The cracked gas compressor (CGC) is the critical system in
modern ethylene plants. This compressor drives gases from the
crackers for downstream separation. Few plants can afford the
luxury of backup compressor trains. Any downtime or reduced
capacity of the CGC negatively impacts olefin facility profits.
Under this highly competitive environment, even fewer ethylene
operators can afford the thousands or even millions of dollars
per day in lost revenues due to unplanned shutdowns to clean
and repair CGCs.
Significant design advancements have developed more robust
systems. Although many olefin plants are making substantial
investments in hardware and metallurgy to improve compressor
performance and to extend run lengths, the reliability, efficiency and
throughput capacity of the CGC are still key influences on
plant profitability. As engineering designs have improved, more
attention is directed to preventing costly fouling of these
critical systems with innovative chemical treatments.
The purpose of the CGC is to compress the gases from the
cracker for separation in downstream units. Due to the low
boiling points of the light gases, very low temperatures are
required for separation at feed-stream pressures. Fig.
1 is a basic schematic for a five-stage compressor
1. Typical compressor train schematic.
Compressor fouling problems are common. They are
particularly troublesome in gas crackers due to the low volume
of aromatic gasoline formed during cracking. Aromatic
hydrocarbons are useful because they help keep fouling
precursors in solution for easy removal. Liquid crackers have
an inherent advantage due to higher aromatics production;
however, these crackers also experience extensive fouling. The
fouling rates in liquid crackers processing a variety of feedstocks such as heavy naphtha,
gasoil (GO) and atmospheric GO residues have increased
Costs associated with CGC fouling are high, and they can
increase exponentially as conditions deteriorate.
Result: The reliability of CGCs remains a
serious issue in spite of improvements in design and system
metallurgy. Therefore, olefins plants are exploring new
techniques to optimize compressor performance and run length.
Many operators are discovering that chemical additives can
offer attractive advantages for fouling control.
Gas crackers usually crack ethane and/or propane, whereas
liquid crackers usually crack naphtha. The increasing volumes
of shale gas and advantageous pricing have shifted more
ethylene plants to rely heavily on natural gas feedstocks. However, this pattern is
changing due to competition and growing demand for natural gas.
Several ethylene producers are responding by turning to
heavier, cheaper feedstocks such as GO and residues
to improve profitability and to operate their systems at more
severe conditions, thus increasing throughput. Variations in
feedstock properties, higher cracking severity, and operations
near design capacity accelerate polymerization reactions that
lead to compressor fouling.
Unsaturates formed during the cracking process are reactive
species, and levels found in effluent gas vary depending on the
feedstock. Although some heavier components are eliminated in
the quenching operation, the cracked gas still contains
virtually all the C4s and most of the C5s
and C6s, along with some heavier fractions in the
gaseous phase. This stream also contains appreciable amounts of
highly reactive di-olefins and acidic compounds that are
subject to oxidation and/or polymerization. Unsaturates
contribute to free radical generation via thermal reaction or
Dies-Alder mechanism or oxidation at the high temperatures
found in the compressors, forming polymers. These polymers tend
to accumulate in the compressor discharge lines, casing and
Most of the worlds ethylene producers rely on wash oil
and wash water as a mitigation strategy. Wash-water injection
reduces the compressor discharge temperature, which helps
control fouling. Conversely, water injection also leads to
erosion and corrosion of the compressor wheels and blades.
Wash oil is used in a similar manner but works in a
different way. It serves as a solvent that dissolves some
polymers, thus allowing them to be removed from the system.
For maximum effectiveness, wash oil should be highly
aromatic (60%) and have a boiling range of 175°C to
315°C. Most important, it must be free of styrene,
naphthalenes and diene compounds. Wash oil that meets these
specifications is increasingly more difficult to obtain
locally, and shipping adds significant cost. High cost and
short supplies are discouraging the use of wash oil. A number
of ethylene producers have found that high-quality wash oil and
wash water alone cannot provide sufficient fouling control.
Potential fouling factors
Fouling is a continuous process driven by free-radical chain
propagation, oxidative and catalytic polymerization, and
Diels-Alder reactions. The polymers initially formed by these
reactions are soluble in aromatic streams unless they undergo
continued polymerization. These factors promote polymerization
reactions in CGCs:
High concentrations of monomers, including dienes formed
by high-severity furnace operations
High compressor-outlet temperatures
Organic peroxide-induced fouling caused by oxygen
Return/recycle stream contamination by peroxides and
other polymerization catalysts from LLDPE and HDPE
Fouling increases with the number of these influences
present within the system. Fig. 2 illustrates
that it is essential to control as many of these factors as
possible to manage the fouling rate before it can begin to grow
exponentially. Failure can make it extremely difficult or
impossible to control compressor fouling.
2. Relationship between fouling rate
and factors influencing polymerization.
Fouling within a compressor train is seldom uniform, and
some areas are more susceptible than others. Fouling is worse
in hotter areas along the wheels near the discharge, discharge
piping, after-coolers and diffusers. Medium-pressure (MP)
casing fouling is worse than high-pressure (HP) casing fouling
because the concentration of reactive dienes and other monomers
increases as condensation occurs in these stages, and the
reactive monomers from the gas phase tend to dissolve (or
diffuse) into liquid hydrocarbons that condense during
Once in the liquid phase, the reactive monomers may undergo
polymerization where typically monomers with reactive double
bonds such as butadiene, styrene, isoprene and vinyl acetylene,
react and polymerize. The condensation of lighter hydrocarbons
can make the problem worse. Polymers forming in the
compressors, if not depositing on the machine, will likely
accumulate on the tube sheet or in the shell side of the
Continuous exposure to high temperatures supports
poly-merization. As the reactions between the monomers proceed,
large polynuclear aromatic compounds form. As the polymer
grows, it loses solubility, begins to cross-link, dehydrates
and transforms into brittle and insoluble coke deposits.
At this point, no aromatic stream, wash oil or dispersant
will solubilize the polymers or prevent their deposition.
Increased wash-oil injection may dislodge some of the polymers
and allow them to move within the compressor. Some will
accumulate in the after-cooler inlets or in knock-out pots,
resulting in high pressure drops.
Experience in a number of plants indicates that polymer
deposition on after-cooler surfaces has a significant impact on
plant run-length. Such deposits limit throughput and increase
power consumption, eventually forcing a shutdown. Few olefin
plants have spare exchangers or bypass routes available, so
after-cooler exchangers must be protected from fouling
3. CGC fouling process.
A certain amount of fouling is inevitable, but it can be
controlled. The key is to control fouling as it initiates
during the polymeric chain propagation rather than to implement
a treatment program for an already-fouled system. During the
initial polymerization steps, the polymers are more likely to
be hydrocarbon soluble.
Advanced mitigation techniques
Conventional chemical mitigation techniques involving
traditional antifoulant/dispersants and wash oil have limited
value. Conventional dispersants work by physically removing
some deposits from the fouling site. However, they cannot
prevent solids generation. They are ineffective on the
insoluble polymers, metal precursors, peroxides and free
radicals that contribute to the fouling. In addition,
dispersants can have costly negative side-effects.
These compounds have an affinity for water; when added in
the CGC system, they can migrate with the wash water. Once in
the quench-water circuit, their surface-modifying properties
can create tight emulsions that make it more difficult to
separate hydrocarbons from process water. Dispersants may also
cause operational upsets and enhance after-cooler fouling by
moving foulants from one location to anotherthe so-called
An innovative chemical treatment program can prevent fouling
by inhibiting reactions. This unique formulation contains a
true inhibitor and a highly effective antioxidant. The true
inhibitor reacts with monomers before they can form insoluble
polymers, and the antioxidant-reduced oxidative polymerization.
Compatible metal chelators or passivators are added as
This unique formulation traps and inhibits free radicals by
making the precursor or reactive species inert, thus retarding
the polymerization reaction rate and reducing solids generation
by altering the formation of insoluble molecules. The method
attacks the polymerization/fouling processes in the initial
phase where it is easiest to control and dramatically reduce
solids generation. The treatment has allowed some compressors
to run efficiently for five to six years with minimal fouling
in the MP and HP casings and after-coolers.
The new fouling control program contains no metals and less
basic nitrogen than amine-based
antipolymerants currently available in the market. This is
important because traditional high-nitrogen chemistry reacts to
form insoluble salts under acidic conditions, as when hydrogen
sulfide (H2S) is present, essentially trading one
fouling problem for another and contributing to the
after-cooler pressure drops that can eventually force unplanned
shutdowns. A new antipolymerant is effective in the presence of
H2S and does not form insoluble salts regardless of
Monitoring tools optimize performance
Monitoring is important when any treatment program used, but
it is essential for optimum CGC treatment. Pressures,
temperatures, flowrates, and wash-water and wash-oil injection
rates (if any) are the keys to understanding compressor
performance. Monitoring these parameters allows calculating the
theoretical efficiency of the compressor system as well as
after-cooler pressure drop limits. Other important performance
indicators include machine vibration, knock-out drum polymer
content and turbine energy consumption.
Fouling is a complex process. Both gradual
trends and sudden changes in these parameters can imply
developing problems. Solutions require an understanding of root
causes that may not be immediately obvious. Proprietary models
and sophisticated simulation tools can be used to explain
complex interrelationships between changes in feedstock and
furnace operating conditions and their impact on the
composition of cracked gas entering the compressor.1
Such tools can be used to evaluate how flow variations in
recycle/purge streams influence cracked-gas composition and the
effects on compressor efficiency. The end result is a set of
realistic performance improvement targets based on a complete
understanding of what is actually happening within the
The chemical composition analysis of fouling deposits
provides a wealth of additional information. Organic and
inorganic content can be identified, along with any corrosion
products that may be acting as fouling catalysts. A full
elemental analysis can identify impurities, and the carbon-to-hydrogen ratio indicates
the extent of polymerization in the compressor train as well as
the nature of those polymers.
The analysis allows treatment to be fine-tuned for maximum
performance, but each analysis is only a snapshot of a dynamic
process. Other proprietary assessment tools can address this
problem by allowing online sampling in real time, using a
retractable screen that can be inserted and withdrawn as often
as necessary for sample collection and
Case 1: Fouling example
A 400,000-tpy (400-Mtpy) olefins gas plant cracked a mixed
feedstock consisting of 80% ethane
and 20% propane. The quench-water tower overhead was compressed
in a four-stage centrifugal unit driven by a steam turbine
operating at 1,500 psig and 950°F. The second-stage CGC
after-cooler design was unusual in that the process side was on
the tube side of the exchanger. This intercooler was
susceptible to fouling and had unacceptably high-pressure
drops. During normal operations, the plant used wash water at
about 0.7% of gas flowrates to maintain compressor discharge
temperature at 194°F and also injected wash oil weekly
(80%90% aromaticity) at about 0.1% of the charge gas
Problem. The plant experienced an unplanned
shutdown approximately 18 months after a previous planned
turnaround. Unfortunately, this plant had no backup reboiler. Reasons
for the unplanned shutdown included:
Stage 2 pressure drop approached after-cooler design
limits (from 7 psig to 26 psig)
Stage 2 polytrophic efficiency decreased 8%10%
below start of run (SOR)
Stage 3 polytrophic efficiency decreased 5%8% below
Operating capacity declined 30%50%.
Considerable fouling was observed in the second-stage
after-cooler during the unplanned turnaround. Because the
compressor had been cleaned during the planned turnaround 18
months previously, the plant expected little fouling and,
therefore, did not inspect or clean it. Conventional
dispersants had been used to wash off the polymers formed in
the second and third compressor stages in an attempt to
eliminate polymer deposits in CGC casings and discharge
Evaluation and recommendations. Foulant
samples were collected and sent to a third-party specialty
chemical research division to identify the fouling precursors
and mechanisms. Based on that analysis, the specialty chemical
company recommended a multifunctional antifoulant to control
fouling where the problem was most serious, in compressor
stages 2 and 3. Dosing hardware and portable feed tanks were
rushed to the site. Due to the rapid pressure drop increase, it
was recommended to install assessment tools to facilitate
polymer sample collection (Fig. 4) and
analysis so that treatment could be optimized and fine-tuned as
conditions changed.2 Fouling declined and plant
performance improved dramatically.
4. Proprietary online fouling monitoring
system with retractable filter and example
retractable filter with deposit
Sample analysis. Data summarized in
Table 1 indicate that although some
polymerization continued to take place, the severity of
polymerization decreased with treatment, as indicated by
polymer molecular weight and degree of cross-linking. Further
analysis showed that the deposit consists mostly of
non-cross-linked polymers. Treatment was successful in
Foulant samples were initially collected
quarterly.2 When the pressure drop across the
second-stage intercooler stabilized, the sampling interval was
lengthened to six months (Fig. 5). Pressure
drop remained stable and sample collection was discontinued
when analysis 4 indicated that the fouling was under control.
In this case, a monitoring program and additives improved
fouling control and increased plant run-length while achieving
substantial energy savings.
5. Second-stage after-cooler
pressure drop trends.
Case 2: Treatment interruption increases fouling
A 400 Mtpy-ethylene plant, using a mixture of ethane and
propane feedstock, experienced fouling problems in the CGC
unit. Evidence of severe fouling was found when the system was
disassembled during a shutdown in September 2010, and a
third-party specialty chemical company was approached for
recommendations. A site survey revealed several problems:
- Decreasing plant throughput
- Increasing energy consumption (specific steam
- High axial displacement in the MP-stage compressor
- Increasing CGC discharge temperatures in all stages
- Increasing wash-oil consumption.
Successful antifoulant trial. Following a
turnaround in 2010, the plant approved a six-month antifoulant
trial program, and treatment was initiated in January 2011.
Several CGC parameters were monitored daily:
- Axial displacement
- Specific steam consumption
- Compressor discharge temperatures
- Polytrophic efficiency.
The treatment program successfully controlled axial
displacement and vibration in all three monitored compressor
stages. The data indicated steady-state performance within
compressor design parameters. Steam consumption per ton of
ethylene stabilized below pretreatment consumption levels.
Discharge temperatures remained below operating targets and
polytrophic efficiency increased during the trial (Fig.
6. MP compressor axial displacement trends.
When the trial was complete, plant management decided to halt
the dosage and to monitor performance without treatment for
comparison purposes. The situation remained relatively stable
for three months. Later, indications of fouling reappeared in
July, and the situation degraded rapidly and became more
difficult to control.
The axial displacement increased, steam consumption began to
climb and the polytrophic efficiency declined markedly. As
discharge temperatures increased, polymer deposits accumulated
at rising rates on compressor internals, causing potentially
damaging vibration and a rapid increase in axial displacement
that soon exceeded the design limit for the MP compressor
Compressor loading reduced to prevent
shutdown. The treatment resumed at the plants
request in August 2011. However, the axial displacement was so
severe that vibrations were almost impossible to control, and
the polytrophic efficiency continued to decline. Management
decided to halt the antifoulant dosing again and reduce
compressor loading by 20% to 30% to prevent the compressor from
going offline and shutting the plant down completely.
The reduced compressor loading forced the plant to operate
below normal capacity, thus lowering gross revenues as steam
and wash-oil consumption increased operating costs.
Unfortunately, the next planned shutdown was many months
Capacity increased with modified treatment.
To improve operating economics and keep the plant running until
the next planned shutdown, third-party specialty chemical
technicians reformulated the antifoulant with enhanced surface
modifiers and dispersants to increase antifoulant activity in
the compressor after-coolers and disperse the organics and
corrosion byproducts (Fig. 7). When the
modified program was implemented in November 2011, the axial
displacement declined and then stabilized slightly above design
7. Plant-monitored polytrophic efficiency
Although the situation was still far from ideal, the stabilized
system allowed the plant to increase throughput approximately
10%, increasing revenues and avoiding an extremely costly
unplanned shutdown. The resulting efficiency improvement
reduced operating costs for steam and wash oil, further
improving plant profitability. The modified treatment was
resumed after the planned turnaround in April 2012, with
excellent results. Wash-oil consumption was reduced by 30%,
axial displacement stabilized within design limits, and
specific steam consumption declined substantially.
Results. The modified treatment allowed a
10% increase in load over the five-month period preceding the
planned shutdown scheduled for March 2012. It also increased
ethylene production by 4%. At an average market price of
$1,400/metric ton, the 15 additional tons of ethylene added
$21,000 to the plants bottom line.3
Average steam consumption declined, thus reducing energy
costs. The plant required approximately 1 ton of steam to
generate 140 kW of energy at a cost of $50/ton (Fig.
8). The additional steam required when treatment was
discontinued between April and August cost the plant nearly
$33,000. Restarting the treatment program offset virtually all
of that additional expense by generating more power with less
steam, saving the plant approximately $31,000.
8. Specific steam consumption patterns.
The plant purchased approximately 25 tons of wash oil
annually at a cost of around $2/kg. The wash-oil savings under
the modified treatment program are estimated at approximately
20%, a savings of $10,000 on an annualized basis.
Even the very best antifoulants are preventive tools.
Fouling is a continuous process. Once fouling accumulates, it
is extremely difficult to stabilize, let alone to remove by
chemical treatment. In this case, the decision to halt the
first treatment regimen allowed fouling to resume and worsen.
Although the second treatment regimen helped to stabilize the
partially fouled system, accumulations occasionally broke free,
causing compressor imbalances and increasing axial shifts. The
on-again, off-again treatment is risky. When fouling symptoms
are observed, the treatment with an effective antifoulant
should be initiated as quickly as possible and maintained
Case 3: Mixed-feedstock cracker compressor fouling
Within six months after a specialty chemical company began
compressor antifoulant treatment in 2008, the olefin plant
changed its feedstock from 100% naphtha to 60% naphtha and 40%
LPG. The treatment allowed the plant to meet a 2012 turnaround
schedule established before the subsequent feedstock changes. The operating
issues were compressor fouling, decreased polytrophic
efficiency and throughput limitations due to increasing suction
pressure. Following the treatment program, plant reliability improved. More stable
operations were possible with less fouling. The olefin plant
could increase throughput capacity and conserve steam
consumption. More importantly, the plant was able to meet the
targeted five-year run length with clean compressors. CGC
efficiency inevitably declines over time as the system wears
and deposits accumulate. Controlling fouling deposits is
essential, given the critical role played by CGCs and their
impact on plant revenues and operating costs.
1 Compressor Advanced Simulation Software
2 Fouling Assessment Tool (FAT)
3 Average market price according to ICIS
Manisha Jain is the lead
engineer for technical services with Dorf Ketal
Chemicals at Dubai, UAE. She holds a BE degree in
chemical engineering from Mumbai University, India. Ms. Jain has six
years of experience in troubleshooting petrochemical plants.
Sudarshan Vijayaraghavan is
technical leader for petrochemicals for Dorf
Ketal Chemicals treatment programs in Asia Pacific. He
is a key member of the training, technology and monitoring
(TTM) initiative, where his work involves program
monitoring modeling for gasoline fractionators, CGC and
light ends units with various licensor
Saurabh Shende is the technical
manager in global petrochemical technical
services group with Dorf Ketal Chemicals. He has about
12 years of experience in preparing prototype models
for plant monitoring and has remotely assisted in the
startup of one of the largest ethylene crackers. Mr.
Shende holds an MS degree in chemical engineering from
UDCT (now known as ICT).
Kyle D. Mankin is the global marketing
manager-petrochemicals for Dorf
Ketal Chemicals, Houston, Texas. He has 32 years of
experience in petrochemicals plants around
the world, including two world-scale ethylene plants.
Mr. Mankin holds a BS degree in chemical engineering
from the University of Texas at Austin.
Mahesh Subramaniyam is the
director of research and development with Dorf
Ketal Chemicals and leads the companys chemical
development. He holds a PhD in chemistry from the Indian Institute of Technology, Mumbai.