Petronas improves ethane extraction of gas processing complex

Oct. 25, 2004
Due to greater demand for ethane in Malaysia and a desire to improve plant operating stability, Petronas Gas Bhd. (PGB) investigated different technologies to increase ethane recovery and improve plant performance at its Gas Processing Plant A (GPP-A) complex.

Due to greater demand for ethane in Malaysia and a desire to improve plant operating stability, Petronas Gas Bhd. (PGB) investigated different technologies to increase ethane recovery and improve plant performance at its Gas Processing Plant A (GPP-A) complex. GPP-A, in Kerteh, Terengganu, Malaysia, contains three turboexpander ethane recovery plants.

This article discusses the process technology upgrades that PGB studied, and the design and construction efforts to retrofit the chosen technology into the existing facilities. Operating experiences encountered during the recent startup and commissioning of the revamped units are also discussed.

Petronas gas processing

PGB, a subsidiary of Petroleum Nasional Bhd. (Petronas), the national oil and gas company of Malaysia, operates six gas processing plants. The GPP-A complex includes trains GPP-1 through GPP-4; the GPP-B complex includes the trains GPP-5 and GPP-6.

The complexes have a combined production capacity of 2.75 bcfd of sales gas that supplies Malaysia's power sector. The GPP facilities also produce ethane, propane, butane, and condensate products, which are supplied to the adjacent Petronas petrochemical integrated complex or exported to neighboring countries.

During the Asian economic crisis of 1997-98, ethane and its derivative, ethylene, enjoyed tremendous demand and high prices. The price of ethane increased to $550/tonne in 1998 from $350/tonne in 1996.

Since GPP-2, 3, and 4 were commissioned in the early 1990s, the design ethane recovery of 80% was never sustained. Recovery levels were 65-72%, and attempts to increase recovery by changing operating conditions resulted in plant instabilities and upsets. PGB, therefore, had difficulties maintaining nominal ethane production for its customers.

Due to these operating difficulties and economic incentives, PGB decided to conduct an ethane extraction improvement (EEI) project in early 2000 to investigate alternatives for enhancing ethane production of GPP-2 through GPP-4.

EEI study

The initial EEI study was intended to assess the current operating conditions of the three GPP trains and to determine the extent of modifications required to:

Improve ethane recovery to more than 90%.

Improve the plant's ability to accommodate changes in feed gas compositions.

Improve the reliability and stability of the plant's operations.

Maintain a sales-gas output of 250 MMscfd/train at the increased ethane recovery rate.

The study indicated that the turboexpander trains could be retrofitted using one of three NGL-recovery technologies: Gas Subcooled Process (GSP), modified Cold-Residue Reflux, or Recycle Split Vapor.1 All three processes could attain the desired sales-gas production and ethane-recovery target value.

Click here to enlarge image

Table 1 shows a summary of the study results.

The technology assessment showed that the main sales-gas pipeline compressors were fed from a common header from GPP-1 through GPP-4. The GPP-1 train is a dewpoint-control plant, which means that "wet" sales gas was mixing with the otherwise "dry" sales gas from the other three plants.

The Recycle Split-Vapor process required a dry-gas stream, recycled from the discharge of the pipeline compressors to each of the three trains, to serve as additional tower reflux. Including the GPP-1 sales-gas stream meant that no dry recycle gas stream was available; this excluded the Recycle Split-Vapor technology.

An additional analysis of the remaining two retrofit options indicated that the cold-residue reflux process would require rewheeling of the existing sales-gas compressor for each train; the GSP option would not require this modification. Rewheeling would cost approximately $300,000, which did not include additional plant down time.

Considering the overall project cost impact and schedule constraints with the cold residue reflux option, PGB determined that GSP was the optimum technology that would provide recovery improvements and operating flexibility.

In addition to increased recovery, the GSP design also produced 10% more sales gas (276 MMscfd vs. 250 MMscfd) for the lean feed composition without further equipment modifications. PGB decided to proceed with the GSP retrofit design for GPP-2 through GPP-4 early in the second quarter 2001.

The final design basis conditions resulted in a GSP retrofit design with a calculated ethane recovery of 95.56% for lean feed gas and 97.46% for rich feed gas without additional residue or refrigeration compression.

Original unit design

The low-temperature separation units (LTSU), or cryogenic sections, of GPP-2 through GPP-4 are the same configuration; each was originally designed to provide a net residue-gas product flow rate of 250 MMscfd at 80% ethane recovery. The cryogenic section is a traditional single-stage expander process with supplemental mechanical refrigeration.

Click here to enlarge image

Fig. 1 shows the basic process configuration.

GSP retrofit

The advantages and basic design features of a GSP retrofit of an industry standard, single-stage (ISS) NGL recovery plant are discussed in the literature.2 3 For this project, the retrofit required three new equipment items and modifications to two existing items (Table 2).

Click here to enlarge image

Fig. 2 shows the basic process configuration for the GSP retrofit of GPP-2 through GPP-4.

Click here to enlarge image

One unique feature of this retrofit is that the plant can revert to its original ISS operating mode. PGB required this flexibility because even though NGL plant retrofits are common in the US, retrofits are a relatively new concept outside of North America. Also, the EEI project was the first major venture that PGB adopted on any of its existing facilities.

The retrofit was therefore designed so that PGB could completely isolate the GSP system if necessary and operate the LTSU in the original ISS mode. PGB installed several mode-switching valves to achieve this flexibility

(Fig. 2).

This switching ability also meant that such existing equipment as the demethanizer and turboexpander-compressor, although modified for the GSP operating conditions, still had to operate efficiently in the ISS operating mode.

Concurrent with the GSP retrofit modifications, PGB studied the design of the demethanizer side-reboiler pumps.

The GPP plants were originally configured with pumped side reboiler loops for the top, middle, and bottom side-reboiler services. Since start-up, these pumps have operated sporadically, resulting in inefficient demethanizer operations. The study concluded that PGB could shut down and completely bypass the three sets of pumps, thereby converting the pumped reboiler loops to thermosyphon circulation.

Project strategy

The EEI project was PGB's first major revamp on any of its existing facilities; therefore, an integrated management team of PGB operations and project management was assigned to coordinate the project design, procurement, construction, and commissioning activities.

The strategy was to use mostly local companies throughout the project while developing in-house technical and management expertise. Petronas' consulting company, OGP Technical Services Sdn. Bhd., executed the detailed design, procurement, construction management, and commissioning activities.

Ortloff Engineers Ltd., the process technology licensor, provided continuous process design support, technical review, and approval of various key design documents throughout the engineering and commissioning phases.

Petronas' subsidiary trading company, Malaysian International Trading Corp., provided the procurement expertise, and a local contractor was awarded the construction contract for all three retrofits.

In this "multi-contracting" strategy, PGB awarded more than 16 procurement and construction packages. By self-performing this work, PGB saved about $3 million compared to a traditional turnkey engineering, procurement, and construction contract.

Project risk assessment

Petronas' expanding global operations necessitate continuous risk-profiling assessments for business ventures that involve potential risk exposure. The EEI project team therefore conducted extensive risk-assessment exercises throughout the project continuously to evaluate risks associated with engineering, construction, and commissioning activities, as well as any other aspects that could negatively affect the project.

Two types of risk profiling were specifically adopted for the project: project risk assessment for overall risk reviews and project independent review for specific areas of concern.

Project risk assessment involved brainstorming sessions of the project team and specific discipline observers. Twelve aspects of the project were considered and evaluated, including scheduling, quality control, contractual, stakeholders, contractor, financing, technology risks, and health, safety, and environment.

In the project independent review meetings, independent parties reviewed the engineering or construction of specific project areas, usually before starting a major milestone, such as the issuance of the invitation-to-bid quotation packages or at the end of detailed engineering. The project team then closely monitored each session report; therefore, each risk ranking was reduced as the project progressed.

Six sessions were conducted during the projects' 21/2 years. This does not include the 10 internal and external audits that emphasized safety, quality, and integrity issues. Due to these activities, total change orders invoiced at the project's end were less than 5% of total contract price and 800,000 man-hr were achieved without a lost-time incident as of January 2004.

Project execution

GPP-2, GPP-3, and GPP-4 are duplicate designs. The only difference is that the GPP-2 layout is a mirror image of GPP-3 and GPP-4. This means that tie-ins, equipment locations, and pipe routings were essentially the same for all plants.

Even though a complete set of drawings was issued for each job, the duplicity of design reduced substantially the engineering man-hours that would have been otherwise expended for three separate designs.

As-built drawings are always a major concern with retrofit projects; consequently, the engineering department spent 3 months verifying the existing plant drawings. This work included underground scanning to confirm subsurface piping locations and photogrammetry to effectively verify existing pipe routings.

These were valuable exercises because early identification of any discrepancies minimized costly delays, not only for redesign costs, but also for construction costs and overall project delays. The engineering effort for all three retrofits was completed in 5 months.

Construction

The construction sequence was different for each of the three retrofit projects and depended, to a large extent, on scheduled turnarounds for each plant. The GPP-4 plant received the first retrofit.

Construction work was performed in two phases. The Phase I included the installation of all the operating mode-switching valves (which also served as tie-in valves), bypass piping for the side reboiler pumps, tray modifications in the existing demethanizer, and the new mechanical center section (MCS) for the existing expander-compressor.

The expander casing was shipped to the vendor's shop in Kuala Lumpur for minor machining to adjust flow clearances. All of this work was achieved within the scheduled, 8-day GPP-4 turnaround.

After Phase I work was completed, GPP-4 started up in the ISS operating mode. At that time, Phase II commenced, which included the installation of new GSP equipment and associated piping and instrumentation.

Phase II was performed during a 7-month period and involved work in a live plant, which required additional attention to the safety of all plant and construction personnel. Construction of the GPP-2 and GPP-3 retrofits was similar except that the sequence of Phase I and II work was reversed.

One of the most difficult aspects of the retrofit construction activities was the modification and replacement of the MCS for the existing turboexpander-compressor units. GPP-2 through GPP-4 shared a common spare MCS. Before the final shutdown of GPP-4, the spare rotor was sent to the US for rewheeling, during which time the three trains had no spare MCS available.

After the GPP-4 expander-compressor was modified, the replaced rotor assembly was sent to the US for modification. This then became the MCS for the GPP-3 expander-compressor. Likewise, the GPP-3 rotor was modified and placed in the GPP-2 expander-compressor and the GPP-2 expander-compressor MCS became the common spare for all three plants.

Retrofit commissioning

Each plant was initially restarted in the ISS mode. After stable operations were achieved, the plants were warmed up and operated in essentially a dewpoint-control mode. Then gas was routed to the new GSP system.

The basic steps in transitioning from an ISS to a GSP operating mode are:

1. Pressure up the GSP system using demethanizer overhead vapor.

2. Cool down the GSP system using cold demethanizer overhead vapor.

3. Divert all demethanizer overhead vapor into the ethane absorber and subcooler.

4. Divert all expander discharge from the demethanizer to the ethane absorber and start up the bottom pumps to return ethane absorber liquids to the demethanizer as reflux.

At this point, the plant is actually running in a modified ISS mode with the ethane absorber functioning like a separator. The final step in converting the plant to GSP operating mode is to route some of the vapor feeding the expander to the subcooler to serve as reflux to the ethane absorber.

It is critical that the operators closely monitor the cool-down rate of the subcooler during this last transition. A brazed-aluminum heat exchanger is limited to a cool-down rate of 1-2° C./min.

The operators must also gradually increase the GSP flow through the subcooler to avoid thermal shock to the subcooler and upsets to the expander-compressor when the inlet flow to the expander decreases.

GPP-4 was commissioned in March 2003, GPP-3 in August 2003, and GPP-2 in December 2003.

GPP-4 retrofit performance

After GPP-4 switched to the GSP operating mode, the demethanizer experienced flooding problems as the plant cooled down to design operating temperatures. A significant increase in the differential pressure in the column's top section (Trays 1 to 13) indicated flooding.

As the condition worsened, the ethane absorber level increased, indicating carryover from the demethanizer. Further investigation showed that the flooding seemed to occur at a consistent low-temperature separator No. 2 temperature of –52° C.

A design review of the internals indicated that downpipes from the top chimney tray to Tray 1 were inadequately sized for the new liquid rates feeding the demethanizer's top. In a GSP design, liquid loading in a demethanizer's top section is significantly higher in GSP mode than in the ISS mode when the expander discharge is feeding the tower's top.

PGB developed a redesign so that the downpipes for GPP-3 could be replaced. It was the next plant modified.

PGB must wait until GPP-4 shuts down before installing the redesigned downpipes in the demethanizer top chimney tray. GPP-4 will therefore continue operating in the GSP mode at slightly warmer conditions to minimize the risk of flooding.

The GSP design still achieved 91.7% calculated recovery with this operating constraint, based on field-performance evaluation test data.

GPP-3 retrofit performance

The downpipe sizing error discovered on GPP-4 was corrected in GPP-3 as part of tray-modification work during the shutdown.

After the retrofit, GPP-3 was commissioned and experienced the same flooding problems as GPP-4 at similar process operating conditions. The project team concluded that there must be other bottlenecks in the demethanizer's tray design that were not showing up on existing column drawings.

PGB cannot shut down GPP-3 and, therefore, the plant continues to operate in the GSP mode with warmer-than-design operating conditions. Although the modification to the downpipes did not eliminate flooding, the change did help the plant achieve a calculated recovery of 93.4%.

Continued flooding in the GPP-3 demethanizer, even after the downpipe modification, led to further investigations into the cause of the operating problems. The effort centered on discovering the cause so PGB could implement changes in the GPP-2 demethanizer.

The ensuing investigation led to two additional discoveries. First, a report issued in 1993 from the original engineering and construction consortium highlighted a similar flooding condition in the demethanizer, which appeared to be directly related to the temperature of low-temperature separator No. 2.

Second, a tower scan in 1993 indicated foaming on Tray 7, which is the feed tray for the separator liquids. Field modifications to the configuration of Trays 4-7 were designed to alleviate the problem in 1993.

Click here to enlarge image

Fig. 3 shows the changes implemented and tray configuration before the retrofit.

Trays 5 and 6 were removed and a seal pan with 2-8 in. drainpipes was installed to feed downflowing liquids to Tray 7 from Tray 4. Although Tray 7 is an active tray, two chimneys were added. All of these steps were attempts to alleviate the problem and stop carryover from the demethanizer.

There was enough uncertainty in the GPP-3 and GPP-4 demethanizers that PGB thoroughly inspected the GPP-2 demethanizer internals before the retrofit to discover what additonal constraints might be causing the flooding. PGB found hydrocarbon liquids in the seal pan directly under Tray 4. This was unusual because hydrocarbon liquids in the top of a demethanizer should be light ends and vaporize rather rapidly once the tower is warmed up to ambient conditions.

Click here to enlarge image

PGB sampled the liquid trapped in the seal pan. Table 3 shows the lab analysis of the sampled liquid.

It is interesting that aromatics such as benzene, toluene, and xylene were present. These constituents have relatively high solidification temperatures vs. typical operating temperatures of an NGL-recovery plant.

Ortloff performed a sensitivity analysis in a process simulation model by adding 0.1 mole % of benzene in the feed gas to the LTSU. Results indicated that this feed-gas composition could indeed lead to concentrations in the low-temperature separator No. 2 liquids that could cause a freezing condition when flashed to demethanizer pressure.

This issue, however, has yet to be verified as the root cause of the flooding. Further analysis is required to determine the overall impact of these aromatics in demethanizer operations.

GPP-2 retrofit performance

After inspecting the demethanizer internals, PGB implemented several additional modifications to alleviate the flooding problem. The most significant changes were:

Downpipes from Tray 4 to Tray 7 were rerouted to feed into the downcomer area of Tray 7.

Even though chimneys had been installed in Tray 7, the active panels were never blanked off. These were replaced with blank panels, which converted Tray 7 into a true chimney tray.

Drain holes were installed on the seal pan under Tray 4 to eliminate accumulations of heavy hydrocarbons as was experienced during the initial inspection.

There were also other modifications with respect to tray valve gauge thicknesses and inlet wier heights.

When GPP-2 was commissioned in December 2003, it experienced the same flooding problems that occurred in GPP-3 and GPP-4. PGB again adjusted the plant operating conditions to maintain a safe margin for flooding; the plant continues to run in the GSP mode.

As with the two other plants, the additional changes did not eliminate tower flooding; however, plant performance improved and GPP-2 achieved a calculated recovery of slightly more than 95% based on field-test data.

Process evaluation

Even with the known operating bottleneck in the demethanizer, plant performance testing was completed on all three plants.

Changes to the trays in the GPP-2 and GPP-3 demethanizers did not eliminate the flooding problem; however, the stepwise performance increase tends to support the idea that the changes helped mitigate the impact. This is further supported by the performance-test data.

Also, reducing the number of theoretical stages in the GPP-3 demethanizer simulation resulted in a reasonable match to the performance-test data, especially the temperature profile. This analysis supports the field observations that the GPP-3 demethanizer was operating at or near actual carryover and the top section was flooded, which reduced tray efficiency.

This observation also correlates well with the increased ethane recovery in GPP-2 that resulted from higher tray efficiencies in the tower's upper section due to the additional changes implemented.

Current situation

As of March 2004, all retrofits were completed and PGB was trying to further investigate the cause of the demethanizer flooding and a possible correlation with aromatic contamination in the LTSU feed gas. The most noticeable plant improvements are currently:

Improved stability in the LTSU, which allows it to accommodate changes in the inlet gas composition.

"Snowball" upsets often encountered in the ISS mode have been eliminated.

Click here to enlarge image

The plants maintain stable operations even with fluctuating concentrations of carbon dioxide in the feed gas.

Increased ethane production has provided Petronas more flexibility to sustain an uninterrupted ethane supply to its customers.

The retrofits have achieved all of the goals set forth in the original study.

References

1. Pitman, R.N., Hudson, H.M., Wilkinson, J.D., and Cuellar, K.T., "Next Generation Processes for NGL/LPG Recovery," presented to the 77th Annual Convention of the Gas Processors Association, Mar. 16-18, 1998, Dallas.

2. Lynch, J.T., Pitman, R.N., and Hudson, H.M., "Texas Plantn Retrofit Improves Throughput C2 Recovery, Goldsmith Gas Plant," presented to the 75th Annual Convention of the Gas Processors Association, Mar. 11-13, 1996, Denver.

3. Wilkinson, J.D., and Hudson, H.M., "Improving Gas Processing Profits with Retrofit Designs for Better Ethane Rejection/Recovery," presented to the Permian Basin Regional Meeting of the Gas Processors Association, May 13, 1993, Midland, Texas. F

The authors

Click here to enlarge image

Adam Abdul Rahman [email protected]) is process engineering manager for Petronas Gas Bhd., Kertih, Malaysia. He joined the company in 1991 and is currently responsible for all gas plants and facilities for Petronas' plant operation division. Rahman has worked in project management, construction, commissioning, process, and utilities operations of Petronas' gas plants. He holds a BS in chemical engineering from Lamar University, Beaumont, Tex.

Click here to enlarge image

Amy Azlina Yusof (azlinay@ petronas.com.my) is a senior process engineer with Petronas Gas Bhd., Kuala Lumpur, Malaysia. She joined the company in 1998, and has worked in process and utilities operations, commissioning, and project management for Petronas' gas processing plants and projects. Yusof holds a BEng in chemical engineering from the University of Sheffield, UK.

Click here to enlarge image

John D. Wilkinson is president of Ortloff Engineers Ltd., Midland, Tex. He has been with Ortloff Engineers as a senior consulting engineer and licensing manager since its formation in 1986 and had worked for its predecessor company, the Ortloff Corp., in a process engineering capacity since 1975. He specializes in the design of cryogenic gas processing facilities for the recovery of ethane or propane from natural and refinery gas streams. Wilkinson holds a BA (1974) in chemistry and chemical engineering and a Master of chemical engineering (1975) from Rice University, Houston. He is a registered professional engineer in Texas and Oklahoma and is a member of AIChE.

Click here to enlarge image

L. Don Tyler (Don.Tyler@ Ortloff.com) is manager of engineering services for Ortloff Engineers Ltd., Midland, Tex., a position he has held since 1999. In 1992-99, he was a project engineer for BE&K Engineering Co. In 1985-92, he was manager of engineering for Harbert Construction Corp. Before 1985, he was a mechanical engineer and piping engineering group leader for Ortloff. Tyler holds a BS (1978) in mechanical engineering from Texas A&M University. He is a registered professional engineer in Texas.