Anne K. Rhodes
Refining/Petrochemical Editor
By relocating and integrating units from the Dow plant with Phibro's 130,700 b/d refinery at Texas City, Tex., and adding a new residual oil solvent extraction (ROSE) unit, Phibro will optimize its Texas refinery operations.
The dismantling, movement, and re-erection phases of the project are all but finished, and installation of piping and new instrumentation for the major relocated units is well under way. When the project is complete, Phibro will drastically reduce fuel oil production at Texas City and increase output of middle distillate.
Resid, which the company now produces in excess, will be converted to a heavy fluid catalytic cracking (FCC) feedstock. Most of this stream will be fed to the oversized FCC unit (FCCU) at Phibro's 71,000 b/d Houston refinery, thus eliminating Phibro's reliance on purchased FCC feed.
OYSTER CREEK REFINERY
During the 1970s oil embargo, Dow was in need of naphtha feedstock for its Freeport, Tex., ethylene plant. The company decided to build what Richard Duszynski, Phibro's senior vice-president of planning and business development, calls "half a refinery."
The 200,000 b/d plant comprised: a crude tower, three distillate hydrotreaters, a large resid desulfurizer (an Exxon Research & Engineering Co. Residfiner), and two sulfur-recovery units, which were built to enable Dow to process Saudi Arabian crude.
Naphtha from the refinery was sent to Dow's steam cracker, the company sold desulfurized distillate and jet fuel, and the Residfiner cleaned up the atmospheric bottoms for sale.
For economic reasons, Dow operated the refinery for only about 9 months. In fact, one of the units was never even started up. But the idled refinery was maintained in excellent condition, says Duszynski, which was one reason for Phibro's interest in the units.
THE DECISION
The driving force for this project was Phibro's desire to reduce residual fuel oil production company-wide.
Because of a lack of bottoms-upgrading capacity at the Texas City refinery, that plant produces a greater-than-average percentage of residual fuel oil. And Phibro buys all its crude and sells all its products on the spot market, which means the company's earnings are tied closely to the fluctuations in the resid market. Phibro's Houston refinery, on the other hand, has excess FCC capacity and consumes purchased FCC feedstock.
This spread relationship (long on resid and short on FCC feed) typically has placed Phibro at an economic disadvantage.
The company examined the prospects of building a coker or a high-pressure hydrocracker, but determined that neither would solve the problem entirely. And for both these cases, huge grassroots capital investments would be required at a time when new construction brings only utility-type rates of return. For this reason, Phibro began searching for alternative means of "balancing' its Texas refining system.
During this evaluation period, Dow abandoned its attempts to sell the Oyster Creek plant "in place" and decided to offer it for sale as used equipment to be relocated. Phibro had already evaluated the option of purchasing and restarting the Oyster Creek refinery, and was familiar with its processing scheme.
The mild hydrocracking nature of the Oyster Creek Residfiner was suitable for converting resid to FCC feedstock, which Phibro was lacking. And, with the combination of the Residfiner and a grassroots ROSE unit, Phibro could all but eliminate the sale of resid on the spot market by converting the product to a vacuum gas oil-type materials much less expensive step than converting resid to gasoline.
Coincidentally, at that time, Dow was looking for an outlet for by-product hydrogen. Dow's excess hydrogen is produced at its Freeport ethylene complex. At the time, Dow was burning the hydrogen as fuel, and was anxious to find a higher-value outlet for it.
Dow also owned an idled pipeline that ran from Freeport to within 3 miles of Phibro's Texas City plant site.
This combination of factors offered both companies attractive solutions to their respective problems. Dow could sell its refinery, idled since the early 1980s, and Phibro could balance its Texas refining system by converting resid to FCC feed via a combination of Residfining (using a transplanted unit) and ROSE processing. And, as an added bonus, a convenient hydrogen source was available to supply the relocated Residfiner. (A new pipeline will carry the hydrogen from Dow's existing pipeline to the refinery.)
Because the companies' goals were so well matched, in August 1993, Phibro completed the purchase of Dow's entire Oyster Creek refinery.
BUILD VS. MOVE
Phibro saved time and money by relocating idle refinery equipment and tapping into an existing hydrogen supply. It is difficult to calculate capital savings directly, says Duszynski, because new construction undoubtedly would have involved a newer technology than the Residfining process.
Instead, the cost of dismantling, transporting, and re-erecting the Residfiner and building a new ROSE unit should be compared to the cost of building a new high-pressure Hydrocracker. A comparison of only the capital requirements, however, would ignore the difference in value provided by the refining margins associated with each process.
It would be more accurate, says Duszynski, to compare the present values of the two alternatives. This results in an advantage for the Residfiner/ROSE combination of better than 2-to-1. If a strict comparison of used equipment with the construction of identical new equipment were made, however, the used equipment would show a savings of about 70%.
It is easier to make a more definitive comparison of the time required to construct new units vs. re-erect used ones. Using existing equipment saved engineering and design time, shop time for the large, high-pressure vessels, and field erection time. Engineering time also was reduced by using Dow's drawings to build foundations identical to the original ones.
The aggressive, 2-year project schedule is estimated to constitute 5060% of the time required to build a new facility.
DISMANTLING, TRANSPORT
Dismantling began at Oyster Creek in first-quarter 1994, and is essentially complete. The major units transported to Texas City are the Residfiner and sulfur-recovery units.
Phibro is serving as general contractor for the relocation project. Matrix Engineering Inc. is responsible for relocation engineering and overall engineering coordination at Texas City and Oyster Creek. Belmont Inc. will perform the re-erection of the 63,000 b/d Residfiner.
Litwin Engineers & Constructors Inc. will perform turnkey engineering and construction of the 30,000 b/d ROSE unit, licensed by Kerr-McGee Corp. Re-erection of the sulfur-recovery units will be performed by Petrofac Inc.
The dismantling process involved several steps:
- Hydrocarbon gases were removed from all units.
- All equipment was bar-coded to produce an accurate inventory.
- All equipment was inspected either before or during dismantling.
- Insulation, small-bore piping (2 in. or smaller), and other unsalvageable items were removed and, where possible, sold as scrap.
- Large-bore piping and associated equipment were removed. (A special tool was used to cut and bevel the pipe simultaneously, so that it could be spooled easily.)
- Vessels, pumps, exchangers, and other hardware were removed.
- All structural steel and support materials were removed.
- The equipment being moved to the Texas City refinery was loaded and transported.
When all equipment is removed, the site will be cleared.
During dismantling, Phibro worked to cross-check that Dow's equipment drawings were accurate and complete. Pictures of the equipment also were taken before dismantling. All instruments, valves, compressors, and pumps are being refurbished before reuse at Texas City.
Fig. 1 shows the Residfiner being dismantled at Oyster Creek.
During the dismantling operation, foundations and other structures were being built at the Texas City site so that, as major equipment was removed from Oyster Creek, it could be transported and immediately re-erected in Texas City. This system enabled Phibro to avoid double-handling of the major equipment.
Because a heavy crane was needed for both the dismantling and re-erection processes, Phibro chose to perform the two processes, along with equipment transportation, in immediate succession. The timing of these events had to be orchestrated carefully to minimize both the cost of renting the crane and the time required to move it between locations and set it up.
While minor changes always occur in an operation of this size, project vice-president Richard McCalla says there were very few unforeseen problems or surprises. Phibro's previous experience moving idled units, although on a much smaller scale, proved very valuable in planning and executing this project.
CONSTRUCTION
At the new construction site, the process just described is taking place in reverse order.
The foundations, major vessels, and most of the structural steel are already in place. Re-erection of the major vessels on their new foundations was completed in January 1995.
During the construction phase, all equipment is being inspected again and the units will be tested before start-up begins.
After completion of the re-erection work, all piping and new instrumentation will be installed. At the same time, construction of the new ROSE unit will begin. (The ROSE unit is considered the critical path item, and is expected to be ready for start-up in first-quarter 1996.) In the meantime, the towers will be reinsulated.
Because the construction site is outside the battery limits of the Texas City refinery, disruptions to operations will be minimal. A plant turnaround is being planned for late 1995 to tie in the new units and fully integrate the added processes.
To tie in the new units, each crude unit will be shut down separately so that second-stage desalters can be installed, and so that the necessary piping and valves can be connected. After the crude units are restarted, they can be operated as before, but will have the capability to divert atmospheric or vacuum resid to the new units when they are ready.
The existing visbreakers at Texas City will be shut down. Some of the equipment will be used to provide additional vacuum tower capacity for processing hydrotreated atmospheric resid.
CRUDE PRODUCTS
Some of the existing metallurgy in the Texas City crude units, along with the current sulfur plant capacity, limits the refinery's crude slate to blends containing about 1.0 wt % sulfur or less. And visbreaking is the refinery's only bottoms-upgrading process. These factors result in an economic incentive to process lighter, sweeter crudes.
By the time the new units start up, the metallurgy of the crude units will have been modified such that the crude oil sulfur level will no longer be an operational constraint. Phibro plans to replace all of its light, sweet crude slate with moderately heavy, sour crudes.
The Residfiner/ROSE combination will provide a high degree of bottoms upgrading, which will enable Phibro to run about 95% of the world's crudes, according to Duszynski. Although Phibro has not yet secured commitments for a crude source, Middle Eastern, South American, and some Californian crudes are under consideration. The company also plans to process an additional stream of atmospheric resid.
To handle the extra sulfur removed from the feeds, sulfur plant capacity will be increased to 600 long tons/day (It/d), with 100% redundancy. The COPE oxygen-enrichment process, licensed by Goar, Allison & Associates Inc. and Air Products & Chemicals Inc., will be installed during the expansion.
A flow diagram of the new portion of the Texas City refinery is shown in Fig. 2. (122620 bytes) The ROSE unit will be placed downstream of the Residfiner. Crude distillation capacity will increase by 10,000 b/d as a result of the project.
The added downstream capacities mean the refinery will be able to boost total output by 20,000-30,000 b/d. This scheme will produce no fuel oil and only 3,000-6,000 b/d of ROSE pitch.
Table 1 (11126 bytes) shows how yields will change as a result of the project. After start-up, the gasoline/distillate split will be about 50/50, and production of heavy fuels will decline dramatically. Production of a vacuum gas oil-type material will increase to provide feed for the FCCU at Phibro's Houston refinery.
The project will not have much effect on Phibro's ability to make "clean" fuels. The Texas City refinery will continue to have ample hydrotreating capacity for making low-sulfur diesel after start-up of the new units. And, other than a little help in meeting the sulfur constraint, the project will have no impact on Phibro's reformulated gasoline production.
OPERATIONAL CHANGES
Changes in existing refinery operations will occur mostly in the crude units, visbreakers, and the FCCU.
The existing crude units constitute a standard, self-contained, atmospheric/vacuum fractionating system, the products of which feed other downstream units. After completion of the project, the crude train will be a three-stage operation, with the atmospheric fractionation feeding the Residfiner, which provides feed for vacuum fractionation, which in turn produces feed for the new ROSE unit.
Even with these changes in the process, the operating parameters of the atmospheric and vacuum units will change very little.
Both existing visbreakers will be shut down permanently. One will be revamped, converted, and used as a supplemental vacuum fractionator. The other will remain idle.
In the future, the FCCUs at Phibro's Houston and Texas City refineries will be fed hydrotreated feedstock; whereas today they typically process virgin feedstock. There will be little if any change in FCCU operating parameters, but an increase in conversion and product quality is expected.
Fig 3 (119636 bytes) illistrates the integration of Phibro's Texas refineries after completion of the project.
UTILITIES
The added units will increase electricity requirements at the Texas City refinery from the current level of 35 mw to 62 mw after the expansion. Of this increase, 7 mw will be provided by existing excess cogeneration capacity in the refinery. The remainder of the power will be purchased from outside the refinery.Steam requirements will not increase substantially, says McCalla, as steam will be a by-product of the waste heat from the new units. A new 60,000 lb/hr boiler will be installed as part of the expansion, however, to supplement existing boiler capacity in order to meet increased steam demand during start-up.
Water requirements will increase by about 500 gpm after the expansion. The additional water is needed for increased water requirements for desalting, cooling tower makeup, boiler feedwater, and process water makeup. Water supply will be expanded by debottlenecking the existing river-water clarification system and boiler feedwater preparation unit.
PROCESS WASTES
Waste water treatment requirements will increase by about 500 gpm after the expansion. This increase will be accommodated by minor changes in the existing refinery-in particular, by recycling to process water makeup some of the water that now enters the water treatment plant. The water treatment plant also will be debottlenecked to increase processing capacity.
All new furnaces will be equipped with low-NOx burners to reduce air emissions. As another means of reducing air emissions, many tank seals in the existing plant will be upgraded to double seals. This also will provide an offset for some of the increased volatile organic compound (VOC) emissions resulting from installation of the new equipment.
In addition, the high degree of redundancy being added in the Claus and tail gas units will avoid excess air emissions as a result of equipment failures.Solid waste quantities from the waste treatment plant will increase slightly for two reasons: 1) as a result of increased wastes from the desalter, caused by the processing of additional residual feedstock; and 2) because the amount of equipment feeding the waste water treatment plant will increase. Disposal of spent catalyst from the Residfiner also will increase off site solid waste disposal.
REGULATIONS
Any time new equipment is added at a U.S. refinery, regulatory reporting requirements increase. Duszynski says the Residfiner/ROSE unit will necessitate additional reporting to the U.S. Environmental Protection Agency, and to the Texas Natural Resource Conservation Commission.
In addition, several additional reports will be required, all under the New Source Performance Standards umbrella:
- Sulfur recovery unit SO2 excess emission report, Subpart J (quarterly)
- Waste water system VOC emissions report, Subpart QQQ (semiannually)
- Equipment VOC leaks report, Subpart GGG (semiannually).
PREPARATIONS
Phibro is in the process of preparing its existing units for the new operation:
- The crude unit metallurgy will be changed during the scheduled turnaround.
- The Texas City FCCU is being revamped to increase capacity and conversion.
- UOP-licensed catalyst coolers and new nozzles are being added to the Houston refinery's FCCU.
Phibro also began training operators at the beginning of this year. This means that, at start-up, the unit operators will have a year's training on the new units. In addition, maintenance and inspection personnel are monitoring the erection of the new units, and thus will be familiar with them when they are started up.
Phibro considers these to be worthwhile investments that will help the start-up procedure go smoothly.
FUTURE
Phibro is investigating outlets for the ROSE pitch from the new processing scheme. Two options under consideration for processing the pitch are conversion to hydrogen and conversion to methanol.
Because hydrogen supplies from the existing catalytic reformer and the new pipeline are expected to be adequate for the foreseeable future, the methanol option may be preferable. It is also more expensive, however, and Phibro has not yet made a final decision on the disposition of the ROSE pitch. Phibro's ambitious plan for optimizing its Texas refining system is well under way. The effect of the project on the company's margins, however, remains to be seen. But, says Duszynski, Phibro is expecting a rate of return well above 20%. In addition, Phibro's overall earnings volatility will be greatly reduced.
In these times of low margins and scarce capital, innovative schemes such as this may become the norm in the refining industry.
Copyright 1995 Oil & Gas Journal. All Rights Reserved.
