WESTERN GAS REVAMPS FORMER EPNG MIDKIFF PLANT

Bruce D. Portz, Tony Marques
Western Gas Processors Ltd.
Denver

Western Gas Processors, Denver, has completed a $10.5 million reconditioning of the Midkiff, Tex., gas plant and gathering system after purchasing it from El Paso Natural Gas Co. (EPNG).

The plant is currently running at about 60 MMcfd, approximately 80% of its new 70 MMcfd capacity, and recovering NGL slightly better than initial simulations projected. Because of improvements to the handling of compression liquids (Table 1), the volumes of liquid being produced are also higher than anticipated.

And the modernized plant is consuming about half the fuel of its predecessor.

PURCHASE STRATEGIES

In summer 1988, Western and several other companies were invited to bid for the purchase of the 130 MMcfd Midkiff plant and associated gathering system near Midland, Tex.

The sale came as a result of EPNG's program of reducing its merchant function and concentrating on the transportation of gas as its primary company objective.

The Midkiff plant is located 4 miles east of Midkiff, Tex., about 40 miles southeast of Midland. The Spraberry gas-gathering system, which feeds the plant, covers a four-county area surrounding the plant site (Fig. 1).

The plant itself was a 1950s-vintage installation and, as offered, consisted of approximately 30,000 hp of compression; an ambient-temperature, lean-oil absorption and fractionation facility; a substantial boiler plant with power generation; feed-water treating; a large cooling-water system; and above and below-ground product storage.

The field equipment consisted of four substantial compression sites and approximately 1,100 miles of gathering lines.

In spite of its age, the existing facility achieved reasonable liquid-recovery levels but at the expense of excessive manpower and utility consumption.

In order to maximize its bid to El Paso, Western evaluated ways of enhancing the value of the facility. This essentially meant looking for ways to maintain or improve its liquid-recovery levels while improving its overall operating efficiency.

Engineering at Western was requested to come up with a modernization plan, schedule, and cost estimate to achieve this goal. Three areas for improvements were investigated by engineering.

The first was the installation of a modern turboexpander plant. This would maintain the liquid-recovery levels, while reducing the level of operator attention required in this area of the plant.

Replacing the lean-oil plant, it was discovered, essentially eliminated the need for steam in the plant. Eliminating steam required the installation of engine driver generators, but this trade off appeared reasonable.

The second area for improvement was modernization of the the plant compression shutdown and control systems in order to reduce the amount of operator observation required, while increasing the safety and reliability of the compressors.

The third area was to produce a demethanized mix of NGL product instead of fractionated products. The driving forces behind this option were that additional capital would have to be expended to replace or integrate the existing liquids-fractionation facilities into a cryogenic plant.

Also, it would be possible to shutdown the bulk of the plant's liquid storage and batch-liquids shipping systems. This would be made possible by existing pipeline connections at the plant.

EXPANDER CHOICE

A replacement for the leanoil plant was the first section of the project considered. Both a straight refrigeration plant and a turboexpander plant were considered.

As far as liquids recovery was concerned, Western felt that ethane in the mid 70% range was reasonable with C3+ running to 99%+. The resulting refrigeration-plant option projected recovery of about 50% of the ethane, 70% of the propane, and 94% of the heavier components.

The expander version was capable of ethane recoveries of 90% with 99.5% recovery for the heavier components. Despite the recovery advantage provided by the expander plant, its higher installed and operating cost made it less than the obvious choice.

Nonetheless, the expander plant option was finally decided upon in anticipation of higher long-term liquid prices. After this decision, the next consideration was establishing how this expander plant could be integrated into the existing facility.

As purchased, compression throughout the Midkiff system was primarily set up to boost feed gas through the plant without residue compression.

About half of the feed to the plant was compressed at four field sites to approximately 150 psig then fed to the main Plant B compressors which boosted it to about 800 psig.

The other half of the feed entered the plant at nearly atmospheric pressure and was compressed in three stages to 800 psig by compressor Train A. In addition, two smaller compressors were provided for the recycle of a small flash-gas stream generated in the lean-oil regeneration loop.

The mechanical limitations of the main plant compression and certain other equipment in the lean-oil plant set the first constraint on the potential expander plant. The maximum allowable operating pressure (MAOP) of certain vessels in the old plant was 880 psig, with the compressors limited to discharge pressures in the 800-psig range.

The feed to the expander plant would be limited to these pressures.

Likewise, Western was compelled to pick an expander outlet pressure that would allow the single-stage "B" compressors to boost residue gas back to the 820 MAOP required by the pipeline.

With these pressure constraints, we had built an envelope in which the expander plant would have to operate. The actual layout of the compressors made the proposed operating arrangement simple to incorporate.

With gas as rich as that which feeds Midkiff, it was also apparent that the expander plant would require refrigeration assist. This load was easily satisfied by use of the two former flash-gas recompressors.

This application was interesting in that the flash compressors are driven by two-cycle engines and are more or less fixed speed-and-load machines. In other words, two machines were found that could each provide 800 hp worth of refrigeration and would consume more or less the same amount of fuel whether utilized fully or not.

The decision was made to limit the process to being supported by one machine with the second serving as a 100% spare. With these operating constraints imposed on the process, modeling the operation and determining just what recoveries could be achieved proceeded.

The result of this process modeling was a potential expander process that operated within the existing compression capabilities with ethane recoveries in the 75% range and C3+ recoveries running 99%+.

This was less than would have been possible if a grassroots expander plant had been installed but very satisfactory when the efficient reuse of the existing compression was considered.

SUCCESSFUL PURCHASE

Having reached this point in analysis of the project, engineering was able to develop and present to Western's management the proposed modernization plan with its associated cost estimate and schedule.

With this information, Western moved forward with the bidding and negotiation process with EPNG through the last half of 1988. Western was selected as the final bidder for the plant in February 1989, successfully completed negotiations for the purchase in May 1989, and took over operations of the facilities on July 1, 1989.

When Western received the news that it had been selected as the final bidder for the Midkiff plant, engineering solicited several contractors for bids to fabricate and install the cryogenic section of the plant.

Recoveries and other operating constraints were specified to ensure that the bids supplied would fit into the existing plant.

The solicited engineering and construction (E&C) firms were also informed of Western's search for a suitable surplus plant and were offered reimbursement for their efforts should that search prove successful.

As the contractors assembled their bids, Western pursued its search for a suitable used plant. The company felt it likely a new plant might suffer from a lengthy schedule due to the long delivery of a couple of its key components.

Over the next few weeks, several potential surplus plants were reviewed and the more likely candidates visited.

Of these, three or four would have sufficed with some modifications. But for the most part, the plants visited were block mounted and did not lend themselves to a timely or cost-effective relocation.

One candidate, however, proved not only a fairly close fit but was skid mounted and included other supporting systems that Western would have had to provide at Midkiff.

Required modifications would include adding mildsteel, shell-and-tube exchangers, rewheeling the expander, changing the demethanizer, and adding a fresh charge of molecular sieve.

Western also specified other systems that would have to be added to the Midkiff facility. Fig. 2 highlights changes in the system. Fig. 3 shows a view of the B feed compressors, demethanizer tower, and other new facilities.

One such system was amine treating.

CO2, H2S CONCERNS

Feeds to the Midkiff plant contained small amounts of CO2 and traces of H2S, which had not caused a problem for the old lean-oil plant.

Western was concerned about the potential problems larger quantities Of CO2 could cause in the cryo plant and the fact that the H2S content of the feeds was slowly increasing.

CO2 had a small potential to cause freezing problems in the cryo plant, but Western's greatest concern was for its accumulation in the liquid products causing them to go off specification.

The addition of an aminetreating system would not only solve potential processing problems, but allow Western to process gas that would have formerly been too sour. The company elected to treat the feed gas with diethanolamine (DEA) and converted a lean-oil absorber in the old plant for this purpose.

Eventually both gas and liquid treating were installed so that pigged and compression knockout liquids could be treated.

Upon installation of amine treating, it was also found necessary to install a tail-gas incinerator to satisfy the Texas Air Control Board and eliminate the venting of H2S. Cryogenic plants normally require molecular-sieve dehydrators, and in larger plants these are often proceeded by some sort of glycol system.

At Midkiff, there were already two glycol contactors in the old lean-oil plant that could easily be made use of. Western had determined that the surplus cryogenic plant under consideration was provided with molecular-sieve beds somewhat smaller than would have otherwise been required.

As a result, the company decided to reuse the existing glycol contactors and keep the on stream time of the surplus molecular-sieve dryers at a reasonable level.

The decision to eliminate steam from the plant made it necessary to install a new glycol-regeneration system that would work with something other that steam as its heat medium.

Installing a new regeneration system also served to replace badly worn and poorly located equipment. Western ultimately purchased both a glycol-regeneration system and a hot-oil system which provided heat for the glycol skid as well as an inlet liquid-stabilization skid.

By the time Western personnel had visited several of the potential surplus plants and ranked them, the bids from the E&C companies were being received. This served to reinforce the perception that the schedule would be better served by locating a suitable surplus plant.

The heart of any modern cryogenic plant is its brazed aluminum exchangers, and the delivery of this item at approximately 40 weeks was just unacceptable.

In spite of the appeal of a new plant and the costs to be incurred in relocating a surplus plant, Western elected to pursue the purchase of the favored surplus plant.

In time, Western became owner of Texaco's Burleson, Tex., plant.

In mid-September 1989, potential contractors were called to Burleson and invited to submit bids for relocation and re-erection of the plant at Midkiff.

Due to contractual requirements, Texaco was required to abandon the Burleson plant site by early November 1989. This tight time frame allowed only 4 weeks for the relocation effort.

By late September a vender was selected and the move begun. The disassembly and move to Midkiff was completed within the required schedule with most of the equipment being staged near its final resting place and items like the demethanizer being sent directly to vendor shops where modifications were performed.

MODIFICATIONS

It was necessary to perform a few modifications to the Burleson plant in order for it to meet Western's needs.

Some of the modifications included increasing the top diameter of the demethanizer and increasing the capacity of its internal liquid distributors. Also converted were two former NGL-pipeline pumps to amine charge-pump service.

In order for the cryogenic plant efficiently to handle the rich Midkiff gas, two shell-and-tube exchangers had to be added. Since these exchangers were installed in the warm end of the plant, they were constructed of mild steel.

Higher liquid rates required addition of a third larger capacity demethanizer bottoms pump.

Three fin-fan coolers, formerly used as discharge coolers for propane, feed, and residue compressors, had been obtained with the Burleson plant.

In the new operation it was necessary to dedicate all of these units to the propane refrigeration system. When installed in series, the three units just fit the required propane condensing load.

The most costly and time-consuming modifications involved the expander itself which was rewheeled along with a spare center section that came with the Burleson plant.

Also purchased with the Burleson plant was equipment that would serve for dehydration of mole-sieve regeneration gas, acid-gas removal from the Midkiff inlet liquids, and enough high-pressure storage to allow reasonable operation at Midkiff.

In addition to the surplus equipment moved to Midkiff, other supporting systems purchased included:

• Inlet liquids stabilization

• A hot oil system which supported the glycol unit and stabilization

• A glycol regeneration system

• Tail gas incineration

• NGL pipeline pumps

• Discharge coolers for the expander recompressor and residue compressor

• Generation facilities.

In the interest of increased safety for both man and machine and increased performance for its compressors, Western added Altronic shutdown panels for all of the compressors involved in the project.

These panels were set up to monitor jacket-water temperature, oil pressure and temperature, discharge temperature from each cylinder, and to provide automatic shutdown and callout in the event of upsets.

Construction was started at the Midkiff plant in November 1989. Because of the scope of the project and the original project schedule providing for an early April 1990 start-up, work was carried on for the majority of the project with a 6-10 hr day work week with a work force of approximately 150 people.

Also, because of the project time frame, the bulk of all design, drafting, purchasing, and materials control was carried out from the Midkiff construction office.

Because construction was being conducted at an operating facility, among the company's most important goals were the desire to ensure that work was carried out in the safest manner possible, to avoid any unnecessary plant shutdowns, and to reduce as much as possible interference with day-to-day plant operations.

These goals were achieved by way of frequent safety meetings, daily issuance and enforcement of hazardous work permits, safety watches and inspections, and the communication of the possibility of a developing hazardous situation between operations and construction staffs.

PROBLEMS

During prestart-up commissioning, two major problems arose.

The first was uncovered during inspection of the high-pressure lean oil absorber that was being converted to serve as the amine-gas contactor. The vessel and its trays were found to be severely fouled with a sludge composed mostly of heavy hydrocarbons and iron sulfide.

This situation required the absorber to be steamed, chemically cleaned, and finally given a tray-by-tray water wash before it was deemed suitable for service.

The second major problem involved one of the brazed aluminum exchangers that had come with the Burleson plant.

It had been an early concern that the brazed aluminum exchangers might be fouled with sieve dust or some other contaminant and therefore restrict flow through the plant. Western installed a second dust filter on the mole-sieve outlet line to help prevent this in the future.

During the nitrogen leak/pressure test of the cryogenic skid, a leak between two passes in the small brazed aluminum exchanger that served as the reboiler for the demethanizer was discovered.

The leak was from the high-pressure side to the bottom reboiler pass. No damage was apparent from the outside of the exchanger with its insulation installed.

When the insulation was removed, however, a half- inch bulge spread out over about 2 ft of one corner indicated damage inside. Apparently the exchanger had been damaged by internal freezing of water which had allowed a short circuit to develop between the high-pressure feed stream and a demethanizer reboiler circuit.

Texaco's turn-down operation at Burleson had given no indication of the presence of a problem. The damage was barely perceptible from the outside of the exchanger but was confirmed by an X-ray examination of the internals.

The type of damage prevented the exchanger's being repaired and forced Western to seek a replacement.

When the damage to the small brazed aluminum exchanger was discovered, operators isolated, tested, and examined the larger unit. It was in good shape and suffered no leaks between passes.

In addition, this opportunity was taken visually to reinspect all the nozzles which were found to be clean.

For a replacement for the damaged exchanger, a relatively large and compatible surplus unused aluminum exchanger was located in Houston.

After a rerate, thorough pressure test, and examination, the replacement exchanger was carefully strapped to a truck and shipped to the Midkiff plant site where it was spliced into the process skid. The resolution of this problem forced a delay of approximately 1 month in Western's planned start-up date.

The shutdown of the old equipment and tie-in of the new equipment at the Midkiff plant was accomplished during a 4-day turnaround during late May of 1990. Overall, work during the turnaround went very well and was completed in 1 day less than what was scheduled.

Plant start-up was delayed by approximately 24 hr due to problems with keeping the electrical generation equipment on line. Once the electrical problems were resolved to the point that we could keep the plant on line, startup proceeded fairly well.

Copyright 1991 Oil & Gas Journal. All Rights Reserved.