Dutch North Sea platform expands on conventional glycol scheme

• The L9 production platform is rolled out to the barge. Visible are the 1.6-m high beams of the cellar deck that extend to the staggered lay-down areas, along with the 24-in. export pipeline curving to the subcellar deck.
• The glycol contactor (upper half shown here) is 2.6 m x 10 m, weighs 87 tons, and integrates with an absorber with structured packing and a separator in the bottom part.

The design minimized total costs (capital and operating) and environmental impact while at the same time maximizing safety.

NAM operates 21 gas-production platforms in the North Sea. In 1995, the company awarded Stork Engineers & Contractors a contract for the design and engineering of the L9 platform, expressing a desire for an approach that included at least 98% availability and unmanned operation.

Although it was clear from the outset that the basic process would be a conventional tri-ethylene glycol (TEG) gas-drying system with combined export of dry gas and condensate, Stork felt considerable improvements were possible.

Subsequently, the substructure of the wellhead platform was installed in December 1996 so that drilling could be started. The jacket of the production platform was installed in August 1997, and the topsides of both platforms, interconnecting bridge and vent stack in April 1998.

First gas was produced on June 14, 1998, all systems performing as expected.

Main process

It consists of a 12-slot wellhead platform bridge linked to a production platform with the single-train, gas-treatment process and a 20-bed accommodation module.

The L9 field is located some 80 km north of Den Helder.

The export pipeline, which ties into the Northern Offshore Gas Transport (Nogat) trunkline to Den Helder, was also part of the contract. The development of the L9 field also required an extension to the existing Nogat gas-treatment plant at Den Helder. This plant, designed by Stork's Amsterdam office (formerly Comprimo), required special provisions to prevent plugging by wax as the L9 condensate has an unusual high cloud point of 5-10° C.

Well fluids from the wellhead platform are routed via the production header across the bridge to the processing facilities on the main platform (Fig. 2 [91,503 bytes]). Fluids are first cooled to about 35° C. in the gas cooler, which is a single-pass shell-and-tube heat exchanger.

After gas and liquids have been separated in the production separator, the gas is dried to water dew point specification in the glycol contactor by countercurrent absorption in TEG.

Separation of water and condensate occurs at train pressure (120 bara) because liquids would be too finely divided after choking over level-control valves. Therefore, the liquid from the production separator flows via an open connection by gravity to the water/condensate separator.

This is a parallel-plate separator in which weirs control the interface level, a design by Stork and proven on the previous L15 platform for NAM. Water flows via an external line to three water weirs of different heights in series.

With valves in these lines, the interface level can be changed from outside the vessel. Produced water is passed via the flash vessel to a skimmer before being discharged to sea.

Condensate from the water/condensate separator goes to the export pipeline for transport to shore together with the dried sales gas.

The water content of the condensate is continuously monitored. When it is too high, the condensate is automatically diverted to the condensate tank from where it is pumped back to the inlet of the gas cooler.

The condensate tank and recycle pump have three functions:

1. Recycle off-spec condensate, during start-up, for example
2. Recycle condensate recovered in the skimmer
3. Recycle water and condensate from the glycol regeneration.

Glycol regeneration; vapor recovery

In the glycol contactor, aromatics such as benzene, toluene, ethylbenzene, and xylene (BTEX) dissolve preferentially in the TEG. Therefore, the glycol regenerator overhead vapors that are condensed in the steam condenser also have a high aromatics content.

These water and hydrocarbon condensates are recycled via the condensate tank to the inlet of the gas cooler. Here the BTEX from the condensed water will partly dissolve in the combined hydrocarbon stream because this has a lower aromatics content. This results in less aromatics discharged with produced water to the sea.

Measured aromatics contents of atmospheric samples are given in Table 1 [18,543 bytes].

The glycol pump recirculates a constant flow of 12.2 cu m/hr; therefore, there is a nearly constant flow of vapor from the glycol-regeneration system. This residual vapor (off-gas) also has a high aromatics content.

After removal of the condensables, the vapor from the off-gas drum (13.7 cu m/hr at 35° C.) is recovered by a small reciprocating compressor (2.5 kw) and is subsequently routed to the fuel-gas system, together with flash gas from the glycol flash vessel.

The off-gas compressor recovers an estimated 180 tons/year of gas and at the same time eliminates the main emission of benzene to the atmosphere.

All rotating equipment essential for gas production has been provided with spares; for this compressor currently only space for a spare unit is reserved.

Power, heat

All the normal cooling is done by a closed loop cooling-water (CW) system circulating fresh water with antifreeze (Fig. 3 [82,712 bytes]). The CW expansion vessel is pressurized by nitrogen. This excludes oxygen so that there is no corrosion in the carbon-steel system and provides leak detection by monitoring the nitrogen pressure.

For the initial cooling duty of 32 MW, there are three CW pumps and three seawater (SW) pumps that cool the three intercoolers. These titanium-plate coolers present three times 50% duty.

The main cooling duty is initially the gas cooler of 28 MW. If in the future gas compression is installed, the addition of one set of pumps and cooler will make it a four times 33% spared system with a total cooling capacity of 48 MW.

Power is supplied from two gas engine-driven generators. For start-up and emergency use, two diesel-engine generators are installed in separate rooms behind firewalls.

Each generator (1 MW each) is rated for 100% of the required electrical duty. This is done to have a highly reliable electricity supply, so that all drivers on the complex can be electrical, including the firewater pumps (via fireproof cables) and the two electro-hydraulic cranes.

The only other diesel drive is a small, hand-started air compressor for initial starting of the engines.

The two firewater pumps are horizontal on the cellar deck and operate in series with the submerged seawater pumps. The FW pumps can be tested via their automatic recycle check valves to the seawater header or via a metered test line to sea.

In case of fire, the SW inlet to the intercoolers is closed to divert all the seawater supply to the FW pumps.

All four generator engines are cooled via the closed loop CW system. The diesel engines have a second plate cooler in their jacket-water circuit which, in case of fire, is cooled with seawater.

The gas engines of the generators have a second plate heat exchanger in the jacket-water circuit that provides heat to the living quarters via a hot water loop.

Despite the additional diesel-driven generator, the cost saving over a conventional system with diesel-driven submerged FW pumps, diesel-hydraulic cranes, and air-cooled generators, is in excess of 1 million EURO (approximately $1.1 million U.S.).

There is always heat input into the working CW system from the engine cooling. Therefore, the temperature of the CW supply can be controlled at greater than the hydrate temperature of the gas by a bypass around the intercoolers.

This eliminates the need for gas-temperature control in the main process: The gas cooler is always on maximum CW flow. On start-up, the gas cooler can even be used to heat the gas to greater than the hydrate point.

The temperature in the return header is a function of the total cooling duty; a temperature controller here switches sets of intercooler/CW pump/SW pump on or off as required.

The sea water runs uncontrolled at full flow through the intercoolers in order to prevent fouling. The intercoolers have individual outlet lines down to sea level to recover part of the static head required to pump the seawater to the cellar deck.

Control; layout

Therefore, there are automatic sequences for starting the utilities and the glycol system and for opening the wells. Only the required gas flow is set from the central control room; such other variables as optimum train pressure are calculated in the control system.

The wellhead topsides (625 metric tons, three decks) were installed over the well-bay part of the lower deck. This part of the lower deck had been integrated with the jacket (525 metric tons), providing a permanent deck for early drilling access and saving one offshore lift operation (removal of temporary drilling floor).

This installation method of the wellhead platform triggered the choice of four skirt piles, swaged into sleeves at the bottom frame of the jacket.

Bridge linked to the wellhead platform is the main production platform which has six legs and four decks plus a helideck. The 4,500 metric ton topsides are a portal frame construction so that there are no braces except for two on the southern cantilever that support the 20-bed accommodation module and the control module plus helideck and a 40-man "free-fall" lifeboat.

The braced production jacket (1,000 ton) features vertical legs, extending up to sub-cellar deck level and is supported by six main piles. The stabbing point of the topsides is just below the lowest deck elevation.

This enabled fabrication of the topsides from sub-cellar deck level as a single lift structure, without the need for deck-leg extensions.

One of the middle legs supports the pedestal crane, the other a tubular vent stack that extends 74 m above the top deck.

In the lay-out, space has been incorporated for future facilities, such as main gas compressors, water injection, separators of satellite fields, and methanol-recovery distillation.

A 100-cu m methanol tank, sized for these future developments, is integrated into the cellar deck. Three of the six 1.38-m diameter main columns of the topsides are used for storing diesel oil.

The utility area with the diesel generators and sea water and firewater pumps is separated from the process area by a blast wall over the full width of the platform.

Export pipeline

When this Nogat line was laid, it had been provided with a number of split tees welded on the pipe.

Thus, it was possible to make an underwater hot tap for the L9 tie-in by drilling through such a tee without any underwater welding. The valve arrangement installed at this tie-in also includes a branch for a future 16-n. pipeline.

The Authors

Leo Festen is a process engineer with Stork Engineers & Contractors, Schiedam, The Netherlands, the company that resulted from the merger of Stork Protech and Stork Comprimo. Festen has 20 years' experience in the design of facilities for the chemical and oil and gas industries. He worked in Stork offices in The Netherlands, the Middle East, and Malaysia. Festen holds an MS in chemical engineering from Groningen State University.

Arjan Gerritse has been a process engineer with Stork Engineers & Contractors B.V., Schiedam, since 1995 and worked on the NAM L9 project during the basic and detailed design phase. He has previously worked at DSM Research, Geleen, and holds an MS in chemical engineering from Delft University of Technology.

Sandor Schrameijer has been a process engineer with Shell International Exploration & Production B.V. since 1990. He has been seconded to Petroleum Development Oman and to NAM, The Netherlands, and worked as process and project engineer. His last job was mechanical/process engineer on the L9 gas-development project. Schrameijer holds an MS in chemical engineering from Delft University of Technology.

Copyright 1999 Oil & Gas Journal. All Rights Reserved.