Refining Gas Processing Petrochemicals: Petroleum Technology Quarterly
Refining Gas Processing Petrochemicals: Petroleum Technology Quarterly
Refining Gas Processing Petrochemicals: Petroleum Technology Quarterly
Q2 2018
REFINING
GAS PROCESSING
PETROCHEMICALS
SPECIAL FEATURES
5 ptq&a
Cover
Neste is rebuilding the wastewater treatment plant at its refinery in Porvoo, Finland. The project involves renewing the
chemical and biological treatment systems of the wastewater plant. The project is scheduled for completion in 2020.
Photo: Neste
©2018. 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.
Vol 23 No 3
Q2 (Apr, May, Jun) 2018
T
here are busy and secure times ahead for refiners and petrochemicals
Editor producers as demand for oil products leaps ahead into the next decade.
Chris Cunningham The International Energy Agency (IEA) has been updating its outlook on
[email protected] the oil industry and bases its forecasts on continued strong economic growth
across the globe and acceleration in oil production to meet the resulting growth
Production Editor in demand, for the next five years at least.
Rachel Storry
[email protected]
The US will lead that surge in crude output, supported by rising produc-
tion in Norway, Canada and Brazil. Naturally enough, there is a caveat to all
Graphics Editor of this, with the IEA pointing out that the oil industry will need to step up its
Rob Fris investment in production after years of decline, otherwise demand will begin
[email protected] to embarrass the supply side.
The oil industry’s investment in new production fell by no less than a quar-
Editorial ter in two of the last three years and stayed level through 2017. During the
tel +44 844 5888 773
fax +44 844 5888 667
same period, oil prices collapsed before OPEC and other producers reined
back their own production.
Business Development Director The situation has changed markedly since, with the US in particular driv-
Paul Mason ing forward, chiefly through the development of stable production levels from
[email protected] shale measures. According to the new outlook, the US will eventually pro-
duce 17 million b/d of crude oil, condensates and natural gas liquids – a rise
Advertising Sales Office of around 4 million b/d from recent levels – to maintain its global lead as a
tel +44 844 5888 771
fax +44 844 5888 662
producer of petroleum products. Through the same five-year period, OPEC
countries will add 750 000 b/d to the global market. In summary, production
Publisher capacity across the world will rise 6.4 million b/d by 2023 for a grand total of
Nic Allen 107 million b/d, while the US will account for 60% of the total.
[email protected] At the same time, rising demand across the globe, led this time by China
and India, will be not far behind output levels, lagging output by 2 million
Circulation
b/d or so and with no prospect of that demand slowing down until 2023 as
Fran Havard
[email protected]
petroleum counter-measures such as electric vehicles and increased energy
efficiency begin to have a significant impact on the balance in oil trade.
Crambeth Allen Publishing Ltd In the meantime, the main driving force for demand will be petrochemicals
Hopesay, Craven Arms SY7 8HD, UK which will account for a quarter of growth in demand for oil over the five-year
tel +44 844 5888 776 period. This trend again places US in pole position as production of ethane
fax +44 844 5888 667 and naphtha from its new oil and gas production sites increases the supply of
relatively low cost petrochemical feedstocks.
Register to receive your regular copy of
With supply outstripping processing capacity in the US, there is also the
PTQ at www.eptq.com/register crude export market to boost the fortunes of oil producers. Exports recently
reached levels of 2 million b/d; according to the IEA, and depending on prog-
ress with the construction of new pipelines and export terminals, that figure
PTQ (Petroleum Technology Quarterly) (ISSN
No: 1632-363X, USPS No: 014-781) is published
could rise to 5 million b/d within the next five years. While the Middle East
quarterly plus annual Catalysis edition by Crambeth will remain the biggest supplier of crude oil to Asian refiners, the US’s devel-
Allen Publishing Ltd and is distributed in the US
by SP/Asendia, 17B South Middlesex Avenue, oping role as an exporter is likely to feed into China’s investment in its petro-
Monroe NJ 08831. Periodicals postage paid at New chemicals industry.
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. CHRIS CUNNINGHAM
PTQ Q2 2018 3
www.CRITERIONCatalysts.com
Q Our ion exchange system for treating boiler feed water is to precipitate directly from the vapour phase. Ideally,
increasingly inefficient. Suggestions please for something more the salt point of a crude overhead neutralising amine
efficient and less costly to run. should be 15°F (8°C) lower than the water dew point
temperature of the column. In this case, any amine chlo-
A Jim Robinson, Senior Technical Consultant, Jim.Robinson@ ride salts formed will be within ‘reach’ of the water wash
SUEZ.com, SUEZ Water Technologies & Solutions employed to remove such corrosive salt deposits.
It is important to consult with a water treatment special- However, these ‘beneficial’ amine corrosion inhibitors
ist that understands your current water supply quality are not the only amine species present in the tower. There
and the treated water requirements for your system. This are several sources of ‘tramp’ amines as well, which are
information will enable the water treatment specialist to any organic amines not intentionally added to the crude
provide the best advice for your facility. overhead tower for corrosion control. Boiler steam con-
A common change many have made to save regener- taining tramp neutralising amines is added to the col-
ation chemical is to replace or assist their ion exchange umn as ‘stripping steam’ to improve the efficiency of the
system with reverse osmosis (RO) treatment of the water. distillation process. In addition to the amines used as
Sometimes RO treatment produces satisfactory water for steam system neutralisers, other sources of tramp amines
the intended use, sometimes the water needs to be fur- include the incoming crude slate and any slop oils, alka-
ther polished by ion exchange or electrodialysis reversal. nol amine units, hydrogen sulphide scavengers, and sour
On the other hand, since your problem is reported to water strippers.
be the ion exchange system becoming less efficient, per- Many of these tramp amines, including many used as
haps the most economical solution will be to find out steam system neutralisers, have elevated chloride salt
what has caused the loss of efficiency. Are regeneration point temperatures, and will form corrosive amine chlo-
flow rates, concentrations and temperatures properly ride salt deposits in tower areas not subject to removal
maintained? Is the proper amount of ion exchange resin by the water wash systems. In addition, as refiners lower
still in the vessel(s)? Are the resin beads fouled or has top tower temperatures to enhance distillate production
their exchange capability been depleted so resin replace- – jet fuel, diesel, kerosene – this corrosive salt formation
ment is required? will be driven into upstream areas of the distillation col-
A good water treatment specialist should be able to umn unreached by the water wash.
help you answer these questions and find the best solu- An analytical profile of the crude tower amine salt
tion to resolve your problem. deposits and overall overhead amine matrix, with com-
parison to the composition of the amines in the steam
neutralising amine blend, as well as the other tramp
Q We are getting amine salt fouling in our distillation column amine sources, can be used to model the behaviour of the
overhead. Are we using the wrong neutralising amine in our steam and other tramp amine species entering the tower.
steam system? Modern modelling software, such as the Suez LoSalt pro-
gramme, utilise this data along with crude slate amine
A Toni Rossi, Senior Boiler Technical Advisor, Toni.Rossi@ profiles, and additional tower operating data, to accu-
SUEZ.com, and Collin Cross, Senior Product Analytics/Support rately predict the potential for formation of corrosive
Manager, [email protected], SUEZ Water Technologies & amine chloride salt deposits.
Solutions If steam system amines are determined to be a signif-
The answer is ‘quite possibly’. However, a systematic icant source of corrosive salt deposits in the distillation
approach is required to determine if the organic amine tower, there are solutions available. One example is the
or, more commonly, the blend of amines applied to the Suez series of Steamate LSA (Low Salting Amine) steam
steam system for corrosion control are primary contribu- system neutralising amine products. These patent pend-
tors to corrosive amine chloride salt deposits in the crude ing technologies were recently developed utilising the
distillation column. Suez LoSalt software via collaboration between the Suez
Carefully selected organic ‘neutralising’ amines are hydrocarbon processing and boiler research scientists
purposely added to crude distillation columns to elevate and engineers to select steam system amines with lower
pH and reduce the corrosivity of the water formed in the salt point temperatures that are compatible with mod-
column through condensation. A key criterion for selec- ern crude tower operation, especially those working to
tion of these beneficial amine neutralisers is their chlo- enhance distillate production.
ride salt point at normal overhead conditions. This is In addition to the low salt point neutralising amine,
the temperature at which the amine chloride salt begins Steamate LSA technology incorporates a unique, vola-
CLG’s bottom-of-the-barrel
upgrading technology solutions can enable you
to produce low sulfur fuel oil to meet the tighter
International Maritime Organization (IMO) regulations.
www.chevronlummus.com | [email protected]
03M012018H
clg_PTQ_Q2_CLG_mar_2018.indd
clg.indd 1 1 1/10/2018 2:19:15
08/03/2018 PM
10:07
able control problems, a matrix of control models (based and coke production, lower VGO conversion, and lower
on step testing to understand process dynamics) is used gasoline yield. If the FCC catalyst instead is well pro-
to calculate the response of each controlled and con- tected by feed pretreatment, the consequences can actu-
straint variable based on changes to manipulated vari- ally be minimised or even avoided.
ables or measured disturbance variables. Step testing Haldor Topsoe is a leading supplier of pretreat cat-
should be small enough to avoid disturbing the process, alysts for FCC units and is able to tailor a pretreatment
but large enough to see the response over process noise. reactor loading that can handle all kinds of contaminants
Advances in technology have resulted in some DCS in the VGO. In particular Ni, V, Fe, Si, As, and P are big
host systems to include embedded MPC. With MPC run- concerns currently and these contaminants all have the
ning in the DCS controllers, it is redundant and fast (once potential of ruining the FCC catalyst workload. The best
per second compared to MPC on top of a DCS which solution for the refiner is to install an effective hydrode-
typically executes once per minute). Embedded MPC metallisation (HDM) catalyst as part of the pretreatment
also provides the option to be part of offline simulation reactor. The best HDM catalysts are recognised to have a
and training. The other advantage of embedded MPC is porosity and activity that will trap the contaminants and
avoiding connectivity issues with MPC added to a DCS accumulate them to achieve the projected cycle length.
which includes database synchronisation, watchdog tim- Nickel, for example, will react with H2S and precipitate
ers, and fail/shed logic design. as NiS inside the catalyst pore system. This means that
When explaining MPC, the analogy of driving a car an effective Ni guard catalyst is equipped with large
down the highway is used to better understand the pores and a significant pore volume to facilitate a high
advantage of using this multivariable control over PID. storage capacity. Based on industry experience, Haldor
For regular PID control, you drive while looking out the Topsoe’s R&D has developed a complete HDM cat-
side window. With a curve coming up on the highway, alyst portfolio, which currently includes MultiTrap,
you need to go off the road before PID makes a correc- SiliconTrap, and top-tier guards, such as TK-743. In any
tion to get back on the road. For MPC, the advantage is case, it is advisable to interact with the catalyst provider,
you are looking out the front windshield; you see the as the best HDM catalyst solution is always tailor-made
curve coming and easily adjust the car direction to avoid to match the exact feedstock type and requires knowl-
going off the road. edge about, for instance, additive properties.
For MPC, you generate models based on step tests
to predict where the process is going. Many controllers A Marco Roncato, Senior Product Manager Process
are manipulated at the same time based on the models, Development & Marketing, [email protected], Chimec
minimising the impact from any disturbance. The result The FCC plant is one of the main conversion units in a
is much less variability while operating closer to con- modern refinery and especially in these last years it is
straints without violating them. A good MPC example called on to process heavier feedstocks more than ever.
is crude switches that have minimal impact to the crude When the feed is a residue, it is unavoidably contam-
fractionator while maintaining side cut specifications. inated by metals which – reducing the catalyst’s activ-
Note, MPC can have different operating objectives that ity and promoting the feed’s dehydrogenation reactions
can be changed easily depending on market conditions. – leads to a reduction in conversion levels and to an
Examples of operating condition changes include but are increase in hydrogen and coke.
not limited to: maximise throughput, minimise energy Such metals can be nickel (Ni), vanadium (V) and
use, maximise one product/intermediate over another. sodium, and they all concentrate in the bottom residues.
It is important to train console operators on how to use All of them are catalysts themselves, triggering unde-
the new MPC, and ensure they understand the operating sired dehydrogenation and condensation reactions, end-
objectives. Also, do not forget about modifying standard ing up in increased hydrogen and coke yields, and at the
operating procedures if required. same time reducing naphtha. Moreover, they all deposit
on the catalyst’s surface hence reducing its activity.
But while sodium content can be easily managed pri-
Q Heavier feeds to our FCC are resulting in excessive marily with good desalter management – that means
hydrogen and coke production. We think nickel contamination minimising caustic injection downstream – on the con-
of the catalyst is to blame. Any solutions? trary both Ni and V are almost impossible to remove
from the residue, because they are present in porphy-
A Michael Tinning T Schmidt, Senior Product Manager, mtp@ rinic form.
topsoe.com, Haldor Topsoe A/S One solution can be a hydrodemetallisation section
It is a fact that many FCC feeds have become more upstream of the FCC, but this is obviously a very expen-
and more hard to process in recent years. Opportunity sive solution, although effective. Otherwise, it is possible
crudes, cracked feeds, heavier cuts, and use of additives to inject, in the FCC feed, a proper nickel passivator in
are enabling FCC units to make more money than before; order to protect the catalyst’s activity.
however, the feedstock changes come at a price when it These passivators – usually antimony (Sb) based,
comes to catalyst performance. If the FCC catalyst is not although many others are claimed to be active – have
sufficiently protected from the contaminants present in been used for many years to reduce nickel’s detrimental
the feed, it will significantly impact the overall perfor- effects. Usually they have to be injected in the feed before
mance of the FCC complex in terms of more hydrogen the riser, with the help of a carrier like light cycle oil.
Maximize volume
swell and produce
more barrels
Nitrogen in the feed limits aromatic saturation, density reduction
and volume swell. Removal of nitrogen is essential for the
yield improvement both in your hydrocracker and
ultra-low sulfur diesel hydrotreating unit.
www.topsoe.com
Learn more about how we can work together to grow your business.
www.catalysts-licensing.com
© 2016 Exxon Mobil Corporation. All rights reserved. ExxonMobil, the ExxonMobil logo, the interlocking “X” device and all product names herein are trademarks of Exxon Mobil Corporation.
Goodnight
reactor worries
Put problems with fouling and pressure drop to bed. Our proprietary
internal filtration design eliminates crusting during purification to reduce
pressure drop. This improves throughput by as much as 50 percent – letting you do more than
dream of longer production cycles. For questions or to learn more, visit crystaphase.com.
API 618
HIGHEST AVAILABILITY
COMBINED WITH BEST Over 120 years of experience in valve Full range of services and top
MAINTAINABILITY; FAST design and manufacturing performing components through
ACCESS TO ALL WEAR PARTS Outstanding durability of Redura® global organization and local service
sealing elements for longest mean centers
Full range: time between overhaul (MTBO) → www.recip.com/api618
Rod load up to 1’500 kN / 335’000 lbs Worldwide Burckhardt Compression
Power up to 31’000 kW / 42’100 hp quality, engineered in Switzerland
Lubricated up to 1’000 bara, Your solution partner – From bare
non-lubricated up to 300 bara shaft compressors to turnkey
Advanced solutions for demanding solutions
and sour gas applications
burkhardt.indd 1
02_API618_Ext_Range_210x297RA_RZ.indd 1 08/03/2018
13.06.16 10:04
14:21
Shell Global Solutions
TECHNOLOGY
INNOVATION
IS OUR DNA
We, at Shell Global Solutions, constantly create
new tailored solutions with our clients to exceed
their expectations.
shell.indd
228646 1
GS_TechnologyDNA_A4_V07.indd 1 23/02/2018 08:55
06/03/2017 16:28
reported extremely quick localised corrosion attack in o Operate upstream accumulators at design conditions
the past, especially at deeper located areas of the feed/ to ensure proper phase separation
effluent heat exchanger network. It is nearly impossible o Resize upstream accumulators to improve phase
to keep the naphtha hydrotreater system dry so that no separation
corrosion will occur. Wash water injection only provides o Install other mechanical methods (such as coalescers)
partially successes and is a limiting factor in naphtha upstream to produce drier reactor feeds
hydrotreaters. • Minimise HCl content in the hydrogen treat gas that is
Kurita’s ACF technology is a very good alternative to fed to the hydrotreater reactor
eliminate or significantly reduce the corrosion poten- o Monitor hydrogen treat gas to ensure that HCl target
tial at naphtha hydrotreaters. The applied ACF products limits are consistently met
are formulations with a very strong organic base, which o Improve performance of chloride traps/guard beds
immediately reacts with HCl or other acidic components. as required.
Already formed NH4Cl salt deposits are dissolved by In many cases, even with proper contaminant control
replacing the ammonia. The former deposits are trans- methods, the formation of NH4Cl in the reactor effluent
ferred into liquid ACF salts, which have a neutral pH cannot be prevented. In these cases, the best mitigation
and a very low corrosivity. During operation, ACF prod- option is to implement water wash injection to the efflu-
ucts can be applied periodically for some time at higher ent system. Depending on the location of salt forma-
treat rates to remove ammonium salt fouling. The better tion, water wash may be needed in the reactor effluent
option is continuous dosage with low treat rates. Dosed exchanger train and/or at the inlets of downstream air
continuously, the ACF programme avoids salt precipita- coolers. Typical wash water sources include boiler feed
tion and keeps the metal surfaces clean, so that no corro- water, stripped sour water and steam condensate.
sion attack occurs. The primary objective of such a water wash system is
When flooding with corrosion attack in the stabiliser to scrub NH3 and HCl from the vapour phase by par-
column top section is observed, Kurita’s ACF technology titioning these contaminants into a condensed water
will also provide a perfect solution. A shutdown of the phase. In this way, the risks of under-salt corrosion as
naphtha hydrotreater for one or two days to perform a well as hydraulic limitations (pressure drop increases)
water wash is no longer necessary, avoiding extra costs can be minimised. At a minimum, a proper wash water
for cleaning or production losses. Based on practical system should include the following features:
experience, the ACF product can be applied during oper- • Water injection locations which are upstream of salt
ation as online cleaning for some hours. formation zones
• Sufficient water injection rate to force saturation + 25%
A Mark G Riley, Fellow, Adsorbents Development, mark.riley@ volume excess
uop.com, Honeywell UOP • Properly designed atomising spray nozzles
Ammonium chloride deposits are the result of chlo- • Divided water injection across parallel exchanger or air
ride ingress into the unit. The most common source is cooler inlets
improper clean-up of make-up hydrogen imported • Flow meters on each water injection line.
from a reforming unit. Chloride removal from reformer In addition to the techniques detailed above, it is criti-
net gas can be accomplished using an adsorbent or by cal to frequently monitor NH3 and HCl levels in effluent
water washing. Chlorides may also originate in the accumulator waters. These measured values, along with
liquid feed either due to poor desalter operation or relevant operating parameters (flow rate, temperatures,
due to organic chlorides in crude; often due to surfac- pressures), should be used as inputs to calculate NH4Cl
tants employed to enhance crude oil recovery. It may formation temperatures. Variations in contaminant lev-
be necessary to periodically switch the wash water els may shift salt formation locations such that the exist-
to the intermittent injection point upstream of the ing water wash system is no longer adequate to provide
normal wash water injection point to flush out desired benefits. In these situations, alternate water
ammonium chloride salts. Ammonium chloride salt wash injection locations, likely requiring increased
deposits at a higher temperature than ammonium water wash rates, would be needed. The effect on efflu-
bisulphide salt, which is the usual focus of water wash ent accumulators to properly separate the additional
programmes. water loading must also be examined to avoid free
water in effluent stripper/stabiliser feed streams.
A Randy Rechtien, Field Technical Services Manager, Randy.
[email protected], Baker Hughes, a GE company Q How can we minimise NOx emissions from our FCC?
There are two primary methods for addressing ammo-
nium hydrochloride (NH4Cl) salt deposition in hydro- A Michael Tinning T Schmidt, Senior Product Manager, mtp@
treater effluent systems: contaminant source control and topsoe.com, Haldor Topsoe A/S
water wash injection. The best approach to handle NOx emissions from the
Contaminant source control efforts can be employed FCC unit is actually done upstream in the FCC pretreat-
to reduce the levels of both NH3 and HCl in the reactor ment reactor. By applying a high-active hydrotreating
effluent. Control methods should include the following: catalyst, nitrogen compounds present in the vacuum gas
• Eliminate free water in the naphtha feed to the hydro- oil (FCC feed) can be reduced significantly by hydro-
treater reactor denitrogenation (HDN). By reducing nitrogen before
grace.com/value
Learn more about how
Grace has delivered
value to its customers
through collaboration.
Visit grace.com/value
1200
Baseline
1000 250
45−55% NOx reduction
NOx, mg/Nm3
using CP P
800 200
600 150
Additional 35%
400 100 NOx reduction
using DENOX
200 Predicted NOx 50
Actual NOx
0 0
Time NOx reduction project 2017−18
Figure 1 Switching from a Pt based combustion promoter to CP P Figure 2 DENOX in combination with CP P, providing an additional
decreased a refinery’s NOx levels by about 70% 35% NOx reduction Graphs by permission of Hellenic Petroleum
NOx, ppm
200
NOx, ppm
150
150
100
100
50
50
0
0 0.2 0.4 0.6 0.8 1.0
0
−100 −120 −60 0 60 120 180 240 300 360 420 480 Ecat Pt, ppm
Time, days
Figure 1 It can take a long time to completely flush platinum Figure 2 Small quantities of residual platinum can have a major
based CO promoters out of the regenerator effect on NOx production
NOx, ppm
O2 100
Aromatic N HCN N2
80
(coke)
O2 60
H2O
40
O2 20
Amines NH3 NOx
(unstripped 0
products) 0 0.5 1.0 1.5 2.0 2.5
O2, %
Figure 3 Recent research has focused on conversion of HCN and Figure 4 NOx formation is a function of temperature and oxygen
NH3 to N2 instead of NOx content in the regenerator
ing an additional 35% NOx reduction. Both technologies the regenerator. The primary products of coke/nitrogen
are currently in use at the refinery, which is resulting in combustion are HCN and NH3. NOx is mainly formed
historically low levels of NOx (see Figure 2). from the combustion of these ‘reduced’ species. The lat-
est NOx reduction research has focused on conversion of
A Ron Butterfield, Senior Technical Service Engineer, Ron. these species to N2 instead of NOx (see Figure 3).
[email protected], Johnson Matthey If switching to a non-Pt promoter does not provide suf-
Major causes of high FCC NOx emissions include the ficient NOx reduction, the next step is the use of a NOx
use of platinum (Pt) containing CO promoters, uneven additive, such as Johnson Matthey’s NO-NOX, which is
air/coke contacting in the regenerator bed and excessive normally added at 1-2 wt% of the circulating catalyst and
additions of antimony nickel passivator. can provide an additional NOx reduction of up to 50%
Pt CO promoters have been mostly eliminated in units over non-Pt promoter alone.
with NOx emissions limits. However, some units using Uneven air/coke contacting in the regenerator bed can
non-Pt promoters still have Pt promoter in the ware- also lead to excessive NOx formation in the regenerator.
house, which can be accidentally added, resulting in NOx formation is a function of temperature and oxy-
higher than normal NOx emissions. Also, be careful that gen content in the regenerator. Higher temperature and
any flushing purchased equilibrium catalyst does not higher oxygen content will both lead to increased NOx
contain residual amounts of Pt. Commercial data show production (see Figure 4).
the decrease in NOx emissions with the shift from Pt There may be NOx generating hot spots in the regen-
based CO promoters to non-Pt CO promoters such as erator due to poor mixing caused by either mechani-
Johnson Matthey’s COP-NP. You can see from Figure 1 cal damage, design problems, operation outside of the
that it can take a long time (four to five months depend- design parameters of gas or catalyst velocities or high
ing on catalyst inventory and fresh catalyst make-up oxygen content. The hot spots may not be apparent with
rate) to completely flush the Pt out of the regenerator. the small number of dense bed thermocouples available
Even small quantities of residual Pt can have a major in most units. It may be helpful to perform a scan of the
effect on the NOx production in the regenerator (see regenerator to see if there are non-uniform flow patterns.
Figure 2). NOx emissions are directly related to the equi- Mechanical problems may require a unit shutdown to
librium catalyst Pt content. correct equipment damage or make hardware changes.
Once the Pt is removed from the regenerator, it may be Excessive additions of antimony nickel passivator will
necessary to trial a new generation NOx reduction addi- tend to increase NOx production. Therefore, close control
tive that can help to prevent the formation of NOx in of antimony additions is very important.
Higher reactor LPG yields increase wet gas compressor There is a strong incentive to maximize FCC olefin
loading. So, in a unit that is already limited by wet gas production with reactor temperature and catalyst
compressor capacity, revamp modifications must be formulation. Executing the correct revamp can
made to accommodate higher reactor LPG yields. economically address main column, wet gas
Replacing the compressor entirely or installing a new compressor, and gas plant constraints so that
parallel compressor are expensive options. theoretical reactor yields become actual barrels of
valuable product.
Instead, the keys to a practical offgas compressor
revamp lie in the polytropic head equation (bottom
right). Look to minimize main column inlet to wet Polytropic Head Equation
gas compressor inlet pressure drop and overhead
receiver temperature. Next, compressor speed and
rotor modifications may afford additional capacity. On
a mass basis, FCC wet gas rates have been increased
by over 40% without installing a parallel compressor.
T
o improve refinery perfor-
mance by reducing fuel oil
make, an FCC revamp was
envisaged to process heavier feeds
and maximise unit conversion and
utilisation. One of the key objectives
of the FCC revamp was enhanc-
ing the safety and reliability of the
unit to achieve a cycle length of
four years. KNPC scouted several
of the latest available technologi-
cal options and selected UOP tech-
nology including VSS, Elevated
Optimix feed distributors and the
Advanced Fluidized (AF) stripper
for the reactor revamp. Following
unit commissioning, the technology
has performed successfully. Heavier
feed was processed and met unit
objectives in terms of yields as UOP Figure 1 Site of KNPC Mina Al-Ahmadi refinery
and KNPC have jointly conducted
optimisation tests Propylene recovery, MTBE and and safety performance of KNPC
alky units were added in the same refineries
Project background revamp. With changing market • Execute a major upgrade of
KNPC operates three refineries in demand for fuel quantity and qual- the MAA and Mina Abdullah
Kuwait. They have a combined pro- ity, KNPC continued exploring (MAB) refineries to convert high
cessing capacity of 460 000 b/d of the possibilities to further increase sulphur fuel oil to higher value
crude oil to produce primary prod- the unit’s capacity, and possibly to products
ucts like petrol (gasoline), diesel upgrade the technology to process • Respond effectively to KPC/
fuel, kerosene, liquefied petroleum heavier feeds and improve yields. KNPC strategic directives for
gas (LPG), heating oils and bitumen. The Clean Fuels Project (CFP) 2020 expanding refining capacity in
KNPC Mina Al-Ahmadi (MAA) provided the opportunity to achieve Kuwait
refinery (see Figure 1) has a UOP this objective. • Provide new employment oppor-
FCC unit which was designed in tunities for Kuwaitis in the refining
1984 for a feed rate of 30 000 b/d KNPC’s CFP 2020 vision sector.
of vacuum gas oil (VGO). The unit As a part of the CFP 2020 project,
started up in 1987 and consists of KNPC set out the following project FCC project objectives
a high efficiency style combustor objectives: The CFP 2020 vision was incorpo-
regenerator, a downturn arm riser • Develop KNPC refineries into rated into the FCC project’s objec-
termination system in the reactor, integrated merchant refining com- tives as follows:
and feed injected into the riser Wye. plexes to meet diversified market • An additional 6% increase in unit
It has since been revamped once by needs capacity (42 500 b/d)
UOP in 1997, to increase the capacity • Meet future market demand and • Improve unit reliability and
to 40 000 b/d with a feed of either specifications for fuels for local and on-stream factor
VGO or HVGO and some coker international markets • Process difficult feedstocks and
gas oil. • Enhance the environmental optimise conversion
Equipment Design Matters In both the main and external fractionators, liquid
distributors must be designed for practical flow rates
Many attractive projects fail to meet expectations at
and to handle solids. Unsophisticated distributor
startup. Disappointing performance often results from
design creates uneven liquid distribution that reduces
bad simulation practices and/or poor equipment design
fractionation efficiency and LCO recovery against
rather than faulty execution. Refineries are currently
the endpoint specification. The main fractionator
considering FCC revamps to increase olefins for
slurry pumparound and quench distributors must
more alky unit feed, maximize LCO product recovery,
eliminate hot spots in the grid and bottoms liquid pool,
and minimize slurry product by producing HCO for
respectively, to prevent coke formation. The picture
hydrocracker feed. These changes raise fractionator
below illustrates the result when equipment design is
operating temperature. Higher temperatures require
left to low-cost vendor solutions.
better process and equipment designs to avoid fouling
and coke formation, which lead to poor reliability and Finally, the bottom product from the external
potentially to an unscheduled shutdown. fractionator (reduced slurry) will be nasty. Stripping
trays must be specially designed to work in this
While getting the simulation right is important, process
extremely fouling service, and bottoms pumps must
equipment design is equally critical to a project’s
be compatible with very low API material containing
success. Consider a project to minimize FCC main
solids.
fractionator bottoms product (Slurry, DCO, CSO, etc.).
As outlined in the top figure, an external fractionator Equipment design matters. Don’t miss performance
can recover substantial quantities of LCO and HCO goals by applying generic equipment design to
from the FCC slurry product, reducing slurry volume by specialized problems.
60% - 70%.
Connected Plant
© 2018 Honeywell International Inc. All rights reserved.
T
he characterisation and need to be determined separately resins, and asphaltenes. In particu-
reactivity of heavy oil frac- and this comes at a price (both time lar, measurements of asphaltenes
tions (vacuum residue) were and money). Organisations such as are important for heavy oils and
studied using a PIONA (paraf- the American Society for Testing and residues to determine solid deposi-
fins, iso-paraffins, olefins, naph- Materials (ASTM) have developed tion probability, usually an issue in
thenes, aromatics) molecular routine tests for determining boil- not only the production industry,
approach. Eleven crude oil assays ing ranges and properties of crude but also in transportation and refin-
along with vacuum residue pilot- feedstock, distillation fractions and ing. Alternatively, the fractions of
scale thermal cracking informa- products, however the bulk of these saturates and aromatics are of less
tion were provided. These feeds techniques were originally intended importance due to their redundancy
were characterised and their sus- to capture the properties and com- with other lighter range properties
ceptibility to thermal cracking position of conventional oils. Heavy typically measured.2-3
was evaluated all using a PIONA oils pose a challenge since they con-
rigorous approach. A reactivity tain a large, non-distillable fraction.1-3 Introduction to heavy oil kinetics
index related to the naphthenes- Commonly measured properties When dealing with reaction kinetics
aromatics-dehydrogenated aromat- in these non-distillable fractions, for thermal cracking of heavy oils,
ics ratio was found for each feed in such as density and viscosity, can a considerable amount of material
order to match experimental results. prove to be helpful when deter- has been published on the subject
This reactivity index was then cor- mining the physical and transport in the last few decades. This mate-
related to the vacuum residue feed properties of these fractions. Other rial focuses more on the molecular
properties. The analysis of these properties, for example carbon res- structure approach to solving the
results points to C7 asphaltenes (or idue, SARA (saturates, aromatics, problem.4 This approach is a more
equivalent), carbon residue and resins, asphaltenes) analysis and rigorous alternative to the lumped
density of vacuum residue as key pour point also prove to be help- kinetic schemes that are being
properties to be measured in order ful when determining not only the replaced due to lack of predictive
to capture the chemical nature of quality of an oil product, but also its accuracy across differing feedstocks.
this fraction, and thus proper reac- chemical nature and reactivity.2-3 Moreover, this surge in molecu-
tivity in a thermal cracking process. Carbon residue is a particularly lar structure modelling comes at
Identifying and quantifying each important characteristic of crude oil a time when personal computers
component contained in oil fluids residues, since it not only can indi- have reached a point at which pro-
has proven to be both impractical cate the quality of the fraction, but cessors can keep up with the enor-
and unfeasible. Characterisation of also can be correlated to a number mous computational demands of
crude oils and their different cuts is of properties such as hydrogen to approaching the problem. It should
commonly performed by measuring carbon ratio (H/C), heteroatomic (S, be noted that molecular structure
different properties. Traditionally, N) content, asphaltenes content, or based kinetics, as well as physical
techniques such as gas chromatogra- viscosity.2 property calculations, have been
phy (GC) for lighter fractions (light The pour point of an oil fraction around since the 1960s and many
ends and C1-C5) and the use of distil- indicates the minimum temperature influential and detailed papers on
lation curves are commonly used for at which this material will flow. This the topic were published as early as
the liquid fractions. In this regard, property is affected by the presence the 1980s.5-6
the properties (densities, molecu- of heavy molecules (which increase When looking at the thermal
lar weight, chemical family, and pour point), thus it can be correlated cracking of heavy oils, or even
so on) of the light ends and lighter to molecular weight and density.2 approaches to catalytic cracking
component ranges are usually well SARA analysis proves to be useful and processing, there is an empha-
known, however the physical prop- for residue fractions, which usually sis on aromatic groups with or with-
erties of the remainder fractions contain large amounts of aromatics, out significant saturated branching.7
hwi.indd 1
HWI_PTQ_Full Page Ad_APRIL2017.indd 1 08/03/2018
2/24/17 10:13
12:11 PM
dupont.indd 1 10/03/2017 10:54
each potential reactive pathway
need to be calculated. For thermal 60
C40
cracking systems where hydrogen 4 aromatic rings
50
donors are not readily available, the 2 naphthene rings
main available cracking pathway
Contribution, %
40
for the larger aromatic components
consists of scissor-type reactions
30
of the saturated chain branches.
Once these branches have been 20
removed, the aromatic rich core
structure left usually will no longer 10
crack and will ultimately become
pitch or coke. These compounds 0
are then being tracked as dehydro- Naphthene Aromatic Dehydrogenated
aromatic
genated aromatics in this method’s
application. 60
With the characterisation of crude C40
feeds, the importance of matching 50 4 aromatic rings
the amount of heavy aromatic core Contribution, % 1 naphthene ring
components then becomes essen- 40
tial in order to get the kinetic rates
that would match what is occurring 30
within thermal cracking reactions in
the process. If too many saturated 20
components were estimated in the
feed characterisation, the calculated 10
cracking rates would over-predict
0
lighter material yields. On the other Naphthene Aromatic Dehydrogenated
hand, if too many ‘aromatic core’ aromatic
type components were estimated,
which would be balanced by very 60
C40
saturated components, this would
3 aromatic rings
lead to over-prediction of both 50
3 naphthene rings
lighter and very heavy range yields
Contribution, %
40
with a lack of middle range liquid
product.
30
Luckily, with many laboratory
analysis methods the carbon residue 20
is something commonly measured
for heavier boiling range mixtures 10
and is shown to relate to H/C.6
From this measurement, the amount 0
of dehydrogenated aromatic com- Naphthene Aromatic Dehydrogenated
aromatic
ponents can also be estimated and
the kinetic rates and yields from
thermal cracking are properly pre- Figure 2 Representation of different C40 naphthene and aromatic ringed species in a
dicted. In many cases, the yield dif- PIONA slate
ferences from two crude feeds that
have similar distillation curves and measured (such as density, carbon Relative density = Sample-density/[Lightest
densities come from these heavier residue, and so on) and the reactiv- Sample density] (3)
component structure distributions. ity of the vacuum residue to ther-
mal cracking. This widespread range proved to
Methodology of characterisation The first step consists of the char- be important in order to capture the
Oil characterisation information acterisation of these feeds. The effect of the different physical prop-
was provided by Fuji Oil Co. Ltd. physical properties of the crude erties on the reactivity of each vac-
(FOC) for 11 crude oil feeds. This oil (and its vacuum residue cuts) uum residue.
information, contained in stand- cover a wide range as can be seen in The oil assays for these crude oil
ard assays, is going to be used in a Figure 3. This plot shows the relative feeds contain yield information for
PIONA backed simulation environ- densities of the crude oil feed and 14 cuts, and physical properties
ment to determine the relationship vacuum residue fraction relative to such as density, sulphur and nitro-
between several physical properties the lightest sample (see Equation 3): gen content, viscosities, and so on
1.14
produce higher liquid/lower gas
1.12
yields from vacuum residue feeds.
lightest sample
Plant
Simulation
Level
sabin.indd 1
Sabin-ThinkGlobally-QR.indd 1 11/03/2018
3/16/17 2:5413:48
PM
I NNOVATION
FLEXIPRO ™
technology,
the next generation
fixed valve tray.
The FLEXIPRO™ valve tray approaches the turndown ratio of movable valve trays without
the drawbacks:
• Valves cannot stick closed
- Uniform vapor distribution maintains tray efficiency over the full operating range
• Non-moving valve legs do not cause erosion or enhanced corrosion
- Maintain uniform open area and good vapor distribution
Pitch yield
found yields
Common
feed
VR
Sa le 7
Sa le 8
9
1
Sa e 2
Sa le 3
Sa le 4
Sa le 5
Sa le 6
0
11
Sa le 1
Sa ple
e
e
pl
pl
ce
pl
p
Laboratory Eureka Plant
m
m
n
m
Sa
Sa
re
scale model model
fe
Re
Pitch pour point
Plant data
Operational
Points 1, 2...
Thermal cracking
VR
0
11
Delayed coker
Sa le 1
e
e
pl
pl
pl
pl
pl
pl
pl
pl
pl
e
ce
kinetics
pl
p
m
m
n
m
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
re
Sa
fe
Re
0
11
Sa le 1
e
e
pl
pl
pl
pl
pl
pl
pl
pl
pl
e
tion model corresponding to FOC’s
ce
pl
p
m
m
n
m
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
re
Sa
fe
Re
Sa le 1
Sa le 2
Sa le 3
Sa le 4
Sa le 5
Sa le 6
Sa le 7
Sa le 8
0
11
Sa le 1
Sa ple
pl
p
m
m
n
m
re
Sa le 7
Sa le 8
9
1
m 2
Sa le 3
Sa le 4
Sa le 5
Sa le 6
0
11
Conclusions
Sa le 1
Sa ple
e
e
pl
pl
ce
p
The PIONA approach offers not
pl
p
m
m
n
m
Sa
Sa
Sa
re
9
1
0
11
Sa le 1
e
e
pl
pl
pl
pl
pl
pl
pl
pl
pl
e
e
p
m
m
m
m
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Quinolene insolubles fere
Sa
0
11
Sa le 1
e
e
pl
pl
pl
pl
pl
pl
pl
pl
pl
pl
p
m
m
n
m
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
re
Sa le 1
Sa le 2
Sa le 3
Sa le 4
Sa le 5
Sa le 6
Sa le 7
Sa le 8
0
11
Sa le 1
Sa ple
pl
nc
p
m
m
m
m
re
Reactivity index
days * $2.0/
be carried
$ 3.5 million/y lars in revenue to India’s oil and (COS and CS2) at lower operating
slurry with chemical settling aids, World
the out (blend of feeds, results
differ- Formulated MDEA reference Analyser solvent
affect catalyst, which
Simulated gas sector are •
lost daily with- Refining,
1.5beingDesalter Reduced
temperatures heat
Nov/Dec1999. transferItin
(<220°C). also heaters,
has a
ent operating
in more conditions,
frequent upset
need for scaling
ends up,
catalyst
Jefftreat Desalter
Ultra
The only meaningful process cost out considering 1 the impact2 of envi- causing
5 Motaghi
low more
M,
pressure
Shreefuel
drop, consumption
K, Krishnamurthy
which is and
S, Anode
impor-
and so on), but also implies poten-
change-out.
for the electrostatic separator is for ronmental waste and the cost of higher cost
grade coke from traditional crudes, PTQ, Q2
tial monetary tant as the SCOT process operates at
flow.saving byscale,
reducing the lost revenue1.0 2010.
Preventative actions that are taken
recycle For this the recy- from the bottom of the 6• Hot spots in heating
near tubes,and which
need tothe
run cracking Elliott atmospheric pressure andan
would experiments.
H2S in treated gas
removed.
crude oils,Assuming
Characterization like
of heavypolarlandfill
oils andresins is $1.0/
bitumens. and 1. oil production. affected.
Victor M Scalco III is Global Commercial
Time, minutes efit from reduced operating costs,
Lb, the cost is:
asphaltenes. To prevent corrosion, it content and1.5emulsion properties Strategic lowerthe
As cost ofDevelopment
desalting depends
Vapor pressure and critical constant prediction Business Manager,
emissions, improved oper-
is essential
method
Figure 2
for
Upset to
heavy pre-process
hydrocarbons,
resilience of the
Jefftreat crude
Energy
Ultra
& will frequently fluctuate
Source: according
Shell analysis primarily on crude oil quality, refin-
Gulftronic
ability
Electrostatic
and increased
Separators with
capacity.
1600
Fuels, tons/year
2007, *
22, 1, 455-462. $ 2000/ton =
oil in the desalter which breaks the to the crude $ 3.2 Further reading
1.0feed in the desalter. eries
General Atomics Electromagnetic Systems
must make a compromise
4 million/year
Kumar H, Froment G F, Mechanistic kinetic 1 Guercio V J, US producing, exporting more Group. Additionally,
He has worked for new for facilities,
more than cap- 20
oil/water
levels and emulsion, separating the Some crudes are difficult to desalt, between the maximum to allowable
modeling of treated gas specifications.
the hydrocracking of complex (Criterion)
slurry oil, Oil &C-834 high
Gas Journal, activity,
4 Oct 2010. low yearsital costs
in theare designexpected be lower
and implementation
water
Critical from
to the
achievingcrude oil, and thus such
2 Platts,as Venezuelan extra heavy salt content
to theand lowest costseparation
of desalt-
feedstocks, such
The annual
removing the salt. savingsagasare
as vacuum step change
oils,$3.5
Industrial temperature
mil- crude, Doba
SCOT
0.5
Methodology andcatalyst.
crude
Specifications
and Canadian
owing
Guide, of
ing
hydrocarbon
when
smaller
choosing
filtration equipment
and
their crude
size
oilfor
or
&inlion
the– $0.04
performance
Engineering Chemistry
million of
Research,the estab-
2007,
+ $3.2 million 46, The
Petroleum Jefftreat
Product & Ultra
Gas solvent
Liquids: US has
Caribbean
= crudes, and form HCl in the pre- blend. and
systems, reduced
involved equipment
in technical count
development
18, To
lished increase
5881-5897. the refining margin,
process has been the devel- superior H2S 0 selectivity over car- the enhanced process.
and Latin America, Jan 2012. and training with EPC and FCC/RFCC licensors
$6.7 million/year
refineries purchase crude oils and of heater of the desalter.
3 Silverman L D, Winkler S, Tiethof J A, worldwide. With experience in programme
bon dioxide (CO2) at high VR temper-
5opment,
Quann R jointly
J, Modeling with Huntsman
the chemistry
density and C7 asphaltenes
crude blends
Corporation,
complex petroleum at the lowest
of the highly selec- atures
mixtures, cost,
Environmental High and
Witoshkin salt iscontent
A, Matrix in heavy
effects
therefore capableand
in catalytic development
Significance
of scoping
Smaller
forofcommercial
plants on-line crude applications,
oil
The road ahead
often as heavy
Health Perspectives, and
1998, sour
106, Supplopportu-
6, 1441. opportunity
cracking,
achieving ultracrudes
presented at also
the NPRAplays an
annual
SO2vsemissions.
lowWestin analysis studies
for and commissioning,
optimised desaltinghe holds
tive
IndiaJefftreat
is looking Ultra solvent
to further and Figure
develop aSmaller
MA degree(B)plants
from costindex
the Universitylessof to
asSan build.
Diego.
nity crudes which are blended with
6 Quann R J, Jaffe S B, Structure-oriented important
Criterion Catalysts & Technologies’
meeting,
The
8 (A)
23-25
C-834
Reactivity
role in index
Mar 1986,
catalyst scaleadds
vacuum residue
accumula-
Bonaventure
further The desalting
density
However, if process
Reactivity
there comprises
is
a function
substantial the
and harness its various renewa- of density and C7 asphaltenes
Hotel Los Angeles, California. California.
other
lumping:crude
describingoils.
theTherefore
chemistry ofthe salt tion in heat exchanger
complex tubes due to following operations:
ESD
Specialized in
high pressure Customized
low/high temperature Valve Design
OHL Gutermuth
Helmershäuser Straße 9+12 · 63674 Altenstadt / Germany Industrial Valves GmbH
Phone +49 6047.8006-0 · Telefax +49 6047.8006-29
www.ohl-gutermuth.de · E-Mail: [email protected]
Product yield
9 AlHumaidan F, Hauser A, Al-Rabiah H, Pitch
Lababidi H, Bouresli R, Studies on thermal
cracking behavior of vacuum residues in Eureka
process, Fuel, 2013, 109, 635-646.
10 AlHumaidan F, Lababidi H M, Al-Rabiah H,
Thermal cracking kinetics of Kuwaiti vacuum
residues in Eureka process, Fuel, 2013, 103,
923-931.
11 Aiba T, Kaji H, Suzuki T, Wakamatsu T,
Residue Thermal-Cracking by the Eureka
11
Sa e 1
e
e
pl
pl
pl
pl
pl
pl
pl
pl
pl
e
Process, Chemical Engineering Progress, 1981,
pl
pl
m
m
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
Sa
77, 2, 37-44.
Sa
12 Takatsuka T, Kajiyama R, Hashimoto H,
Matsuo I, Miwa S, A practical model of thermal
Figure 9 Scale-up results obtained
cracking of residual oil, Journal of Chemical
Engineering of Japan, 1989, 22, 3, 304-310.
Glen Hay is Vice President of Business Lante Carbognani is a Key Research and Hideki Nagata is General Manager of the
Development with Virtual Materials Group Developer in the PIONA Reactors department operations management department at Fuji
Inc., Alberta, Calgary, Canada. His experience of Virtual Materials Group Inc. He has Oil Company Ltd. Chiba, Japan. His experience
is focused on reactors, heat transfer units, experience in process modelling, engineering and interests are focused on petroleum refining
and overall plant modelling and optimisation. design and R&D in refining processes. He technology and optimisation. He holds a
He holds a bachelor’s degree in chemical holds a BSc in chemical engineering from the bachelor’s degree in chemical engineering from
engineering from the University of Calgary and University Simon Bolivar, and a MSc in chemical the University of Kagoshima and a master’s in
a master’s in advanced process control. engineering from University of Calgary. reaction engineering.
A New Way
bolted bulk material silos
Johnson Screens offers a complete solution to your hydrocarbon processing industrial needs, including:
– Down Flow Reactor Internals – Hydrocarbon Refining Reactor Internals
– Radial Flow Reactor Internals – Petrochemical Processing Reactor Internals
– Installation of 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.
We look forward to providing the quality engineered products you have come to expect when specifying
Johnson Screens.
North & South America Asia Pacific Europe, Middle East & Africa
Phone +1 651 636 3900 Phone +61 7 3867 5555 Phone +33 5 4902 1600
Fax +1 651 638 3281 Fax +61 7 3265 2768 Fax +33 5 4902 1616
[email protected] [email protected] [email protected]
www.aqseptence.com
O
il refineries and petrochem- The combination of alkaline and
ical plants operate with acid cleaning is not the method
quite a large number of heat of choice when viscous, tenacious
exchangers. Fouling is an omni- materials such as heavy fuels or
present problem, causing signifi- Reactor Furnace additives from upstream processes
cant losses in energy recovery or need removing. High-performing
generating an increase in pressure aqueous cleaning programmes from
drop. Periodical cleaning is manda- the Kurita CD series can significantly
tory even if distillation equipment is reduce downtime with excellent
Feed/effluent
well designed. If fouling is observed, heat exchanger results. Metal surfaces heavily fouled
it can initiate a series of down- system with oils, grease, tars, waxes, fatty
stream problems. In many cases, 1 Packinox heat exchanger oils or other substances are cleaned
heat exchangers are taken out of ser- 2 Texas Tower heat exchanger or without the need for costly organic
vice because of severe pressure drop set of several heat exchangers solvents such as light cycle oil
increase, not reduced heat transfer. (LCO), diesel or white spirit. Water
Heat recovery is essential in pro- Figure 1 Feed/effluent heat exchanger is sufficient for the preparation of the
cess units which are operated with network chemical cleaning solution, which
reactors. Typical process units with is circulated with suitable equip-
feed/effluent exchangers installed procedures are sometimes still used ment. Heat exchanger networks are
are hydrotreaters, hydrocrackers, in oil refineries and petrochemical cleaned in a few hours, with deposits
catalytic reformers, paraffin dehy- plants to remove the organic portion and sticky layers dissolved into the
drogenation plants, paraxylene of deposits such as oil and grease, aqueous cleaning solution.
plants, and methanol synthesis followed by acid cleaning to dissolve Chemical cleaning programmes
plants. Feedstock is heated to a high scale and soften deposits. But the such as Kurita CD-9931 offer a num-
temperature to be vaporised and results with such classical cleaning ber of advantages when it comes to
reacted with a gas, which in most programmes are often only moder- cleaning complex heat exchanger
cases is hydrogen. The desired reac- ate, and further mechanical cleaning networks (see Figure 2):
tion products and unreacted gases is required with additional costs. • Significantly reduced downtime
are de-superheated and condensed for cleaning
for separation. Some process units • Less labour intensive work com-
still operate with heat exchanger net- Hot condensate, pared to mechanical cleaning
water, steam
works of at least six to 12 exchangers • Cleaning solution reaches inacces-
installed horizontally. More common Kurita CD sible areas
are Packinox plate heat exchang- Series • Metal surfaces are not mechani-
ers or Texas Tower tubular heat cally damaged
exchangers, which are installed verti- • Cleaning can be done in situ.
cally (see Figure 1).
Heat exchanger performance can Texas Tower and Packinox heat
be restored by traditional mechan- exchangers
ical cleaning methods, tailor-made Texas Tower tubular feed/efflu-
chemical cleaning programmes, or ent heat exchangers are installed
a combination of both. Mechanical Mixing and in an upright position, with two
cleaning of complex heat exchanger circulation tank exchangers usually connected in
networks using high pressure clean- series to provide good heat recovery.
ing technologies can take several Figure 2 Chemical cleaning of heat Generally, such heat exchangers do
days. Classical alkaline cleaning exchanger network not show significant pressure drop
Further reading
1 Speight J G, The Chemistry and Technology of Petroleum, Fourth
Edition, 2006, CRC Press, ISBN 0-8493-9067-2.
2 Jafari M, Iranshahi D, Rahimpour M R, Progress in catalytic naphtha
reforming process: a review, Applied Energy, 2013.
3 Aitani A, Catalytic naphtha reforming, Encyclopedia of Chemical
Processing, 2006.
www.eptq.com
nalcochampion.com
T
he Preem refinery in Lysekil preheat crude from 189°C to 208°C, technology for this revamp of the
is a modern refinery of high recovering 8.5 MW from the bot- crude preheat train.
complexity, located on the tom reflux. The heat transfer area As the Compabloc exchanger
west coast of Sweden. In 2010, the was 308 m2, the diameter was would recover so much more
company conducted a refinery- 950 mm, and the tube length was energy from the bottom reflux into
wide energy analysis study, cov- 7315 mm. the crude, one downstream shell
ering among other process units a The new heat recovery service and tube heat exchanger could be
11 milllion t/y integrated atmos- would instead recover 21.2 MW removed from the preheat train,
pheric and vacuum distillation unit, from the bottom reflux, preheating and the new Compabloc with a plot
with two parallel trains for crude the crude from 192°C to 234°C. space of 2.0m x 2.0m was installed
preheating. This study concluded As in all other refineries, the in its place (see Figures 1 to 3).
that by increasing the recovery of Preem crude preheat train does not In addition, another of the down-
energy from the column bottom have a lot of plot space available stream services, recovering energy
reflux into crude, Preem would not for adding/increasing heat trans- from heavy vacuum gasoil into
only improve the energy efficiency fer area for more energy recov- crude in two shell and tube exchang-
by preheating the crude to a higher ery. Instead, the refiner decided ers in series, could be moved to the
temperature before the furnace, but to go for an unconventional heat preheat train A, providing additional
would also offload the top section of transfer solution, using a high effi- preheating in this line.
the distillation column and thereby ciency compact welded plate heat Finally, in order to maximise the
reduce the overhead pressure drop. exchanger called Compabloc. reliability of the new preheat train
This latter parameter means that In Preem Lysekil refinery, design, the original shell and tube
the refiner could reduce the flash Compabloc exchangers had already heat exchanger recovering energy
zone pressure in the column and been in service as lean/rich amine from the bottom reflux for crude
thereby increase the yield of heavy interchangers in the amine regener- preheating was kept on standby for
gasoil over light vacuum gasoil, a ator system and sulphur recovery the new Compabloc exchanger.
more valuable product that would unit since 2006 and 2008 respec- The Compabloc exchanger was
increase the refinery’s profitability. tively. Because the exchangers were installed at the end of 2013 and,
The original shell and tube performing well, Preem had no since the beginning of 2014, it has
heat exchanger was designed to doubts other than to use the same been operating at full capacity. In
HVGO
3A2
Desalter
1A 2A 3A1 4A 5A 6A 7A 8A 9A 10A
CDU Naphtha Naphtha LGO Kero HGO AR LVGO VR LGO
OVHD PA PA PA quench
Desalter
1B 2B 3B 4B 5B Compabloc
CDU Naphtha Kero Kero Kero
OVHD PA PA
6B 7B 8B 9B 10B
HGO HVGO VR HVGO
quench PA
1A 2A 3A1 4A 5A 6A 7A 8A 9A 10A
CDU Naphtha Naphtha LGO Kero HGO AR LVGO VR LGO
OVHD PA PA PA quench
Desalter 7B
133ºC 165ºC
1B 2B 3B 5B 4B Compabloc HVGO
CDU Naphtha Kero Kero Kero
OVHD PA PA 372ºC
220ºC 249ºC 270ºC
6B 9B 8B
HGO HVGO VR
PA quench
Figure 2 Layout of crude preheat trains A&B post desalters after revamp
February 2017, it was taken out of which can be explained by the lower easily be taken out of operation for
service for a first internal inspection, crude inlet temperature, the lower cleaning or inspection, without hav-
as per Preem’s safety regulations. bottom reflux to crude flow ratio ing to interrupt the operation of the
Looking at the average operat- and the overall lower flow rates, plant.
ing data for these first 34 months reducing the heat transfer efficiency Figure 4 shows the fouling ten-
of operation, it can be seen that the of the exchanger slightly. dency of the shell and tube
performance is very stable without During the period when the exchanger in comparison to the
any tendency to fouling. Compabloc was taken out of ser- fouling tendency of the Compabloc
Table 1 shows that the average vice for inspection, the original shell exchanger, when in operation.
operating performance in terms of and tube heat exchanger was put Fouling in crude heat exchangers
heat duty and logarithmic mean back into service. This was another is very much related to crude tem-
temperature difference (LMTD) is improvement made to the pre- perature and crude shear stress in
very similar to expected design val- heat train during the revamp. By the heat transfer channel. The higher
ues. The crude outlet temperature installing the new exchanger with the crude temperature, the higher
is 10°C lower than design values, block valves on both sides, it can the fouling tendency, and the higher
the shear stress in the heat trans-
fer channels, the lower the fouling
tendency.
As Figure 4 shows, the Compabloc
exchanger suffers less from fouling
than the original shell and tube heat
exchanger, even though the crude
outlet temperature from the lat-
ter was more than 20°C lower than
from the Compabloc exchanger and
the shell and tube exchanger oper-
ated with a slightly higher (5%)
crude flow rate.
The lower fouling tendency of the
Figure 3 Compabloc installed in Preem’s CDU preheat train B, and outer dimensions as Compabloc exchanger was also con-
per the drawing firmed during internal inspection
of the heat transfer channels, when
Design data vs average operating data Preem was expecting to find more
fouling material.
Design Operating (Apr 2014 – Feb 2017) In summary, the main advantages
Bottom reflux flow, kg/h 484 645 401 016* (455.7) of the new exchanger in the Preem
Crude flow, kg/h 696 763 604 620* (671.8) crude preheat train are:
TG in/out temp, °C 265/207 264.6/191.8
Crude in/out temp, °C 192/234 170.8/224.3 • High heat transfer efficiency: one
Heat duty,** MW 21.2 22.5 Compabloc replaced four shell and
LMTD, °C 21.4 29.6 tube exchangers and provided in
* Mass flow rates are calculated based on measured volumetric flows (m3/h, in brackets) and fluid density at 15°C. total 7 MW more than the original
** Heat duty is calculated based on operating data and fluid specific heat at average operating temperature. exchangers.
• Compact design: one Compabloc
Table 1 recovering 21.2 MW into crude pre-
Visit DigitalRefin
ing.com to see
what content is
relevant to you
A
new energy efficient method
for processing the aro-
matics mixture to pro- Cooling
water (CW)
duce pure benzene, toluene, and Clay Air cooler
Condenser
treating
xylene (BTX) in an aromatics pro- tower Vessel Vessel
duction plant was conceptual-
ised for a significant increase in
Steam /
process energy efficiency and hot oil Reflux Reflux
financial savings. A techno- Feed Pump Pump
economic study was carried out to heater
Benzene
estimate the value of energy sav- Toluene
Distillation
ings, capital investment and addi- EX-1 column I
(DCI) Cooling
tional capital requirement for water (WC)
Toluene Distillation
revamping the existing design of + Xylene column II (DCII)
an aromatics production plant.
The results of the study show large MP MP
steam steam
scale savings in hot utility, cold
Reboiler Reboiler
utility, electricity and overall oper- Extract
ating costs as a result of implement- hydrocarbons Cooler
ing the new method. The payback from SRC Xylene
overhead Toluene + Xylene Pump
period for additional expenditure Feed Cooling
required to install heat exchange pump water (WC)
area piping modifications would
be less than a year, while pro-
cess emissions of CO2 would fall
significantly. Figure 1 Schematic of a conventional process for BTX processing
BTX is produced on a large
scale by processing the reformate grassroots unit for BTX production Conventional process
and pyrolysis gasoline streams in seems very promising. However, A schematic of the conventional
an aromatics production plant. there is a need for 110 trays for BTX process used for processing the
The aromatics rich stream is first fractionation. Therefore, the appli- aromatics mixture (BTX mixture)
processed in the extraction sec- cation of DWC for revamping an obtained from a solvent recov-
tion to remove non-aromatics. existing two column distillation ery column is shown in Figure 1. A
Subsequently, the aromatics stream process does not seem feasible. hydrocarbon stream is heated in the
is further treated in a clay tower A new method for BTX process- heat exchanger with the clay treat-
to remove olefinic impurities, fol- ing was conceptualised by opti- ing tower bottoms, and then with
lowed by two distillation columns mising the operating conditions steam to the desired temperature,
to produce pure benzene and tolu- and adopting new heat integration before entering the clay tower for
ene. Nowadays, divided wall col- for the revamp of a conventional removal of olefinic impurities.
umn (DWC) technology, which two column process to deliver sub- The clay treated BTX mixture
claims 30% energy and capital sav- stantial energy savings and reduce is processed in a benzene separa-
ings over the conventional two col- the process’s carbon footprint. A tion column (DCI) and toluene-xy-
umn arrangement, seems to be the techno-economic analysis was car- lene separation column (DCII) to
upcoming approach for fractiona- ried out to estimate the quantita- obtain pure BTX (see Figure 1). The
tion of a BTX mixture in the aromat- tive benefits of the proposed scheme xylene stream, which is generally
ics plant.1,2 A DWC application in a over a conventional scheme. processed downstream, is either
BE A SMOOTH OPERATOR.
plant-wide optimization.
www.eptq.com
Europe, Africa, Russia Middle East and India North and South America Asia Pacific
Sulzer Chemtech Ltd. Sulzer Chemtech Middle East SPC Sulzer Chemtech USA, Inc. Sulzer Shanghai Engin. & Mach. Works Ltd.
Neuwiesenstrasse 15 3rd Floor Jawhara Plaza 1 Sulzer Way No.1688, Feizhou Road, Nanhui District
8401 Winterthur P.O. Box: 21558 Al Spef Tulsa, OK 74131 Shanghai, 201306
Switzerland Kingdom of Bahrain USA China
Phone: 0041 52 262 50 28 Phone: 00973 1756 8400 Phone: 001 918 446 6672 Phone: 0086 21 3807 1000
[email protected] [email protected] [email protected] [email protected]
Achieving the right sulphur plant capacity and configuration is crucial to the
proper operation of a refinery
R
efiners around the world are
processing a wider crude slate. 34.0 1.3
API
They are also faced with the Sulphur
33.5
challenge of meeting a tighter sul-
1.2
phur specification for the products,
Sulphur, wt%
33.0
hence the role of the sulphur plant
and its design are gaining more
API
32.5 1.1
importance.
A proper sulphur plant configura- 32.0
tion needs to cater to the operational 1.0
flexibility of the refinery. This needs 31.5
to be decided based not only on the
refinery turndown operation but 31.0 0.9
also on the various crude assays that 2008 2010 2015 2020 2025 2030
are being handled.
The best sulphur plant configura- Figure 1 Global crude oil quality trends
tion avoids bottlenecks in refinery
operation with minimum economic mental requirements are gradually International Maritime Organisation
investment. This article will pres- becoming more stringent. (IMO) to limit the sulphur content
ent several case studies on achieving The average API gravity and sul- in marine fuel to 0.5% by the start
the optimum sulphur plant config- phur content of refinery crude slate of 2020, whereas the same limit for
uration with varying crude com- varies by region, and over time the the Sulphur Emission Control Areas
positions, refinery configurations average quality of global crude slate (SECA) is already 0.1% from the
and capacities, and provide generic has been declining gradually. Figure start of 2015 (see Figure 2). This sul-
guidelines for making decisions 1 shows how crude slate is chang- phur containment will have a huge
related to sulphur plant design dur- ing and is expected to change over impact on the global refining sector,
ing the refinery configuration phase. the next few years. This confirms requiring major changes in refinery
The feasibility of a petroleum that total world crude oil reserves configuration and operation. New
refinery depends on the inherent (crude oil of the future) are of lower process units will be required to
interaction between the choice of API gravity and higher sulphur upgrade and/or desulphurise the
crude oil used and the desired type content than current world pro- residue streams.
and quality of finished products to duction (crude oil being processed The impact on overall sulphur
be generated. Using a more expen- presently). The harsh impact of the management in the refinery will
sive light, sweet crude oil requires changing nature of crude oil has also become vital, based on the
a simpler refinery configuration been moderated due to the effect of modifications made on the over-
and hence a lower capital invest- an increased usage of natural gas all refinery configuration and
ment. But supplies of light, sweet liquids and other unconventional operation. The sulphur plant con-
crude oil are decreasing, and at the crude oil sources, but the graph still figuration needs to be optimised
same time the gap between heavier shows that the change in sulphur to limit any unnecessary impact
and sourer crude prices is increas- content in crude is more prominent on the total investment cost of the
ing. Refinery configurations are also than crude complexity. project without sacrificing flexi-
inherently defined by the type of Simultaneously, product spec- bility in the overall refinery oper-
products expected and the quality ifications with respect to sulphur ation. The sulphur recovery unit
to be achieved. Meeting the quality content are becoming increasingly (SRU) design should avoid bottle-
specifications of the final products more stringent. The most recent necks with respect to the crude slate
is of utmost importance as environ- is the decision adopted by the to be processed, the product spec-
Table 3
Conclusions
Refiners around the world are feeling the crunch of
meeting tighter sulphur specifications for products
and at the same time facing the challenge of gradually
depleting reserves of sweet crude. While modifications
are being made to process units, in particular develop-
ments in hydrotreating processes and catalysts, the sul-
phur load in the refinery is continually increasing. This
calls for an update of the sulphur balance of the refin- More than just
ery and an upgrade of the sulphur plant configurations
in most refineries around the world. vacuum
Arriving at the proper sulphur plant capacity and
configuration is of utmost importance for the proper Process-integrated solutions
operation of a refinery. The plant capacity and config- for many types of vacuum
uration should take into account the ‘most sour’ crude
system
assay/blend to be processed in the refinery. It should
also consider adding margins to meet the required
operating flexibility in design with respect to overall GEA supplies steam jet vacuum systems
operation. This should not induce unnecessary capital and hybrid vacuum pumps, optimizing
investment, and hence these margins need to be opti- production processes to reduce costs and
mised. For revamp scenarios, it is in many cases advis- environmental pollution. You can rely on
able to look for opportunities in upscaling the sulphur 90 years of experience and thousands of
plant capacity by modifying the sulphur recovery pro- references in numerous industrial sectors
cess to improve overall sulphur recovery and increase all over the world. And thousands of
sulphur handling capabilities. satisfied customers can‘t be wrong.
Proper balance of refinery sulphur followed by a judi- We‘d like to prove it to you also.
cious selection of sulphur plant capacity and train con- Contact us and we will show
figuration provides a smoother path for overall design, you that we are the right
www.eptq.com
UOP-Callidus-RSR-Ready-Flares-8009-297x210mm.indd
uop callidus.indd 1 1 12/14/17 12:21
15/12/2017 PM
16:38
Replacing a corroded column with
packing internals
A corroded column was replaced with a new column equipped with high capacity
packing. The cross sectional area is less than half that of the original
A
direct contact after cooler
Main specifications of the new
(DCAC) column equipped Cooled air
DCAC column
with sieve trays in an air sep-
aration unit (ASU) was severely New column Old column
corroded and had to be replaced. Height (T/T), m 14 14
Rather than go for a one to one Chilled water
Inner diameter, m 2.7 4.0
replacement, a modern packing Packing type MellapakPlus Sieve tray
252.Y
solution was selected based on a
lower pressure drop and much Table 1
higher capacity. During the process
design stage, the packing height The proposed operating condi-
specified was confirmed to be ade- tions and process flows for the new
quate to meet the heat transfer Cooling water column were unchanged. Based on
requirement. During the detailed the stream data and packing type,
engineering phase, computational the capacity and the pressure drop
fluid dynamics (CFD) simulations were evaluated, with the results
were utilised to assess the accepta- listed in Table 2. It can be seen from
bility of the existing inlet arrange- Hot air the packing hydraulics that the new
ment. Trial runs were conducted column was not highly loaded.
shortly after the new column was The key function of the DCAC
erected. The measured temperature Water column is to achieve the required
of the air from the top of the new temperature specification of the
DCAC column matched the design cooled air stream. A temperature
value very well. Figure 1 Sketch of a DCAC column outside this range will negatively
impact the vital downstream molec-
Background a severely corroded DCAC col- ular sieve adsorbers’ ability to con-
In an ASU, compressed air is umn. The existing column was dition the incoming air stream.
brought into contact with chilled 4 m in diameter and 14 m T/T, and Table 3 summarises the top pressure
and cooling water in the DCAC. was equipped with 10 sieve trays. of the DCAC column and the tem-
The primary function of the DCAC The dimensions of the new column peratures of various feeds. Based
column is to cool the hot air and were specified as in Table 1. on past operation of the old col-
reduce moisture.1 As the water Besides, a detailed design of the umn, the best ever achieved temper-
vapour content in compressed air air inlet was provided in the specifi- ature of cooled air from the top of
is linked directly to temperature, cation sheet: an open pipe type with the DCAC was 15.5°C. Sulzer was
the compressed air must be cooled a disk beneath the downward open- asked to evaluate the specifications
to 8°C~15°C. Otherwise, the down- ing to direct the incoming vapour and guarantee an approach tem-
stream molecular sieve adsorbers upwards. perature (the temperature differ-
may be laden with water in prefer-
ence to CO2. If CO2 is not removed Packing hydraulics
from compressed air, it can prefer-
entially freeze and cause plugging
of downstream equipment. A sim- Upper section Lower section
Top Bottom Top Bottom
ple sketch of a DCAC column is Capacity, % 51.0 51.9 65.3 71.5
shown in Figure 1. ∆P, mbar/m 1.2 1.2 1.4 1.7
An operator in the Asia Pacific
region approached Sulzer to replace Table 2
Inlet piping
For this ASU plant, one of the two
feeds was supplied by a compres-
sor; the other feed came from a
nearby plant. The feeds were com-
bined in front of the DCAC col-
umn, and the distance between the
joint of the two inlet pipes and the
DCAC column was about four times
that of the diameter of the air inlet
to the column. Common practice in
industry is that, in order to avoid
maldistribution of vapour inside a
column, any unavoidable upstream
piping disturbance should be at
least 10 times that of the inlet size
in distance from the column. In the
existing tray design, maldistribu-
tion of vapour would likely be mit- Figure 4 Vapour flow trajectories in the DCAC column and its external inlet pipes
HIGH-PERFORMANCE
EMISSION CONTROLS –
FOR THE LOWEST LEVELS
OF CO AND NOX AS
LOW AS 2PPM.
COMPLETE SYSTEMS
CUSTOM DESIGNED
FOR THE MOST
STRINGENT CUSTOMER
SPECIFICATIONS.
RENTECH BOILERS.
AS TOUGH AS TEXAS.
PRODUCTS SOLUTIONS
Heat Recovery Steam Generators Refining and Petrochemical
Waste Heat Boilers Power Generation
Fired Packaged Watertube Boilers Manufacturing & Institutional
Specialty Boilers Food Processing
CHP
WWW.RENTECHBOILERS.COM
Conclusions
A corroded DCAC column was
replaced with a new packed col-
umn. Due to the high capacity of
packing, the cross sectional area
of the new column was reduced
by more than 50%. Revamp of the
external inlet piping was eliminated
after using an advanced gas inlet
device. The operating temperature ACCURATE • FAST • CUSTOMIZABLE
of the cooled air matches prediction.
TM
Yang Quan is a Sulzer Senior Technical Expert, • Share data with HTRI Xchanger Suite®
specialising in mass transfer components,
such as trays, packing, column internals
and separators. He is also responsible for CONTACT HTRI FOR
air separation knowledge management
MORE INFORMATION.
within Sulzer Chemtech. He holds a PhD in
environmental engineering from the National
www.htri.net
University of Singapore.
www.jmprotech.com
LOWELL T PLESS
Tracerco
R
efinery and chemical plant be obtained for a gamma scan of a
operations depend heavily packed tower includes:
on distillation and separation • Condition of packed beds –
towers. Tower gamma scanning elevation of packing, depth of pack-
is well established in the process ing elements, uniformity of packing
industries as a qualitative tool to A B C
elements, and so on
help troubleshoot towers. Advances • Collector and distributor liquid
in data analysis have led to a quan- Figure 1 Typical gamma scanline levels: Damaged? Overflowing?
titative approach in expressing orientations: A for diameters 1.5 m and • Base or bottoms liquid level
gamma scan data in numerical larger; B for diameters 0.5-1.5 m; C for • Flooding present?
terms easily understood by process diameters less than 0.5 m • Foaming present?
and operations engineers. • Excess liquid retained in packing?
For packed towers, a grid scan of tion or liquid volume fraction. If • Liquid maldistribution?
three or four equidistant scans cross- the measured liquid retention den- The purpose of doing multiple
ing through the beds of packing sity is divided by the process liquid scans of a packed tower, as shown
would typically be used to investi- density at bed conditions (the liq- in Figure 1, is to determine the state
gate the quality of liquid distribu- uid density at the actual operating of the liquid distribution. The foun-
tion. The conventional approach to temperature and pressure), liquid dation for this analysis is the Beer-
‘analysing’ a gamma scan has been hold-up or liquid volume fraction Lambert law:
to visualise how well the scan data can be established. A comparison
I=Io e-ρµχ
from the individual scans matched of the liquid hold-up fraction to the
each other or how well they ‘over- packing operating capacity curves where
laid’ with each other. This is a totally provides an objective appraisal of Io is the initial gamma ray intensity
subjective analysis lacking consist- current operating capacity. measured at a given distance with
ency, open to varying interpretation It is always easier to understand no material interfering with the
and does not translate well from and discuss technical issues when radiation transmission
tower to tower. Therefore, the result- quantitative information can be I is the radiation intensity (counts)
ing conclusions from this approach used to compare operational param- measurement after passing through
can be very ambiguous regarding eters with engineering design. This the tower
the magnitude of any detected liq- advanced analysis provides a new µ is the absorption coefficient of the
uid maldistribution. method for extracting quantitative material the radiation is passing
An alternative analytical approach, information from gamma scan data through (material physical property)
termed PackView, has been devel- to diagnose and characterise the ρ is the density of the material the
oped whereby a relative density operation of distillation and sep- radiation is passing through
scale is calculated from data that the aration towers. It is our goal that χ is the thickness of the material the
grid scan provides. The density scale the use of the advanced analyt- radiation is passing through.
begins at the density of the dry or ics presented improves the value Beyond questions concerning
non-operating packing. The density of gamma scan data and facilitates damage to internals and flooding
scale displays the calculated density improvements in the operation of within a packed tower, the next
of liquid retained in the bed of pack- mass transfer equipment. biggest concern is the state of liq-
ing based on the scan data results. Figure 1 shows the most typ- uid distribution through the beds.
Another calculation by which to ical orientations for conduct- Historically, gamma scan analy-
put the liquid distribution into per- ing a gamma scan on a packed sis has relied upon performing two
spective and to get a measure on the tower. These scans are conven- sets of parallel scanlines (given the
useful capacity of the packing is to tionally called grid scans. Some of tower diameter is sufficiently large)
calculate the liquid hold-up frac- the process information that can through the packing (see Figure 1a),
PTQ Q4 2017 49
70 PTQ Q2 2018 www.eptq.com
12/09/2017 16:51
q2 tracerco.indd 2 13/03/2018 10:46
Figure 3 (A) Initial gamma scan results from small diameter column (B) Gamma scan results, enhanced with liquid retention scale,
showing missing packing at top of bed
on tower dimensions, scan path the bed of packing was supposed be nearly equal to the dry packing,
length (variable χ) and gamma to be as per the reference tower either there was no liquid (obvi-
scan response through the pack- drawing. There was a reduction ously not the case) or packing was
ing. The results are used to super- in radiation counts from the clear missing. This was a bed of random
impose a density scale onto the vapour region at the expected ele- packing. The PackView analysis
scan data. The baseline of the vation for the top of the bed. After provided conclusive evidence that
density scale is the dry bulk or a short distance into the bed the portions of the packing were miss-
non-operating packing den- radiation counts decreased. So the ing. The high density in the bot-
sity. Calculated densities greater question was, what was the operat- tom of the bed was likely crushed
than the dry packing density ing condition of the bed? Were the packing retaining an excess of liq-
represent liquid and/or solids lower counts (higher density) liq- uid. Eventually, entry into the tower
retained in the bed of packing. uid hold-up (flooding) at the bottom confirmed these results.
Figures 3b, 4b and 5b show exam- of the bed? Or was the radiation
ples of this liquid retention density count at the bottom of the bed ‘nor- PackView Example 2
scale. mal’ and something had happened Figure 4a shows the gamma scan
As with the qualitative gamma to the packing in the top of the results from a crude vacuum tower
scan analysis, if the multiple scan- bed? A visual or qualitative evalu- (the fourth scanline was not per-
lines have matching liquid reten- ation of the radiation counts could formed due to limited access). One
tion densities then the implication not answer these questions with of the biggest problems with crude
is the liquid distribution is good. confidence. vacuum tower operation is manag-
However, if there is a difference Figure 3b shows the gamma scan ing the coking (fouling) of the wash
between the scanlines, the reten- results from Figure 3a with the liq- bed. In this example, plant staff sus-
tion density gives a numerical com- uid retention scale added. Note that pected the wash bed had coked up
parison from which to gauge the the density of the dry packing is 160 so the gamma scan was carried out
extent or severity of any liquid kg/m3 and the top of the bed has to assess the situation. As shown in
maldistribution in terms of liquid a density essentially equal to this. the tower diagram on the right side
density. However, the tower was operat- of Figure 4a, this wash bed consisted
ing and there was liquid travelling of two different types of packing.
PackView Example 1 down through the packing. Where The top section was typical struc-
Figure 3a shows the gamma scan is the additional density from the tured packing while the bottom sec-
results from a small-diameter retained liquid? The overall density tion was grid packing. The scan
packed tower (only two scanlines at the top of the packing should be results showed the bottom section
performed due to the small diam- the combination of retained liquid to be very dense, as the radiation
eter). The tower diagram on the and packing density. Since the scan intensity was less than five counts,
right side of Figure 3a shows where results show the overall density to essentially background (no radiation
passing through the tower). Based results from Figure 4a with the liq- the grid was coked and/or flooding.
on this result, the diagnosis was that uid retention scale added. The dry Therefore, the scan radiation source
the grid section was coked and/or density of the grid packing was 255 used on this tower was more than
flooding with liquid, but plant man- kg/m3. The density of the process adequate. The advanced analysis
agement was not convinced. Grid material inside the grid packing, proved that if this wash bed had not
packing is dense so a question was based on the scan results, was cal- been fouled with coke and flooding
asked concerning whether the radi- culated to be approximately 300 kg/ there should have been 30-50 counts
ation source was too small. In other m3. Grid packing is very open pack- passing through the tower, rather
words, some radiation would pass ing and the typical liquid rate on a than less than five.
through the packing from using a vacuum wash bed is very low. A An additional computation con-
larger radiation source and then per- typical liquid retention density for firmed this diagnosis. The liquid
haps the tower would not appear to grid packing in this service is 80-100 volume fraction was calculated by
be flooding. kg/m3. The density calculated from dividing the liquid retention den-
Figure 4b shows the gamma scan the scan data was far above this, so sity of 300 kg/m3 by the process
liquid density (800 kg/m3). The
resulting liquid volume fraction was
0.20
0.38. Figure 5 shows typical capac-
0.18 ity curves from studies of packing
0.16 capacities. Note that packings usu-
Liquid fraction hold-up
maldistribution. The typical quali- fraction calculated, based on a pro- only general guidelines and may
tative evaluation appeared to show cess liquid density of 800 kg/m3, differ based on the specific service
there was a problem in the bed of ranged from 0.01 (very low liquid the tower performs and the overall
packing. The four scanlines did not fraction) to 0.16 (bordering on flood- liquid operating rate.
match each other very well, indicat- ing). Thus the advanced analysis
ing that the overall density between conclusively proved that the bed Conclusions
the four sets of data were different of packing was suffering from very • Using PackView advanced anal-
from each other. The radiation inten- poor liquid distribution. ysis, gamma scan data is able to
sity counts varied from 1800-3600 Figure 7 shows general guidelines measure (quantify) liquid maldis-
as read from the horizontal scale of for the magnitude of liquid reten- tribution through a packed bed in
Figure 6a. But what was the severity tion density differences seen across terms of liquid density.
of the liquid maldistribution? The packed beds from gamma scans and • The liquid retention in packing
radiation intensity or counts gave the relative quality of liquid distri- can be measured from gamma scans,
no perspective or evaluation of the bution. Please note that these are indicating excess liquid hold-up
actual liquid distribution quality. caused by fouling or flooding.
Figure 6b shows the same gamma • If the actual liquid density is
scan results from Figure 6a with 30 kg/m3 (2 lb/ft3) known, a simple calculation can
the liquid retention density scale determine the % liquid fraction, a
added. The dry packing density is Good liquid distribution measure of the maximum useful
∆ = 0−30 kg/m3 (0−2 lb/ft3)
115 kg/m3. The PackView analysis capacity of the packing.
showed the liquid retention ranged
PackView™ is a trademark of Tracerco.
from 10-130 kg/m3. In terms of liq- Fair liquid distribution
uid density across a bed of packing, ∆ = 30−80 kg/m3 (2−5 lb/ft3) Lowell Pless is Business Development
this was a large density differ- Manager - Distillation Applications with
ence. Reinforcing the liquid mald- Poor liquid distribution Tracerco, located in Pasadena (Houston),
istribution diagnosis was the fact ∆ => 80 kg/m3 (>5 lb/ft3) Texas. He has been applying radioisotope
(worrisome) techniques in process troubleshooting for
that one scanline (the blue data
30 years, originally with Tru-Tec Services
curve in Figure 6b) had almost the 80 kg/m3 (5 lb/ft3)
(acquired by Tracerco in 2006). He holds a
same overall density as dry pack- BS degree in chemical engineering from the
ing alone. Thus, this scanline was Figure 7 General guidelines for evaluating University of Texas at Austin, is a registered
nearly dry or there was very lit- liquid distribution quality based on gamma Professional Engineer in the State of Texas,
tle liquid on that side of the scrub- scan results using the advanced analytics and participates on the Design and Practices
ber. Furthermore, the liquid volume of the liquid retention scale committee for Fractionation Research (FRI).
www.jonellinc.com
Uniting two unit offerings into one column shell can lead refiners to improve the
profitability of their gas plants
D
ividing wall column (DWC) high operating pressures and pro-
technology provides refiners pane losses in the off-gas. In some
Off gas
with a unique way of improv- (C1−C4, heavies) processes, these losses are mitigated
ing the efficiency and profitability Fuel gas (C1, C2) by using refrigeration in the over-
of traditional refining techniques. head system. This increases the cap-
This technology, which has been in ital cost of the process.
use since the 1940s, can be modi- LPG These problems can be recti-
fied for various applications ranging (C3) fied by using a top DWC. Doing so
from naphtha/reformate splitters allows the top section of the column
to gas plants, and can even be used to be operated under two differ-
to transform established technolo- ent unit operations: absorption and
gies like isomerisation and naphtha distillation. The section of the col-
hydrotreating. C4,
umn where the feed enters operates
In this article, we discuss the dis- heavies under reboiler absorption. The other
Deethaniser Depropaniser
tinct advantages that DWCs can section of the column separates a
offer to improve and modify tradi- light liquid product and a heavy
tional gas plant process schemes, Figure 1 Conventional LPG recovery unit bottom product using regular distil-
with emphasis on top DWCs. DWCs lation. This arrangement offers the
are often characterised by the pres- ate in a sequence and under similar following advantages.
ence of a vertical wall in a regular operating pressures and temper- First, use of absorption allows the
distillation column. When the dis- atures. The DWC will potentially column to be operated at a lower
tinctive wall is present at the top combine the operation of two or pressure. The absorbing solvent
of the column, the column can be more columns, thereby minimising captures the valuable C3-C4 com-
referred to as a top DWC. both capital and energy consump- ponents in the off-gas and moves
Top DWCs offer a host of advan- tion of the whole sequence. them towards the bottom of the col-
tages, especially in gas plants umn. These components are then
which are generally operated at Uniting wall columns in LPG concentrated on the other side of
high pressures. Unlike a conven- recovery from off-gas the DWC using distillation. For
tional distillation column, the top Liquefied petroleum gas (LPG) feeds which contain a good amount
of these columns is segregated recovery units are an integral part of C5 and heavier components, a
into two separation zones within of any refinery. In general, in an portion of the heavy bottom prod-
the same shell with no intermixing LPG unit two or more columns uct can be used as the absorption
between the two zones. As a result, remove the C1-C2 components along medium (see Figure 2a). Using the
they can be operated in separate with other non-condensable gases, bottoms is highly beneficial when
modes and basically behave as two while concentrating the C3-C4 com- the feed contains at least 10 wt% of
independent columns. Absorption ponents to generate LPG. Figure 1 C5s. Table 1 shows a sample GT-LPG
and distillation can be carried out shows a standard design consist- Max feed.1 If the C5 concentration in
in a single column. This principle ing of a deethaniser column fol- the feed is low, the bottoms solvent
forms the basis of uniting two unit lowed by a depropaniser column. can be supplemented by an addi-
operations into a single column The deethaniser column operates at tional heavy oil solvent from nearby
shell known as the GTC Uniting a high pressure, wherein the C1-C2 processes.
Wall Column (GT-UWC). components are removed as the A water-cooled partial condensa-
The concept of top DWCs can also top product. The C3 rich (or C3-C4) tion is used in the off-gas side of the
be implemented in a reformer unit stream is removed in the next col- column. This is highly economical
gas plant and an FCC unit gas plant, umn. The recovery of LPG in this compared to using refrigeration to
wherein two or more columns oper- design is generally poor due to the mitigate C3-C4 losses.
Table 2
Uniting wall columns in reformer umn. The C4 and lighter cut is then
unit gas plants processed in a deethaniser column,
A typical reformer unit gas plant which removes an LPG cut at the
Figure 2(a) GT-UWC process scheme usually has a depentaniser column, bottom.
scheme; (b) Inside view of a uniting wall followed by a debutaniser and a The three columns in this
column deethaniser. In Figure 3, the depen- sequence operate at high pres-
taniser column removes the C5- cut sures. Additionally, the debutan-
Additionally, this DWC design from the reformer bottoms as the iser and deethaniser columns see
sees a reduction in energy consump- top product. C6 and heavier cuts are partial overhead condensation.
tion by about 10-30% compared removed as the bottom product. In Due to these, a major portion of the
to traditional columns with better the next column, the C5s are con- LPG components (C3-C4) end up
product recoveries (see Table 2). centrated in the bottom of the col- in the off-gas. These components
can be recovered by use of a DWC
configuration.
Fuel gas
As Figure 4 shows, the operation
of the deethaniser can be squeezed
into the bigger depentaniser col-
umn by means of a dividing wall.
The column has two distinct zones
Feed from within: a pre-fractionation zone
reformer and a main fractionation zone. The
pre-fractionation zone sees a com-
Debutaniser Deethaniser bination of absorption and distil-
lation. The lights are separated on
Depentaniser this side as the off-gas, with the
absorption medium reducing the
liquid losses. On the main fraction-
C5 cut LPG
C6+ to ation side, the C5 and lighter com-
reformate splitter
ponents are removed as the other
top product. Similar to a GT-UWC
b
Figure 3 Conventional reformer unit gas plant configuration design, a portion of the heavy
EVOLVING NEEDS.
SOLVING.
M O R E S O L U TIO N S .
M O R E R & D.
M O R E E FFIC IE N CY.
M O R E C O ST S AV IN G S .
CHANGE IS COMING.
W W W . C R I C A T A LY S T. C O M
ACHEMA
Made
Hall 9.1
Your Specialist for in
Germany
Stand C4
PROCESS SAFETY in the
OIL & GAS INDUSTRY
© REMBE | All rights reserved
worldwide
RAHUL C PATIL, PINAKIRANJAN PATRA, AJAY GUPTA, VED PRAKASH MISHRA and ASIT KUMAR DAS
Reliance Industries
A
delayed coker in a petroleum increase the run length of the coker
refinery processes vacuum Flue gases heater so as to increase the produc-
residue from the vacuum tivity of the delayed coker unit as
distillation unit as feed and ther- VR inlet well as the life of the coker heater
mally cracks it into useful products Convection tubes.
including liquid petroleum gas, section A coker heater (see Figure 1) con-
naphtha, gasoline, diesel, heavy gas sists of horizontal tubes where feed
oil, and petcoke. While it is desir- conventionally enters from the con-
able to have the majority of the vection section to the radiant section
cracking and coking taking place of the heater (a down-flow config-
in the coker drum, a small amount uration). The outlet temperature of
of these reactions occurring inside the heater, the COT, is measured at
the heater tubes is inevitable. The the radiant section outlet. The burn-
coke so produced inside the heater Radiant ers are normally floor mounted at
tubes provides additional resist- section the bottom of the radiant section of
ance to heat transfer between the the heater where they fire fuel with
flue gas and the process fluid. Since air. The radiative heat from the
the coil outlet temperature (COT) combustion of fuel gas in the radi-
is to be kept constant, one may end ant section is transferred to vac-
up by firing more fuel to achieve uum residue from the convection
VR outlet
the desired COT, thereby increas- Burners section. The remaining heat from
ing the tube skin temperature or the combustion gases is transferred
tube metal temperature (TMT) due to preheat vacuum residue in the
to the additional resistance offered Air and fuel convection section. In the process,
by coke. Coke depositing on heater vacuum residue cracks into lighter
tubes thus usually limits the run components. As pressure reduces
length by limiting the TMT which a Figure 1 Schematic of delayed coker heater in the heater tubes, the lighter com-
heater tube can be allowed to expe- ponents, typically from C1 to light
rience in view of its metallurgy. For tubes by two processes: by utilising naphtha, evaporate to form two
example, if the TMT at start of run the difference in thermal expansion phase flow in the heater. Along with
(SOR) conditions is 550°C and the coefficient between coke and tube cracking, asphaltenes in the vac-
maximum allowable TMT is 650°C by applying sudden temperature uum residue contribute to the cok-
(generally governed by its metal- variations (spalling); or by physi- ing phenomena, and part of this
lurgy), for a rate of 2°C/day rise in cal scraping of coke from the tube coke deposits on the heater tubes
TMT, the heater would run for 50 with the help of a moving pig (pig- to increase the TMT as time pro-
days. Therefore, it is desirable to ging). The typical time required for gresses. Besides feed characteristics
restrict the rate of the TMT increase spalling and pigging is 1-1.5 days such as asphaltenes, saturates, and
to a low value, which as such is a and 3-5 days, respectively. During residence time, coking inside the
function of the rate of increase in this period, the throughput of the coker heater is a very strong func-
the thickness of coke deposits. Once coker unit is reduced, which results tion of the temperature it reaches
the limiting value of the TMT is in a loss in production. Further, the in the tubes throughout the radi-
reached, coke needs to be removed tubes are subjected to a harsh envi- ant section of the heater. Various
from the inner surface of the tubes, ronment in both removal methods, measures are taken when designing
which normally would require and the frequency of coke cleaning the heat exchanger to eliminate or
downtime and consequent produc- eventually determines the life of the reduce the possibilities of localised
tion loss. Coke is removed from the heater tubes. Thus it is desirable to peak temperatures.
WWW.ZWICK-ARMATUREN.DE
low fluid temperature in the top cess fluid to the upper portion with the heater. The model is used to
low-temperature region are smeared a lower heating rate, resulting in mimic the flue gas side heat trans-
with a high emissivity coating so as lower TMTs, reduced coking rates fer in actual plant operation. Lobo
to absorb more heat. Because of these and higher run length. and Evans have presented a general
different emissivity coatings, the This article attempts to study the and simple theoretical treatment
heating rate experienced by hot fluid effects of change in process flow of the radiant section in a heater
reduces and that experienced by direction on heat transfer distribu- to correlate various variables and
cold fluid increases, which causes a tion in the radiant and convection their effect on its performance. The
more uniform cracking temperature. sections of a delayed coker heater. calculations to determine the radi-
Such coatings have also been used to This was determined by applying a ative and convective heat transfer
make the heating rate uniform across heater model employing commercial in a heater is carried out by using
the circumference of the tube in a as well as newly developed tools. commercial software which uti-
single fired heater. lises more rigorous heater models to
In a new configuration of coker Coker heater modelling and determine localised heat fluxes.
heater with respect to the flow of simulation On the process side, vaporisation
process fluid, termed upflow con- A heater model which takes into and molar expansion due to crack-
figuration (see Figure 3b), vacuum account both process side heat ing were computed using a com-
residue enters the radiant section transfer and combustion side heat mercial simulator which employs
from the bottom and exits at the top transfer is not available for predict- a kinetic model having five feed
of the radiant section. This means ing the performance of a delayed and product lumps to predict the
that fluid at a lower temperature coker heater. Thus, this study was yield pattern and properties of liq-
has a higher heating rate and fluid carried out by dividing the model uid and gas products at the outlet
at a higher temperature has a lower into three parts: a heater model, a of the heater. Here, the feed is char-
heating rate compared to conven- cracking model and a coking model. acterised by its distillation curve,
tional heaters, leading to a more A heater model for evaluating asphaltenes, density, and so on. The
uniform cracking temperature. This the performance of box heaters model breaks the feed into stand-
design claims to give improved with dual side firing was utilised. ard feed cuts, and applies calculated
coker heater performance by migra- It accounts for both convective as values for conversion and severity
tion of the hottest part of the pro- well as radiative heat transfer in to determine product yields. The
FUEL ANALYSIS
MINISCAN IR VISION
• Portable High Speed FTIR Fuel Analyzer for
Gasoline, Diesel, Jet Fuel and Fuel Blends
• 100+ Preconfigured Fuel Parameters
- Concentrations of Oxygenates, Aromatics, Benzene, FAME
- Prediction of Octane, Cetane, AKI, Vapor Pressure, Distillation
• Complete Fuel Report within Minutes
• Ready to Measure: Factory Calibrated Databases
• Remote Access for Diagnostics, Help and Troubleshooting
!"
= kC! (1) 0.5
!"
where CA is asphaltene
Model
concentration. 0
The reactions were considered to
occur inside the fluid film at an aver-
age temperature of the fluid film −0.5
and the bulk fluid. The oil film tem-
perature and TMT were estimated
based on API 530 calculations. Heat −1.0
−1.0 −0.5 0 0.5 1.0
flux required for determination of
Plant
the fluid film temperature (FFT)
and TMT were taken from the com-
bustion side simulations. The heat Figure 4 Coking model: plant values vs model predictions for TMTs
transfer coefficient (HTC) required
for the calculation of the oil film one of the delayed coker units in dent from the figure, the model was
temperature can either be calculated the refineries. TMT profiles pre- able to accurately take into account
from API 530 or estimated from the dicted from the model matched the all these changes and predict the
thermal cracking simulations. plant data very well. The coefficient TMT rise rate.
Each heater tube was further of determination (R2) observed From the above analysis of dif-
divided into ‘n’ number of parts between the model and the plant ferent models, it is apparent that
and solved for mass conservation in data was around 83%. The rise and the coking model with a commer-
each of these parts by the finite dif- fall seen in Figure 4 is because of the cial heater and cracking model
ference method. Properties like liq- actual variations occurring in the can be used effectively to predict
uid viscosity, liquid density, vapour plant with respect to feed flow, radi- the run length of a delayed coker
viscosity, vapour density and bulk ant heat flux, and so on. As is evi- heater. The different configurations
velocity, required for the calcula-
tion of the film velocity and film
thickness, were estimated from the 1.70 0.78, 1.61
1.60 Upflow
thermal cracking simulations. A Bottom
Downflow
1.50 IN
calibration factor was also intro-
1.40
duced in the model to quantify the
Heat flux ratio
1.00, 1.30
1.30
amount of coke depositing inside Bottom
1.20 OUT
the heater tubes from the total 1.10
amount of coke formed. The reac- 1.00
tion kinetic parameters were tuned 0.90 0.77, 0.84
to match the actual operating data. Top
0.80 Top OUT
If the maximum design limit of the 0.70 IN
1.00, 1.72
heater tube metallurgy is known, 0.60
run length can be predicted for var- 0.75 0.80 0.85 0.90 0.95 1.00
ious scenarios from a coking model. Bulk temperature, non-dimensional
The model was tuned to match
actual plant operations data from Figure 5 Heat flux ratio vs bulk temperature ratio in down- and upflow configurations
0.90
grated model for upflow configura-
tion in terms of percentage change
0.85 from a downflow configuration for
various parameters.
0.82
0.80 The estimated average radiant
Upflow heat flux for an upflow configura-
0.78 Downflow tion is greater than for a downflow
0.75 configuration by 4.79%. Better distri-
Run length
bution of heat in the radiant section
improves the heat input to the feed.
Figure 7 Rate of temperature rise to maximum TMT in down- and upflow configurations This results in a higher average heat
Further reading
1 Adam J, Hughes G C, Coker furnace online on-line spalling – safe, clean,
proven & profitable, AFPM AM, San Diego, California, 11-13 Mar 2012.
2 Gupta A, Patil R C, Mishra V P, Das A K, Hydrocarbon Processing, 2013,
92, 101.
3 Catala K A, Karrs M S, Seili G, Faegh A A, Hydrocarbon Processing, 2009,
87, 45.
4 Melton M S, US 5078857, 1992.
5 Golden S W, Barletta T, PTQ, 2006, 11, 105.
6 Zhu N, US 6626663, 2003, Fosbal Intellectual AG.
7 Gibson W C, Gibson R L, Eischen J T, US 6241855 B1, 2001, Petro-
Chem Development Co. Inc.
8 Lobo W E, Evans J E, Trans. Am. Inst. Chem. Engrs., 1939, 35, 748.
9 Bhirud V L, US 9359555B2, 2016, SBT Technology Inc.
It’s not easy to keep every control valve and every instrument
in your plant at it’s peak performance. Even a small problem
in any one of them could result in major issues for
the entire system.
Software that can be configured to solve any problem in the supply chain is
critical to planning in the oil and gas industry
DENNIS OSTENDORF
Quintiq
T
he oil and gas industry expe- improve. A move towards sustain- and transportation have increased
rienced the most severe able profitability is the new norm, demand for innovative solutions
downturn in recent his- with many companies looking to in refinery processes to produce
tory with the price of oil declin- achieve optimal profits while reduc- more and better fuels, while break-
ing by 75% from mid-2014 to early ing expenditure. throughs in exploration and extrac-
2016, throwing oil and gas compa- tion have increased the options for
nies into a state of survival. Many Why innovate? supply. Efficiency improvements
resorted to cost cutting measures While the worst has passed for in planning have proven especially
such as drastic reductions in their now, the crisis is far from over. It beneficial, particularly for primary
workforce, cancellation or post- will take years for the industry to distribution. In 2015, BP reported
ponement of new projects, and recover, and companies that invest record profits on the back of major
reduced investments in large scale in innovation to reduce costs and efficiency improvements in its
capital projects. increase their profit margins will downstream sector, which in turn
Today, companies are cautiously be the ones to come out of the cri- partially offset the negative effects
optimistic as the industry contin- sis stronger than before. Shell’s of a weakened environment in its
ues to recover. While prices are takeover of BG Group is a prime upstream division. The company
still far below the pre-downturn plans to continue streamlining its
peak of $115/bbl (March 2011), downstream operations to deliver
they are slowly but surely stabilis-
In the current oil annual cost savings of $1.6 billion
ing. An unprecedented agreement and gas climate, by 2018. With optimised logistics
between OPEC and non-OPEC planning in the downstream sector,
nations to reduce production and innovating can products can be delivered to stor-
tighten imbalances in supply and age facilities in the right amounts
demand have contributed to this mean the difference and the right time, with an empha-
positive outlook. Additionally, sis on cost effectiveness.
global energy demand is expected between surviving Ideally, innovations to increase
to rise by a third over the next two efficiency would be applied before
decades, while the shale revolu- the crisis and falling a crisis happens, enabling com-
tion and advances in oil recovery panies to be more robust in chal-
technology are expected to provide
even further behind lenging conditions. However,
an abundant supply to meet that innovation during a crisis can also
demand. example of a post-downturn invest- help to mitigate its adverse effects.
While the downturn was dev- ment in innovation. This mega- It is never too late to innovate; in
astating, it has led to an increased merger grants Shell extensive the current oil and gas climate,
focus on exercising capital and access to LNG assets and acceler- innovating can mean the difference
operational cost discipline. ates its global LNG and deep water between surviving the crisis and
Efficiency improvement was strategies, enabling the company falling even further behind.
emphasised, resulting in projects to adapt and thrive in a changing
with break-even prices as low as energy landscape. Three core capabilities of
$25/bbl, as claimed by Statoil for Oil and gas companies can also innovation in planning
its flagship oil project in the North look beyond the industry for inspi- The oil and gas industry is
Sea. This is a level that would ration to innovate. Even before the extremely complex – while there
have been unthinkable prior to the downturn, innovations in other sec- are key concerns that affect all play-
downturn and signals the possibil- tors have benefitted the oil and gas ers, no two companies will have
ity of even lower break-even prices industry – both directly and indi- the exact same planning issues. Oil
as planning processes continue to rectly. Advances in manufacturing and gas companies have historically
Table 2 Table 3
DECIE AUTIN RON HUIJSMANS WALTER PINTO ANDY O'CONNOR TAYLOR AUBURG
Vice President – U.S. Gulf Coast Senior Director of Director, Digital Vice President, Projects
Project Development, Program Director Global Projects Development FREEPORT LNG
EXXONMOBIL DOW CHEMICAL LYONDELLBASELL BASF NA
MUHAMMAD JUNAID BOYD PARKER JOSE PIRES RANDY POUND HERMAN VERHOEVEN
Director - Capital Project Global IT Executive Global Excellence Global Manufacturing VP - Global
Management Americas CHEVRON & Innovation Leader Director – Maintenance Reliability Excellence
HUNTSMAN CORPORATION ANDEAVOR & Reliability COVESTRO
OLIN CORPORATION
www.downstreamevent.com
downstream.indd 1 08/03/2018 10:08
loss of economic dominance. These ative impact the oil industry has data and KPIs and is also capable of
companies are also at a higher risk on the environment and shifting generating and performing what-if
of being affected by negative mar- demand on fossil fuels as sustaina- analyses on alternative scenarios in
ket conditions. ble options become more common real time. The results for this com-
While the thought of innovation – can be predicted and managed pany are that planners can now
can seem daunting, the application in a way that enables companies to schedule the movements of marine
of it in planning processes does not remain robust, no matter the chal- and inland logistics, monitor fuel
have to mean a huge shift in the lenges they face. The use of tech- and stock demand, handle spot
status quo. The benefit of innova- nology in their operations can also trades and monitor for potential
tive planning is that it incremen- have a positive impact on their outages. With end to end visibil-
tally improves a company’s existing brand, signifying better produc- ity, planners are equipped to create
strategies and profit potential. Each tivity, efficiency and customer plans that are sustainable, feasible
incremental improvement then con- satisfaction. and cost effective.
tributes to a larger, accumulated
positive value. Adopting innovation Integration: mastering complex
The reality of planning in the digi- No matter the industry, planning multi-modal logistics
tal age is that there are so many var- puzzles can be largely addressed This company initially made its
iables to consider that a company by applying the three capabilities name as a worldwide transporter of
could be losing money without even of visibility, integration and opti- grain by-products before establish-
realising it. An example of this is misation. The following cases are ing itself as a key player in crude oil
in vessel utilisation in primary dis- examples of how applying each logistics. Today, it transports crude
tribution. Variables include capac- innovative planning capability has from its source location by train to
ity and load of each vessel, sailing brought real value to the company. a terminal, and from there by barge
speed in relation to fuel use and age to refineries.
of chartered vessels. Compromises A 1000% increase in operations
are inevitable, but the decision on No matter the in the span of six months resulted
which compromises to make and in an incredible number of bottle-
when can often make the difference
industry, planning necks. The difference in capacity
between breaking even and making puzzles can be between its barges, trains and ter-
a profit. minal also created challenges in
Here is a simplified example of largely addressed matching the barge cycling rate,
the real value that can be obtained optimising tank levels and antic-
from incremental improvements as by applying the ipating the supply received. The
a result of better vessel planning. company sought a logistics solution
An oil shipping operation is out- three capabilities of that could integrate both rail and
lined in Table 2. barge operations. It needed a solu-
Even with small reductions in visibility, integration tion that could process information
shipping costs, the potential addi- from different sources on a day-to-
tional profits are significant (see
and optimisation day basis and propagate it through
Table 3). the entire system.
From these calculations, it can be Visibility: optimising sourcing, With such a system in place, the
seen that a reduction of shipping opportunities and utilisation company’s planners now have a
costs as a result of better logistics While not strictly within the oil and granular view on the cycling of all
planning has led to a significant gas industry, this energy provider trains and barges. Having a unified
increase in revenue. Better logistics deals in an alternative to fossil solution also means that planners
planning also enables oil and gas fuels, giving a possible future direc- were able to improve overall plan-
companies to have better insight tion for companies in oil and gas. A ning efforts to significantly cut costs
into where to ship and sell their large producer of renewable power, and utilise the company’s tank cars
products to gain the most economic it provides energy via a power sta- and barges more efficiently.
benefit. However, this capability tion fuelled by biomass and coal.
requires close integration between As the company grew, it came Optimisation: cutting down
all relevant departments in the face to face with a major challenge: planning time, increasing overall
company, which can be challenging the need for end to end visibility of efficiency
without the right planning solution. its supply chain. This included vis- As a new energy provider in a
The right planning solution ibility over transport movements, growing market, this company
can also help oil and gas compa- inventory levels, buying and sell- utilises over 200 trucks to deliver
nies to adapt quickly in an envi- ing opportunities and possible billions of litres of petroleum prod-
ronment that is evolving at an disruptions. ucts from dozens of terminals and
unprecedented pace. The effect of For this organisation, the solution depots to nearly 700 service stations
constant changes – such as new was a planning system that gives throughout a continent.
laws designed to reduce the neg- planners a clear view of all relevant One of the key challenges for this
Level Measurement
■ Visual Level Indicators VLI
■ Tank Level Instruments TLI
FRANK BAKKER STACEY OLSON DAVID HOOKER SAJJAD AHMED RICHARD WOLFLI
CEO PRESIDENT - PRESIDENT FORMER BECHTEL PROJECT FOUNDER & DIRECTOR
US METHANOL LLC APPALACHIA MOUNTAINEERING MANAGER ON PTTGC OHIO US METHANOL LLC
CHEVRON NGL STORAGE, CRACKER PROJECT
ENERGY STORAGE VENTURES SABIC
SECRETARY H. WOOD THRASHER SENATOR MIKE TURZAI STEFANI PASHMAN ROBERT RICHARD
CABINET SECRETARY CAMERA BARTOLOTTA SPEAKER OF THE HOUSE CEO SENIOR VICE PRESIDENT OF
WEST VIRGINIA STATE SENATOR PENNSYLVANIA HOUSE ALLEGHENY CONFERENCE MAJOR ENTERPRISE
DEPARTMENT OF COMMERCE PENNSYLVANIA SENATE OF REPRESENTATIVES ON COMMUNITY DEVELOPMENT PROJECTS
DTE ENERGY
www.petchem-update.com/northeast
NEW & EXCLUSIVE FOR 2018:
NEP WILL BE COLOCATED WITH:
John Hague
AspenTech
N
ew methods and cutting edge How we got here itoring, and computational power
technologies are driving asset The foundation for realising a new continued to increase. Systems
performance management APM vision has existed for some were still isolated, but in separate
(APM) well beyond historical capa- time, with engineers having applied systems, engineers started to see
bilities, rapidly increasing its bot- performance models for decades. something resembling real-time
tom line value. Technologies like Such pioneering APM adopters asset-level data.
cloud computing, data science and faced the challenges and constraints By the late 2000s, this situation
machine learning are now being of the technologies surrounding had changed significantly and we
integrated with automated method- their models. Disparate systems started to see multiple parallel tech-
ologies directly into APM solutions. evolved to manage and optimise nology innovations coalesce into the
This wave of integration places maintenance functions, to develop modern state of the art APM meth-
advanced analytical techniques into risk assessment and criticality, and odology. Best in class systems could
the hands of operators and engi- to perform continuous condition now incorporate detection of pre-
neers with previously unimagined monitoring. cise patterns of normal and failure
scale and ease of use. The incremen- behaviours, and perform the com-
tal progress in APM over the last 20 Low-touch machine putational isolation of key indi-
years pales in comparison to what cators of degradation. Especially
is now possible through digital learning ushers important was the 2006 debut of
transformation. Amazon Web Services for scala-
Low-touch machine learning is in a new era of ble cloud computing. Advances in
the key catalyst to scale APM’s structured and unstructured data-
potential beyond existing first prin- performance and bases and operational data pools
ciples based solutions and ‘armies’ were tested and improved at the
of consultant engineers and data
optimisation for enterprise level during this period.
scientists. A widespread integra- every industry Around the same time, smart
tion of machine learning in APM sensors saw a dramatic shift in
marks a transition from estimated performance, size, reliability and
engineering and statistical models These systems were isolated, price. Added to this was a dra-
towards measuring patterns of asset resulting in limited connectivity matic improvement in the compu-
behaviour. and integration, as well as workflow tational and analytical capability of
Operators of refineries can now impairments. Due to limited inte- machine learning called ‘deep belief
readily extract value from decades gration, early computers processed networks’ or ‘deep learning’. This
of existing design and operations small volumes of available related breakthrough was pioneered by
data to better manage and opti- data in batch mode instead of in real Geoffrey Hinton at the University
mise asset performance. This low- time, when insights are most valua- of Toronto, who is now tightly cou-
touch machine learning method ble. Outputs came too late, typically pled with Google.
continuously embraces changes in in days or weeks. Computational The result was a quantum leap
asset behaviour, empowering real- power limited the advancement of in capability, and this has enabled
time APM value creation. Vetted new algorithms. And assured static machine learning to surpass the
and tested across diverse indus- models were fixed, low frequency performance of previous analytical
tries, scalable across multiple assets, and not adaptable to new failure techniques, which limited modelling
and powered by cloud and parallel behaviours and incremental opera- and statistical methods. Machine
computing, low-touch machine tional changes. learning is now the dominant ana-
learning ushers in a new era of As the 2000s arrived, leading lytical method in all IT fields around
performance and optimisation for industries started to better organ- the world. It is used for credit card
every industry. ise assets for condition based mon- fraud detection, facial recognition
REFCOMM
- TRAINING
- CONFERENCE
- EXHIBITION
Delayed Coking, CatCracking, Sulfur Recovery (Amine, Sour Water, Claus, Tailgas)
LASZLO MAARTEN
VARRO WETSELAAR
REGISTER
TO HEAR
FROM THESE SAAD CHARIF
I N S P I R AT I O N A L SHERIDA SOUKI
AL-KAABI
LEADERS
AND GLOBAL
S T R AT E G I S T S
W W W. G A S T E C H E V E N T. C O M / P T Q 1
SPONSORS INCLUDE
DMG116
TRANSFORMING
OUR INDUSTRY
TOGETHER
RMC18 has something for every member of your team! REGISTER TODAY!
• Focused presentations by industry subject matter experts AFPM.ORG/CONFERENCES
• Roundtable discussion sessions with your industry peers
• Tech Talks presenting the newest technologies
Visit www.afpm.org/conferences and select Reliability & Maintenance Conference to learn more. Photo ©Marathon Petroleum Corporation
afpm new.indd 1
AFPM_2018_RMC_Meeting_Ad_PTQ_v1.indd 1 09/03/2018 10:43
3/8/18 8:35 PM
Wireless instruments in corrosive and
hazardous locations
Wireless instruments introduce a range of benefits in corrosive and
hazardous process areas
T
his article explains the dif- maintenance, and minimal impact that are not permissible in today’s
ficulty, time and expense of late changes, workmanship and automated plant. Temperature and
involved in installing and materials. humidity control is the single most
maintaining a traditional instru- In a multiple wiring methods/ important factor affecting corrosive
ment wiring system or single protocols system, it is not a mat- rates when corrosive vapours are
wiring method and shows how ter of counting the number of I/O present in the environment.
multiple wiring methods/proto- signals and figuring out if they are There are many available plans of
cols along with wireless instru- 4-20 mA analogue or 0-24 Vdc, or if action designed for safety in process
ments can avoid those problems. they are input or output; it is about installations. The simplest method
In general, an instrument wiring selecting the right I\O type and to prevent, minimise or control
system (including cables, junc- wiring protocol to achieve process the risk of corrosion and chemical
tion boxes, conduits, termina- functionality by reducing effort and vapour attack is to keep instrumen-
tion racks, cabinets, enclosures, cost. The article details different tation out of the area altogether,
cable tray, tray support system, designs, engineering, deliverables or to make the area less hazard-
multi-cable transits [MCT] and and cost impacts during the imple- ous or less reactive with controlled
marshalling panels) contributes mentation of multiple wiring meth- humidity and temperature through
significantly to the plant’s total ods/protocols along with wireless process improvements. However,
instrumentation costs. The article instruments on the same project. when instruments must be installed
focuses on different impacts dur- Oil and gas, petrochemicals, in corrosion prone areas (such as
ing the implementation of multiple chemicals and other process indus- waste gas treatment and sulphur
wiring methods/protocols along tries often face problems of elec- handling areas), all the instrument
with wireless instruments on the trical downtime, electrical service wiring system (including cables,
same project in order to prevent, disruption, erratic instrumentation junction boxes, conduits, termi-
minimise or control the risk of cor- system behaviour, and loss of I/O nation racks, cabinets, enclosures,
rosion and chemical attack on an cards in DCS and PLC. Increased cable tray, tray support system and
instrument wiring system or cable resistance across the connection MCTs) must be installed and main-
infrastructure to meet the plant’s and contactor results in false data tained to appropriate standards to
requirements.. transmission through the instru- prevent deterioration of the metals
The availability of multiple wir- ment data cable to the system due used.
ing methods/protocols provides to an open or intermittent connec- Generally, metals used in wiring
numerous benefits over single tion of the power or instrument terminals have platings or coatings
type of wiring methods/protocols data cables. Wet, salt-mist, dirty, for protection against corrosion and
including reduced wiring expenses dust contaminated or corrosive chemical attack. These sacrificial
and marshalling panels, fewer vapours increase the risk of corro- or barrier layer coatings over wir-
power supplies and safety barriers, sion and chemical vapour attack ing terminals metals are intended
smaller equipment room, less docu- on the metals used in wire termi- to be consumed during a corro-
mentation and design effort, fewer nal connections, switches and con- sion reaction or to provide a protec-
input/output (I/O) channels in the tactors. Wiring connections can also tive seal. They are effective as long
distributed control system (DCS) be affected by creep corrosion and as the metal in wire terminals is
and programmable logic control- metallic dendrites due to the reac- undamaged and impermeable. The
ler (PLC) cabinets, improved self- tion of corrosive vapours with the expense of protecting wired 4-20
diagnostics and remote diagnos- metal of a connector at high or var- mA conventional, HART, Fieldbus,
tics, increased functionality and iable humid conditions. Intermittent Modbus, Profibus or Profinet instru-
transparency, more information terminations of wire connections ments in a corrosion prone envi-
from the field level, easier exten- in instrument wiring are expen- ronment or hazardous area is
sion during operation, feasibility of sive and time consuming problems significant. Wiring for field instru-
ments (such as transmitters with or dency to undergo a corrosive reac- as crushed cables, excessive length,
without display, switches and con- tion with a variety of industrial mechanical fatigue, poor cladding,
trol valves) requires cables, junction gases/vapours such as sulphur, cable routing complexity, routing
boxes, conduits, termination racks, chlorine and their compounds. between moving components, and
cabinets, enclosures, cable tray, tray A harsh environment requires weight carrying.
support system, MCTs and mar- detailed engineering design of
shalling panels. If the instrument is appropriate corrosion control meas- Implementing wireless instrument
source input or four-wire, it must ures to avoid unnecessary electri- technology
have separate power wiring. All cal downtime, planned/unplanned Introducing wireless instruments to
of the wiring system must meet maintenance and inspection costs. oil and gas, petrochemicals, chem-
the requirements of protection that An instrument wiring system or icals and other process industries
the circuit is certified for. Installing cable infrastructure affected by provides such benefits as robust-
an instrument wiring system can short/long term exposure to cor- ness, fast real-time response,
be quite expensive and additional rosion displays a variety of typical reduced installation time, less wir-
operational maintenance expenses behaviour patterns: ing infrastructure, and lower power
are required on preventing, mini- 1. Electrical shutdown or total dis- consumption. Significant gains are
mising or controlling the rate of cor- ruption of electrical services due to possible with the implementation of
rosion to ensure the protection level short/open circuits wireless instruments in unclassified
is being maintained. 2. Unexplained loss of I/O section areas of process plant, hazardous
Typical corrosion related issues of a circuit locations or corrosive environments.
found in a wiring system dur- 3. I/O cards lost in DCS and PLC, These can be viewed in terms
ing inspection in a corrosion prone data reading and transmission prob- of engineering impacts, design
environment are shown in Figure 1 lems, and erratic behaviour impacts and construction/fabrica-
and include: 4. Terminals, leads, switches and tion deliverable impacts.
• Incorrect or damaged cable gland electrical contractors become cor-
• Damaged cables roded which reduces the conductiv- Engineering/design impacts
• Corroded terminals, switches and ity of the metals The introduction of wireless instru-
contactors 5. Connection of the data/power ments in process plant has many
• Corrosion across the connec- cables inside junction boxes located positive effects on project engi-
tion and contactors, resulting in in the field, and marshalling of sys- neering, such as lower installa-
increased resistance tem cabinets located in local equip- tion costs, quicker installation time,
• Corrosion fatigue failure of ment/control rooms are affected, so faster commissioning, more efficient
terminals. resistance increases across the con- change order management, removal
A wireless infrastructure could nection and contactors, resulting in of the requirement for a power sup-
possibly avoid these instrument false data transmission ply and protection barriers with
wiring system problems, corrosion 6. A combination of a corrosive replaceable battery packs, increased
related inspections, and planned/ environment and vibration assists system/vendor compatibility, and
unplanned maintenance. in the formation of corrosion fatigue device and system compatibility.
cracks in terminals leads, leading to The comparative ease of moving
Effects of corrosion on wiring wiring system failures. or adding I/O points during con-
systems In addition, non-corrosion related struction enhances the cost effec-
All instrument wiring systems or stress in a wiring infrastructure may tive management of on-site changes.
cable infrastructures have a ten- impact measurement quality, such Implementation of wireless instru-
pa da
Ca rs th
Fr
pe y 4
ll clo Ma
i
fo se y
r s
27 – 30 November 2018 // Grand Hyatt Cannes, France
GASES
RENEWABLES
& ALTERNATIVES
TOGETHER
SUCCEED
WATER
axens.net/blog
axens.indd 1
ad_corporate_210X297mm.indd 1 23/02/2018
02/02/2018 14:47
16:28:43