
system determines the permissible
values of variations and implements
them via control systems without
operator interference.
Optimisation can be carried out
at two different stages of a projects
lifespan. The first stage consists of
selection of the size and appropriate location of the equipment; the
second stage includes continuous
optimisation and online profitability control (operational/economic)
of the existing unit. RTO may be
implemented in a variety of
processes but has higher potential
in the oil and gas industries
because of the complexity of their

PTQ Q3 2015 131

Rate of implementation of RTO in

the worlds oil, gas and petrochemical
industries1,2

Comparison of profitability in various

units following RTO implementation2, 3
Processing units

Rate of
profitability
FCC
7-15
Distillation
3-5
Catalytic reforming 10-40
Hydrocracking
10-20
Hydrotreating
10-20
Isomerisation
10-50
Gas refining
7-12
Sulphur recovery
35
Olefin production
4

Unit

Number of units
(RTO implemented)
FCC
34
Distillation
25
Ethylene
17
Water, electricity & steam
7
Refinery complex
6
Hydrocracking
5
Alkylation
5
Gas complexes
4
Catalytic conversion
4
Polymerisation
3
Hydrogen
2
Hydrogen purification
2
BTX
2
RFCC
2
NGL
2
Gravity reduction
1
Sulphur recovery
1
Boiler
1

the world was carried out. The

statistical reports included the
profit incurred by implementing
RTO technology and the software
used to facilitate this.
Reports indicate that 133 cases of
RTO technology have been implemented in the worlds oil and gas
industries. Based on this information, the units that gained most by
implementing RTO technology
were considered to be the best
potential candidates (see Table 1).
Using the information gathered,
the rate of profitability through
implementation of RTO technology
in the processes listed in Table 1
was categorised. In order to
compare profitabilities, it was
decided to compare the rate of
profitability with respect to inlet
feed or finished products of the
units separately. Table 2 compares
the rate of profitability of different
processes by unit feed or final

processes and the number of variables that influence decision making.

For this reason, different projects
have been carried out worldwide
that have resulted in high profitability and fast return on capital
deployed. This article investigates
the implementation of RTO as well
as calculating the rate of profitability of selected units in the oil and
gas industries.

RTO technology worldwide

By studying databases, documents

a
and
available
references,1,2
wide-ranging
investigation
of
implementation of RTO throughout
40000

Profitability, $1000s/year

35000
30000
25000
20000
15000
10000
5000

at

er
y
in
re
f

as

O
le
fin
G

U
SR

yd

ro
t

re
at
in
g

C
FC

rm
in
yd
g
ro
cr
ac
ki
ng
H

al
y

tic

is

re
fo

til

la
tio

/barrel
/barrel
/barrel
/barrel
/barrel
/barrel
$/MMSCF
/ton
$/ton

Table 2

Table 1

Plants

Figure 1 Rate of profitability with RTO implementation in selected processes

132 PTQ Q3 2015

Units

products. As can be seen, the rate

of profitability through implementation of RTO technology for
catalytic reforming, crude distillation, olefin production, FCC, gas
refining, isomerisation and sulphur
recovery is $0.03-0.50 per barrel of
feed.

Implementation of RTO

With regards to the rate of profitability obtained by implementing

RTO
technology
within
the
processes shown in Table 2 and
with respect to the capacity of
domestic units, the rate of increased
profitability was calculated. As
Table 2 shows, these units are for
refining, petrochemical and gas
processes. Figure 1 compares the

RTO technology
yields higher
profitability for an
oil refinery. This is
because oil refineries
have more added
values in their
products
rate of profitability for the listed
units. As Figure 1 shows, for units
where RTO technology is implemented, gas refining enjoys most of
the profitability, followed by distillation units in oil refineries. It
should be noted that some of the
processing units do not exist in all
nine of Irans oil refineries, for
instance
catalytic
conversion,
hydrogen purification and FCC
units only exist in two oil refineries.
Looking at the relative rate of profitability for various units, a
comparison should be made for a
typical unit with a mean value of
feed. For example, a distillation
unit with a capacity of 201 667 b/d
and a catalytic conversion unit with
a capacity of 20 444 b/d were
considered. By using this technique,
the rate of profitability was
obtained and is shown in Figure 2.
From an economic point of view,

www.eptq.com

Conclusions

The application of RTO technology

in oil refining was investigated in
this research. First, the RTO technologies applied in oil and gas
industries worldwide were investigated and classified to find out
which chemical processes featured
the highest levels of RTO implementation. It became clear that
most processes suited to RTO
implementation included FCC,
distillation, olefin production, utilities, hydrocracking, alkylation,
catalytic reforming, and polymerisation respectively. With the aid of
data provided through available
sources of information, the rate of
profitability due to implementation
of RTO technology in these
processes was calculated. With
further investigation of processing
units in Iran with regard to their
capacities, the rate of profitability
of RTO implementation was
obtained. The calculations showed
RTO implementation in the selected

www.eptq.com

4500
4000
3500
3000
2500
2000
1500
1000
500

ry
re
fin
e

le
fin
O

G
as

SR
U

FC
C
yd
ro
tre
at
in
g
H

at
al

is
til
la
tio
yt
n
ic
re
fo
rm
H
in
yd
g
ro
cr
ac
ki
ng

Plants

Since the rate of profitability with

RTO technology is not reported for
integrated processing units in an oil
refinery in the resources available,
it was decided to compare the total
profit of RTO implementation in all
units of an oil refinery with that of
a gas refinery. It was assumed that
an oil refinerys processes consisted
of distillation, catalytic reforming,
hydrocracking, hydrotreating and
FCC. Figure 3 compares the profitability of RTO implementation in oil
and gas refineries. It can be seen
that RTO technology yields higher
profitability for an oil refinery. This
is because oil refineries have more
added values in their products and,
by implementing RTO technology,
the profitability rises accordingly.
Also, oil refineries have units such
as catalytic reforming and hydrocracking units which produce final
products with high added values.

Profitability, $1000s/year

Case study

5000

Figure 2 Comparison of profitability for various processes with medium capacity

70000

Profitability, $1000/year

the priority for implementation of

RTO
technology
in
selected
processes is: gas refining, distillation, FCC, hydrotreating, olefin
production, catalytic reforming,
hydrocracking
and
sulphur
recovery.

60000
50000
40000
30000
20000
10000
0

Oil

Gas

Refinery
Figure 3 Comparison of profitability through RTO implementation in oil and gas refineries

units led to approximately $114

million of annual profit.

PhD in process engineering from Iran University

of Science and Technology (IUST), Tehran, Iran.

References
1 Klein S, Online Process Economics Optimizers,
PEP Review, 2003.
2 Databank of Apc-network Company www.
apc-network.com
3 Oil & Gas journal, News and Technology, 21
April 2003.

Mahdi Ahmadi Marvast is a Project Manager

with RIPI, a subsidiary of NIOC (National
Iranian Oil Company). With more than 15
years of research and development experience
in developing such technologies as GHT, GTL,
SRU, RTO and LDS, he holds a BSc in chemical
engineering from the Petroleum University and
MSc and PhD in the same field from Amirkabir
University of Technology in Iran.

Hamid Ganji is Director of the Process

and Control Development Group and an
Assistant Professor at the Research Institute
of Petroleum Industry (RIPI), Tehran, Iran. He
has 15 years experience in process modelling,
simulation, optimisation, automation and
control of chemical processes and holds a PhD
in chemical engineering from Tarbiat Modares
University.
Saeid Shokri is a Project Manager with RIPI
with interests in optimisation, modelling and
simulation of chemical processes. He holds a

Sorood Zahedi Abghari is a Project Manager

and Assistant Professor with RIPI, specialising
in product scheduling, optimisation, process
modelling, simulation and optimisation. He
holds a PhD in chemical engineering.
Behrouz Nonahal is a Research Engineer with
RIPI, specialising in real time optimisation
of chemical processes, and has presented
seminars on hydrodesulphurisation and gas
sweetening technologies. He holds a MSc from
the University of Greenwich, London, UK.

PTQ Q3 2015 133

MOSCOW REFINING &

PETROCHEMICALS WEEK 2015
10% DISCOUNT
available with promo code

PTQ2015MW
Send the promo code, your
name, job title and company
name to:
marketing@europetro.com

21-25 September 2015, Lotte Hotel, Moscow

Early Confirmed Sponsors:

The Week Incorporates:

Op-Ex Russia & CIS - Operational Excellence in Refining,

Gas & Petrochemicals
21 & 22 September

RPTC - Russia & CIS Petrochemicals Technology

Conference & Exhibition
22 & 23 September

RRTC - Russia & CIS Refining Technology Conference

& Exhibition
24 & 25 September
Please contact us for more
information:

Moscow Refining and Petrochemicals Week offers:

Op-Ex Russia & CIS 2015

Marina Marchovska
marina_marchovska@europetro.com
+359 2 427 22 05

5 days of intensive networking and up to date presentations

RPTC & RRTC

Maria Zaytseva
maria_zaytseva@europetro.com
+ 7 (495) 517 77 09

Over 30 exhibition stands

Over 400 delegates

Opportunity to hear more than 70 reports from industry experts
Interactive panel sessions
Exclusive marketing opportunities

Register at www.MoscowWeek2015.com before 21.08.2015 to benefit from the early bird rate

Technology in Action

Figure 1 Butane blending at Milwaukee refinery

www.eptq.com

Ju 12
l2
Au 012
g
2
1
Se 012
p
20
1
1
O
ct 2
20
1
1
N
ov 2
20
1
1
D
ec 2
2
1
0
Ja 12
n
2
1
Fe 013
b
2
1
M 013
ar
20
13
1

31 Mar
1 Mar
2013

12

28 Feb
1 Feb
2013

20

31 Jan
1 Jan
2013

31 Dec
1 Dec
2012

Ju

30 Nov
1 Nov
2012

31 Oct
1 Oct
2012

01

100

20

200

r2

300

ay

400

Ap

500

600

1200
1100
1000
900
800
700
600
500
400
300
200
100
0

Butane volume, gal1000

Total butane, gal1000

700

Many countries legally mandate the observation of the

maximum gasoline vapour pressure to prevent air
pollution and ensure the safe transport and storage of
petroleum. In addition, gasoline vapour pressure is an
important factor in preventing fuel vapour lock and in
providing easy motor starting in cold weather or at
high altitudes.
The regulation of maximum vapour pressure limits
requires that refiners carefully blend gasoline. That
blending affords refiners opportunity for incremental
profits, especially when low-cost butane is available to
blend into the more expensive gasoline. Butane has
good anti-engine knock qualities but is a major overall
contributor to gasoline vapour pressure, such that the
maximum amount of butane blended into gasoline is
limited by the maximum allowed vapour pressure of
the fuel.
The spread between the price of gasoline and butane
is a key factor in blending profitability. For example, if
the market price for gasoline is $3 per gallon and the
price for butane $1.5 per gallon, every gallon of butane
used to replace a gallon of gasoline in a blend yields
$1.5 per gallon in extra profits.
Typically, between 2% and 3% butane by volume can
be blended into gasoline, depending on the vapour
pressure of gasoline before blending. The most profitable time for blending butane into gasoline is during
the transition between the summer and winter driving
seasons, when sometimes up to 10% of butane by
volume can be blended into gasoline. During colder
seasons, higher vapour pressure is acceptable.
Two examples highlight how much profit can be
achieved. During a six-month period, (October 2012
March 2013) one American refinery in Milwaukee,
Wisconsin, was able to blend an average of 4.8%

butane, or 2.6 million gallons, into its gasoline. With an

average price spread between butane and gasoline of
$1.5 per gallon, the refinery was able to achieve an
extra $4 million in profit (see Figure 1).
Similarly, a refinery located in Green Bay, Wisconsin,
blended 4.3 million gallons of butane into gasoline
between April 2012 to March 2013, with most blending
occurring between September and March. At a gasoline-butane spread of $1.5 per gallon, the result is an
extra $6.5 million of profit (see Figure 2).
Given the potential level of profit, it is easy to understand that a 10 000 b/d refinery that installs a blending
facility worth $1 million can expect a return on its
investment within a matter of weeks rather than years.
Both the Milwaukee and Green Bay refineries have
installed a Technics blending facility that employs two
Grabner Minivap On-Line vapour pressure process
analysers and duplex sampling loops. The blending
system is designed to deliver product on-spec, on-time
and homogenised to a precise blend that is within 0.2
psi (1.3 kPa) directly on the pipeline. Extremely tight
dosing to within 0.05% of butane is achievable.
Apart from exact butane dosing, high-accuracy
vapour pressure testing contributes significantly to
profitability by allowing refiners to use a narrower
margin of safety and still guarantee that the gasoline
vapour pressure is within limits. As a result, more
butane can be added and greater profits generated.
Grabner analysers are known for high accuracy in
standard compliant vapour pressure testing. The
Minivap On-Line RVP tester used in the Technic
system features the triple expansion method with a
reproducibility of 0.1 psi (0.7 kPa), compared to the
ASTM D5191 standards reproducibility of 0.4 psi
(2.75 kPa).
Better precision leads directly to significantly more
blending profit, even at very low butane blending
ratio. If the vapour pressure of gasoline increases by

Optimising profits by blending butane

Figure 2 Butane blending at Green Bay refinery

PTQ Q3 2015 135

1 psi (7 kPa) through blending, the Grabner analyser

allows up to 0.5% more butane to be blended into
gasoline, compared with the ASTM D5191 standard
method.
As a consequence, a medium-sized refinery may save
more than a million dollars per year and recoup the
analysers costs almost immediately.
For more information: info.grabner-instruments@ametek.at

Measuring metal mobility in FCC

equilibrium catalysts
In recent years, a number of new techniques have been
used to study metal mobility and metal effects in FCC
units. These techniques brought new insights into how
deleterious metals do their work in the FCC and have
created solutions to counter these effects. One of the
recent, more surprising discoveries was that iron from
crude, deposited on the FCC catalyst surface, can move
from particle to particle. Vitrification of the catalyst
surface was extensively documented in the mid 1990s
and though the vitrification mechanism was known
and widely accepted, one of its effects has been largely
ignored: when a catalyst surface is in flux due to
temperature and contaminant metals, transfer of
contaminant iron oxide crystallites is possible when
FCC catalyst particles collide in the regenerator.
The study of this is best carried out by a survey of
catalyst particles from a unit suffering from iron
poisoning. For every particle, the concentration of
contaminant metals on the particle can be analysed,
and the distribution over the particles can be derived
from this examination. As catalyst is daily added and
removed from the FCC unit, it will exhibit a residence
time distribution (exponential decay, the highest fraction of the particles is the newest, and only very few
very old particles are present). If a contaminant metal
cannot move from particle to particle, the metal
concentration will show its age and an exponential
decay can be observed. For metals that move from
particle to particle in the FCC unit, a bell curve type of
distribution will be observed.
The new analytical methods, such as imaging and
chemical analysis by scanning electron microscope, are
straightforward to use, though their proper application
requires some attention. What can go wrong? One
thing is not using the average concentration of contaminant metal on the FCC particle, but using

Figure 1 Cat-Aid is an iron-trapping FCC additive: left, catalyst

particles without Cat-Aid; right, with Cat-Aid

136 PTQ Q3 2015

concentrations at arbitrarily chosen spots on a very

limited amount of particles.
Imagine the following example. During a flu
pandemic, two doctors were sent out to study whether
the flu is contagious. One doctor rounded up four
people from within an area and took 250 spot temperatures from each person. He correlated the spot
temperatures with the location where he had taken the
samples: mouth, ear, every finger, every toe, and so on.
His analysis showed the flu cannot transfer from
person to person and he could prove this with 1000
temperatures (collected from four people).
The other doctor took the oral temperatures of 1000
people within one area. He correlated the results via
distribution of people within the area and concluded
that the flu is contagious.
Every study of metal mobility on iron contaminated
FCC catalyst has shown that iron moves around. In
hindsight this is not a complete surprise knowing what
has been discovered in the previous century: contaminant iron exhibits its destructive effects by vitrification
of the FCC catalyst surface. Under FCC conditions this
glass phase is not completely solid, as the dense outer
surface of the catalyst particle testifies. Sticky surfaces
can transfer matter upon collision and the distribution
of iron on FCC E-cat is testament of this behaviour.
Cat-Aid, an effective iron trapping FCC additive that
protects the base catalyst from well-known adverse
effects, has been proven to alleviate the inter-particle
migration of contaminants (see Figure 1).

For more information: Bart.deGraaf@matthey.com

Safe conveyance of sour crude oil

Sour crude oil contains a high amount of impure
sulphur. To convert sour crude oil into valuable and
refined fuel, the impurities occurring as elemental
sulphur in solution and as hydrogen sulphide gas need
to be removed. Due to the presence of hydrogen
sulphide, sour crude oil is toxic, hazardous and
extremely corrosive which makes its transportation to
the preparation system challenging.
Critical applications of conveying mixtures of acidic
hydrocarbon/water to the washing system have
always been a challenge for any pump manufacturer.
This was also the case in one recent project Lewa
handled in the Emirate of Abu Dhabi. Lewa has
designed and supplied 16 pumps to a major construction company in Abu Dhabi specifically designed to
handle this critical application (see Figure 1).
On the project, Lewa was asked to provide a solution
to pump liquid hydrocarbons from the flare knockout
drum or the blow-down drum respectively and convey
the acidic hydrocarbon/water mixture back into the oil
line to the washing system. Since the differences in
these processes are only minor, it was possible to
supply 16 mechanically identical pumps of the Lewa
Ecoflow series, driven by 55 kW electric motors,
measuring 2.7 x 1.9 m, 1.9 m high, and weighing
4200 kg.

www.eptq.com

ASIA

DOWNSTREAM
WEEK
13-16 October 2015, Bangkok
Co-Hosted by:

Organised by:

ASIA-TECH
Asia Downstream
Technology
& Safety
13 & 14 OCTOBER 2015
BANGKOK

Operational Excellence

IN OIL, GAS & PETROCHEMICALS

OpEx

Asia

15 & 16 OCTOBER 2015

BANGKOK

THE INDUSTRY STANDARD FOR EXCELLENCE

Asian Rening and Petrochemicals Perspectives
Pertamina's Reneries Upgrading Projects
Latest Hydroprocessing Technologies
Polyethylene & Polypropylene Technologies
Rening & Petrochemical Integration Benets
Process Safety Management for Enabled Workforce
Challenges to PSM Implementation at PTTES
PLUS: Axens' Asia Seminar, 12 October

CREATING COMPETITIVE ADVANTAGE

Implementation of Operational Excellence in PTTGC
Petronas' OE Agenda: Our Game Plan & Strategies
The Importance of Culture in OE
How Centres of Excellence can Drive
Implementation of OE
Meeting Global Challenges With Global
Collaboration
Integrating Your Operations Team into Projects

Sponsored by:

www.ADW2015.com

High-purity ethane from wet NGL

Figure 1 In all, 16 mechanically identical Ecoflow pumps were

delivered by Lewa. They work under difficult process conditions to
convey a highly volatile mixture of hydrocarbons and hydrogen
sulphide, conveying water back into the oil line to the
washing system

To ensure high reliability in operating the system,

despite the projects difcult conditions, the customer
dened a series of strict selection criteria. The critical
uid characteristics that had been taken into consideration while designing the pumps are: high content of
sulphur of about 5%, high chlorine content, high
temperatures and highly humid conditions under
which even stainless steel corrodes. Since the medium
conveyed is very volatile and extremely hazardous,
Lewa selected hermetically tight diaphragm pump
heads made of superduplex stainless steel. The
diaphragm itself, due to the eroding effect of the particles in the corrosive liquid, is made of wear-resistant
PTFE.

Special suction valves for low NPSHr value

While designing the pumps, it was also necessary to
take into consideration that the mixing ratio of water
to hydrocarbons in the liquid phase varies signicantly
and that the pressure can vary between 20 and 45 bar.
Lewa designed the pumps to handle a ow rate
between 0.5 and about 15 m/h of the acidic mixture.
To ensure that the pumps work reliably despite the
changing ow rate, the net positive suction head of the
system (NPSHa) also had to be kept higher than that
required at the pump suction (NPSHr). When pure
hydrocarbons are conveyed on hot summer days, their
vapour pressure rises and the risk of cavitation
increases. Various options were analysed with the
customer to tackle this issue. Lewa offered integrated
special suction valves to keep the suction pressure
requirements low and satisfy all process conditions.
According to the customer, since commissioning, the
installed Lewa pumps have been running smoothly
and with high reliability, fullling the expectation of
the maintenance and operating personnel. The on-call
time agreement that permits uninterrupted work at the
plant even when one of the pumps needs maintenance
is also working seamlessly.
For more information: lewa@lewa.de

138 PTQ Q3 2015

Producing high-purity ethane (95+ LV% ethane) in an

NGL fractionation plant requires a refrigerated
condenser for the deethaniser. If the NGL feed is wet or
saturated with water, hydrates can form in the top
section of the deethaniser and in the deethaniser
condenser where the operating temperature may be
below the hydrate formation temperature. Formation of
hydrates can plug the equipment and piping to hinder
the operation. The conventional method of preventing
hydrate formation is to use dry desiccant beds or a
glycol contactor to remove water content in the NGL
feed. However, both methods utilise a complicated
dehydration unit that not only requires additional
capex and opex but also additional space. To overcome
these shortcomings, Wood Group Mustang has developed a novel high-pressure deethaniser that produces
95+ LV% ethane from wet NGL without the need for a
dehydration unit. This commercial-scale plant is
currently in operation and has been running
successfully.
Typically, molecular sieve or silica gel beds are used
to dehydrate light liquid hydrocarbons such as LPG or
NGL. As shown in Figure 1, a dry desiccant system
consists of adsorption and desorption (regeneration)
beds with a complicated switching valve arrangement.
The regeneration system consists of a regen gas heater,
regen gas cooler, regen gas compressor, and other
miscellaneous items. A dry desiccant dehydration
system is a batch-operated system. The periodic switching among adsorption, heating and cooling cycles is
complicated although the switching operations can be
manipulated by an automatic timer control.
The operation of a glycol dehydration unit is simpler
than that of a dry desiccant unit. However, a glycol unit

C2 and lighter
to condenser
Molecular sieve dehydration unit

Dehydrated
NGL

Reflux

Molecular sieves
No. 1
No. 2
adsorbing
regenerating

Regen gas from

cooling recycle

Deethaniser

Hot regen gas

Filter
Reboiler
return

Cold regen gas

Regen gas heater,

cooler and
compressor

C3 and heavier
to reboiler

Water

Regen gas
from heating
cycle

Wet NGL
Feed

Drain

Figure 1 Deethaniser with molecular sieve dehydration

www.eptq.com

C2 and lighter
to condenser

C2 and lighter
to condenser

Glycol
cooler

Reflux

Lean glycol

Glycol
contactor
Dry gas

Wet gas

Rich glycol

Deethaniser
Water

Wet NGL
feed

Inlet
scrubber

Liquids

Wet NGL
feed

Reflux

Glycol
regeneration

Deethaniser

Reboiler
return

Reboiler
return
C3 and heavier
to reboiler

C3 and heavier
to reboiler

Figure 2 Deethaniser with glycol dehydration

is commonly used for gas dehydration (instead of

liquid dehydration) due to the small amount of glycol
used in the contactor. In order to use a glycol unit, a
vapour stream needs to be drawn from and returned to
the deethaniser after it is dehydrated. This means the
deethaniser column requires extra length for the internal vessel heads that are necessary for side vapour/
liquid draws and returns (see Figure 2). In addition, a
dehydrated gas compressor or a side draw liquid return
pump may be needed in order to compensate for the
pressure drops in the draw piping, return piping and
the glycol contactor. A glycol dehydration unit consists
of a glycol contactor and a glycol regeneration system.
The glycol regeneration system usually includes a minimum of a ash drum, a glycol reboiler/still column, a
glycol/glycol exchanger, a high-head lean glycol pump,
and a glycol cooler. Furthermore, injecting stripping gas
to help regeneration is necessary in some cases.
As the pressure increases, the temperature of the
deethaniser condenser increases rapidly, while the
hydrate formation temperature increases slowly. Based
on this phenomenon, high-purity ethane can be
obtained by increasing the deethaniser pressure without
dehydration (see Figure 3). For example, the temperature of the deethaniser overhead condenser is about
46F at 390 psig and 63F at 490 psig to produce ethane
at 95+ LV%, while the hydrate point of this high-purity
ethane is about 51F at 390 psig and 61F at 490 psig.
Comparing the operating temperature and the hydrate
point, hydrates form in the condenser at 390 psig but
do not form at 490 psig if the NGL is not dehydrated.
Although the operating pressure of a deethaniser is
typically 200-400 psig, which is well below the critical
pressure of the hydrocarbon mixture, Wood Group
Mustang has designed a deethaniser that operates
successfully near 500 psig at the bottom of the
deethaniser to produce 95+ LV% ethane from wet NGL
without dehydration.
The cost of a high-pressure deethaniser column is
higher than that of a conventional deethaniser column,

www.eptq.com

Figure 3 Deethaniser without dehydration

but the savings from not needing a dehydration system

and from the reduction in horsepower of the refrigerant
compressor due to a higher condenser temperature can
compensate for this increased capex cost. The increased
requirement of the deethaniser reboiler duty caused by
a higher pressure in the deethaniser column can be
offset by the reduction in depropaniser reboiler duty
due to a higher outlet temperature of the deethaniser
bottoms. Therefore, the capex and opex of the
high-pressure deethaniser design without dehydration
are not signicantly different from those of a conventional deethaniser design with dehydration.
In conclusion, a novel deethaniser process without
dehydration has been developed and proven a successful alternative in commercial operation. This process
eliminates the need of installing a complicated molecular sieve or glycol hydration unit, providing simplicity
for both installation and operation.

For more information: chyuan.chen@woodgroup.com

Reducing refinery boiler emissions

A 185 000 b/d US East Coast renery was primarily
processing low sulphur, light crude oil from Canada,
the North Sea and West Africa to be used as transportation fuels, heating oil and liqueed petroleum gas. Due
to the EPAs Petroleum Renery Initiative, the facility
agreed to install NOx reduction equipment (selective
catalytic reduction SCR) on its combustion units as
well as ultra-low NOx burners with the goal of achieving 0.040 lb/million BTU or lower NOx. On the SO2
emissions side, the renery operator had its boiler fuel
oil usage restricted to less than 900 b/d on a 365-day
rolling average, with plans in place to eventually eliminate all use of fuel oil.
To meet renery steam requirements while complying
with tougher emissions standards, the operator
replaced older boilers with two more efcient units
from Rentech Boiler Systems. These custom units were
D-Type (accelerated natural circulation of water) boilers,
with water-cooled membranes, each delivering

PTQ Q3 2015 139

250 000 lb/hr of superheated steam at 575 psig and

730F. The burners can use natural gas and refinery gas.
Emissions control systems bring NOx emissions down
to 2 ppm, CO to 20 ppm and NH3 slip to 5 ppm. Either
boiler can produce enough steam to run the entire
plant, with the other operating as a hot standby. Each
has dual fans as well as monitoring and safety systems.
Both Rentech boilers were shop assembled water-tube
package boilers with separately packaged economisers.
They were designed with complete membrane wall
construction of the furnace, including the front wall.
The walls were constructed of 2.0in OD x 0.135in MW
tubes on 4in centres connected by 1/4in x 1in carbon
steel membranes to form a water-cooled enclosure. The
furnace is 12.5ft W x 10.75ft H x 44.0ft L, providing a
total volume of 5913 ft3. Because the structure is water
cooled, it avoids the usual maintenance issues associated with using refractory and firebrick, and also
minimises startup time.
As the furnace can run at a lower temperature, NOx
formation is reduced. In the rear of the convection
section, fins are added to the final rows of carbon steel
tubes to increase boiler efficiency. Longitudinal, vertical
baffles are employed there to eliminate acoustic vibration problems. The steam drum is provided with a
primary belly pan, chevrons and cyclones to ensure that
moisture carryover does not exceed 0.03%. The chevron
separator is designed to keep operating even if the
water level is as much as 2in above the high-high water
level, so that water carryover is still reduced when the
water level rises. There are two dedicated, unheated
downcomers at either end of the steam drum, to help
stabilise the drum level during rapid changes in the
firing rate. The downcomers also support the steam
drum and reduce the stress on the convection tubes.
The horizontal-tube drainable, convection-type superheater is located behind the division wall and the
screen tubes to protect it from direct radiation. The
single-stage unit without attemperation is designed to
provide a constant steam temperature of 740F
+10/-30F, with temperatures varying based on the
type of fuel, feedwater temperature, continuous blowdown rate and changes in air temperature and relative
humidity. Six superheater thermocouples are used to
monitor tube wall temperatures.
Each boiler comes with two Arrangement #7 (wheel
supported between bearings), 1200 rpm forced draft
fans from Robinson Industries. The fans are driven by a
Dresser-Rand steam turbine with a Lufkin Industries
gear. The turbines are controlled by a Woodward electro
pneumatic governor with a Flowserve pneumatic actuator. Induced flue gas recirculation (FGR) is used to
reduce NOx emissions and a vertical gas flow economiser captures waste heat. The economiser tubes are
horizontal and fully drainable with five 0.75in x 0.060
W solid fins per inch.
Each furnace has a single John Zink low NOx burner
that produces 20 ppm NOx, 80n ppm CO and 8 ppm
(0.0032 lb/mmBTU) VOCs. To bring these levels down
further, the exhaust passes through a catalyst that
reduces CO from to 20 ppm and VOCs to 7.5 ppm. The

140 PTQ Q3 2015

flue gas then passes through an SCR that takes NOx

down to 2 ppm
In addition, Rentech provided the 100ft exhaust stack,
lubrication system, silencers, and most of the ductwork,
insulation and lagging. The refinery operator supplied
the motor controls and starter, the ammonia storage
tank, the continuous emissions monitoring system
(CEMS) and the O2 analyser.

For more information: hkumpula@rentechboilers.com

Continuously improving machinery

performance
Through regular inspection and selective or continuous
monitoring, traditional maintenance measures are
intended to uncover damaged valves or signs of wear
in rings and packings. Maintenance strategies such as
these often disregard the fact that process conditions,
loads, operating times and the like have changed, and
that reciprocating compressors are subject to changing
stresses. The REE service is a methodology for analysing improvement potential.

Audit for reciprocating compressors highlights

improvement potential
Operators of process equipment today expect broadbased suggestions on how to optimise their plants. The
servicing of compressors in many instances the core
component of a plant is no exception. Processes
change during the life cycle of these machines, necessitating adjustments.
Through regular inspection and selective or continuous monitoring, traditional maintenance measures are
intended to uncover clogged valves, signs of wear in
seals and bearings, or lubrication problems.
Maintenance strategies such as these often do not
consider that process conditions, loads, or operating
times have changed, and that the reciprocating
compressor is subject to varying stresses not to
mention using maintenance as a way to optimise the
machine to the new requirements.
Compressor specialist Hoerbiger intends to change
this. In October 2012 the company created the service
offering known as REE: reliability, efficiency, environmental soundness (see Figure 1). Based on the premise
that reciprocating compressors are usually critical
process components, so that the productivity and profitability of an entire process plant hinges on their
reliability and efficiency, the Austrian compressor
equipment manufacturer developed a proprietary methodology for analysing improvement potential.
Systematically uncovering improvement potential
The systematic approach has already proven itself in
practice. A refinery operator in the Middle East, for
example, had 40 reciprocating compressors analysed,
resulting in the identification of energy savings potentials of as much as 10% of total power. In detail, the
proposal was to improve the operating concept in

www.eptq.com

Reliability

Environment

10
9
8
7
6
5
4
3
2
1
0

Industry best
Enhanced
configuration

Value added
improvement
Current
configuration
(x,y,z)

Out of operation

Efficiency

Figure 1 The REE Index is calculated from a combination of three

factors: reliability, efficiency, and environmental soundness

general, and specifically compressor control. Following

an audit of a fertilizer manufacturer, improvements in
lubrication and different cylinder liners were suggested,
resulting in less debris and oil contaminating the reactor, and hence extended service life for the catalyst.
During a REE audit, meetings are held to determine
the DNA of a compressor. The DNA of a compressor
has several aspects the best DNA is the optimum
combination of process, compressor, components, maintenance strategy, and mode of operation. On site,
compressor specialists learn about the design, operating
conditions, and condition of the compressors used in
the process. A standardised report summarises the evaluation. The report highlights improvement potential,
and an index identifies factors that influence reliability,
efficiency, and environmental compatibility on a range
from zero (out of operation) to 10 (presently the best
configuration in the industry). The Index Tool calculates
the potential for improvement.
It is not impossible for the goals of an optimisation
process to conflict with each other: if availability is all
that is important to an operator, this may come at the
expense of efficiency. If efficiency is what is maximised,
wear may increase. The goal of the benchmark is to
highlight and optimise the effects in all three respects,
taking the customers requests into consideration.
Optimum solutions may even guarantee increased efficiency and higher reliability at the same time.

Operator directly defines optimisation goals

The plant operator ultimately personally evaluates and
decides which machines and improvements will be
tackled. This decision then serves as the basis for an
engineering study which outlines specific solutions and
their return on investment. Recommendations range
from performance upgrades, through adapting the
compressor design to new process conditions, to optimising the entire compressor fleet.
Tools that the specialists have at their disposal
include software programs for compressor simulation,
and methods for checking the impact of gas pulsations
and torsional vibration. In its mission, Hoerbiger draws
on data gained over decades of development work,
such as information about flow conditions in cylinders
and through valves, and the distribution of forces and

www.eptq.com

tribological effects on rings and packings. The database

captures the wealth of knowledge gained from more
than 55 000 installed compressors.
Even decades-old compressor installations actually
never need new machines; instead, a custom upgrade
can deliver a considerable increase in reliability and
efficiency. This is important, since replacing compressors while plants are operating is not just about
investment costs, but also brings issues of safety and
time. An upgrade can generally be completed much
more quickly than a replacement. REE is designed to
boost the compressors added value to the operators
business model, while establishing a continuous
improvement process. This is the only way to keep
compressors up-to-date, regardless of maintenance
measures.
For more information: nikolaus.lubega@hoerbiger.com

Managing uncertainties and

increasing profits
In response to the rare-earth supply crisis in 2010, and
the tight oil revolution in the United States, Grace
commercialised eight new low- or no-rare earth FCC
catalysts and additives. The first in the series, Achieve
400 FCC catalyst, addresses the octane debits that were
being encountered by North American refiners. The
technology has delivered an economic uplift on the
order of $3-7M/yr in multiple commercial applications
due to the boost in slurry conversion, FCC naphtha
octane, and butylene yield.
Five key functionalities were developed in the development of the new catalyst series, to provide yield
flexibility for refiners:
Increase distillate yields with high diffusivity
matrices
Reduce dry gas with advanced metals traps
Drive conversion with ultra-high activity zeolites
Maximise resid processing with leading coke
selectivity
Boost refinery octane with dual-zeolite technology.
The following examples illustrate how Achieve catalysts have been applied commercially to improve FCC
unit yield and profitability.

Maximising conversion
When processing lighter feedstocks, insufficient catalytic activity requires that the catalyst circulation rate
increases so that conversion, and thus the coke yield
from the catalyst, increases to satisfy the FCC heat
balance. If the FCC unit cannot physically circulate
enough catalyst, it will be necessary to either reduce the
unit charge rate or the reaction severity to stay within
the catalyst circulation limit. Use of a high activity catalyst can counter the effects of low delta coke, but it is
important to select a catalyst with the proper coke
selectivity (coke to conversion relationship). Achieve
100 catalysts are formulated with ultra-high activity
zeolite to counter the effects of low delta coke, while

PTQ Q3 2015 141

Maximising fuels yields

Grace has been successful at incorporating both nickel
and vanadium metals trapping into FCC catalyst, mitigating the negative impacts of the metals. An example
of metals trapping performance is evident in Achieve
200 catalyst. Refinery operating data after reformulation
to Achieve 200 catalyst resulted in higher gasoline
yields and lower hydrogen, delta coke and slurry yield.
The superior metals tolerance of the catalyst allowed
the refiner to increase conversion without increasing
catalyst addition rate. The increase in gasoline yield and
drop in slurry resulted in a benefit of ~$0.70/bbl for the
refinery.
Maximising refinery octane
A common challenge reported by refiners operating on
unconventional feeds, such as shale or tight oil, is a loss
of gasoline pool octane, caused by reduced volume of
alkylation feedstock. Achieve 400 catalyst is formulated
with multiple zeolites with tailored acidity, to deliver an
optimum level of butylenes to keep the alkylation unit
full and maintain refinery pool octane. Incorporation of
isomerisation activity into the catalyst particle itself
results in a more desirable yield pattern than would be
realised by use of a traditional octane boosting FCC
additive. In addition, Achieve 400 has been shown to
increase the octane of FCC naphtha. In multiple
commercial trials, the catalyst is delivering incremental
octane and butylene, worth on average $0.6/bbl.
Managing unconventional metals
A resid unit experienced deteriorating performance as
a result of processing high carbon, high iron opportunity feedstock and experienced iron poisoning. The
symptoms of iron poisoning include a loss of bottoms
cracking, as feed particles are blocked from entering
the catalyst particle, and a drop in conversion. Catalyst
design can be optimised to resist the effects of contaminant iron (and calcium) found in many opportunity
feedstocks. The high diffusivity matrices in Achieve
possess an optimum distribution of mesoporosity,
maintaining performance because diffusion to active
sites remains unhindered, even at high-contaminant
metals loadings. Commercial data demonstrate an
improvement in gasoline and distillate yield, and a
reduction in both dry gas and bottoms yields due to
improved coke selectivity. Using base case economics,
the shift to Achieve resulted in an annualised yield
improvement of $10M over the base catalyst
formulation.
As refiners adjust their operating conditions and yield
objectives, Graces newest FCC catalytic solutions will
help them maximise profit and manage the uncertainties of increased opportunity crude processing.
For more information: greg.teneyck@grace.com

142 PTQ Q3 2015

Savings in exchanger network cleaning

The Super Clean System (SCS), an online heat
exchanger cleaning technology, is being used by a
major oil refiner in Japan at a refinery equipped with a
120 000 b/d crude distillation unit (CDU) with 44
exchangers in the heat transfer network. Distillation
units are large energy consumers in the refinery and
maintaining the heat transfer efficiency of the network
is of major concern for operators. Exchanger fouling is
complex in nature but for crude oil fouling can be categorised into the following types: chemical, corrosion,
particulate and inorganic. Despite attempts to mitigate
against fouling, inevitably there is a requirement to
clean the exchangers and recover the heat transfer
efficiency.
Historically, the exchangers of the Japanese refinerys
CDU had been cleaned by a conventional water jetting
method. However due to time constraints during turnarounds it was very difficult to effectively clean all of
the exchangers in the network, with priority given to
those that were badly fouled. The time and costs associated with opening and cleaning all of the heat
exchangers was unacceptable to the refiner, who in
2002 decided to trial the Super Clean Technology. The
process involves the injection of very low concentrations of an oil based detergent in a closed oil
circulation loop around the whole exchanger network.
The cleaning solution is circulated, heated and periodically sampled. Progress of the cleaning process was
monitored as a direct correlation with the lightness, L
value, of the solution (see Figure 1). As products of
fouling are removed, the L value decreases until all
products of fouling are removed and the L value tends
to constant.
After a successful trial, the technology has been used
on an annual basis at the refinery and was once again
chosen for the June 2014 turnaround. Cleaning of the
whole heat exchanger network was completed in less
then 48 hours, without the need to open exchangers.
This resulted in significant time, labour and costs
savings compared to the conventional water jetting
method.
80

L-value of fresh LCO = 94.3

70

L-value

delivering the proper coke. In commercial application,

Achieve 100 delivered enhanced activity over a competitive catalyst. At constant coke, the switch to Achieve
100 resulted in higher activity, higher gasoline and
lower bottoms and an economic uplift of ~$0.40/bbl.

60

50

Inj.

Circ.

40
13/6 18:00 13/6 21:00 14/6 00:00 14/6 03:00 14/6 06:00 14/6 09:00
12 hours

Sampling time

Figure 1 Lightness( L) value of cleaning solution vs time

www.eptq.com

Figure 2 Tube and shell, desalter-preflash

SCS is a cost effective, environmentally sustainable

In the case of the CDU, three out of the 44 heat
exchangers were pulled for inspection of cleaning effec- technology for reducing cleaning costs, turnaround
tiveness. The cleaning effect was high, as Figure 2 time and energy conservation for the CDU/VDU. The
shows, with products of fouling being effectively technology has superseded traditional mechanical
Readers can download technical articles and other
removed. The used cleaning solution was subsequently cleaning techniques, is now the method of choice for
information from over 40 categories.
reprocessed with the crude feedstock; there was no the operator of this refinery, and is being rolled out to
other
refineries
under
its operation.
waste water to treat; and, importantly, the cleaning Find
relevant
industry
events
worldwide.To date, the SCS
chemical was derived from a renewable source. The technology has been applied 75 times to successfully
news
items focusing
on contracts, company
clean
hundreds
of heat exchangers.
opportunity to clean all exchangers in the network in a Read
given time constraint manifested itself in energy savingsnews and new products.
For more information: ibaxter@cat-tech.com
through improved heat recovery at the furnace.

www.digitalrefining.com is an easy-to-use website for the refining,

gas and petrochemical processing industries

www.eptq.com

PTQ Q3 2015 143

Alphabetical list of advertisers

Advanced Refining Technologies

27

John Zink Hamworthy Combustion

Aerzener

79

Johnson Matthey Process Technologies

53
119

AFPM Fall Events

123

KBC Advanced Technologies

Albemarle Catalysts Company

IBC

Kobelco Compressors America

85

AltairStrickland

42

Kurita Europe

18

AMACS Process Tower Internals

56

Magnetrol International

64

Ariel Corporation

76

MAN Diesel & Turbo

96

Asia Downstream Week 2015

137

Merichem Company

109

Axens

OBC

Moscow Refining & Petrochemicals Week

134

BASF Corporation, Catalysts Division

IFC

Neuman & Esser

90

Bilfinger Water Technologies

50

OHL Guttermuth

87

Bhmer

12

Optimised Gas Treating

89

Bryan Research & Engineering

99

Paqell

34

Cat Tech International

114

Process Consulting Services

8 & 10

Prognost Systems

83

Refining India 2015

75

Rembe

69

Rentech Boiler Systems

32

CS Combustion

41

Sabin Metal Corporation

121

Delta Screens

45

Sandvik Materials Technology

24

72 & 143

Siemens Energy International

93

Streamlight

66

Chevron Lummus Global

ContiTech Schlauch
CRI International
Criterion Catalyst & Technologies

DigitalRefining.com

17
101
59

DuPont Sustainable Solutions

102

Emerson Process Management

60

Sulphur 2015

ExxonMobil

39

Sulzer Chemtech USA

Flexim

63

ThyssenKrupp Industrial Solutions

106

Grabner Instruments

95

Turbomachinery & Pump Symposia 2015

130

Grace Catalysts Technologies

31

UOP, A Honeywell Company

Hoerbiger Kompressortechnik

80

Wood Group Mustang

Intergraph CADworx & Analysis Solutions

62

Yokogawa Europe

13

ITW Technologies

14

Zwick Armaturen

70

127
46

23, 29 & 36
113

For more information on these advertisers, go to www.ptqenquiry.com

144

PTQ Q3 2015

www.eptq.com

AFX
No matter the feedstock quality, from light gas oils to heavy residues, Albemarles AFX
catalysts provide breakthrough performance. If maximizing the propylene yield from your FCC unit is high
on your agenda, you require more than a traditional catalyst. Thats why Albemarle supplies the innovative AFX catalyst
technology to a majority of customers who operate to maximize propylene. Through their unique morphology with high
accessibility, highly active and selective matrices and proprietary zeolite technology, AFX catalysts deliver unsurpassed
propylene yield. Ready to raise the bar on your FCC units propylene yield? Its time to maximize your potential with AFX.

In addition to maximizing C3=, AFX also minimizes dry gas and bottoms
and delivers superior yields of transportation fuels.

For more information on Albemarle AFX catalysts or our exceptional portfolio

of products and services, call (281) 480-4747 or visit www.albemarle.com.

REFINING SOLUTIONS

ALB15-001_AFX_prodA4.indd 1

3/18/15 6:00 PM

210/297.qxd:A4

13/08/10

12:12

Page 1

Your objectives
in focus
Make the most of todays and tomorrows
challenges with leading-edge solutions from Axens
- Clean and alternative fuel technologies
- Petrochemicals
- Energy efficiency
- High performance catalysts & adsorbents
- Revamps

Single source technology and service provider

ISO 9001 ISO 14001 OHSAS 18001
www.axens.net

axens.indd 1

13/9/10 13:27:18