Topsides equipment, operating flexibility key floating LNG design

March 9, 1998
Use of a large-scale floating liquefied natural gas (LNG) plant is an economical alternative to an onshore plant for producing from an offshore field. Mobil Technology Co., Dallas, has advanced a design for such a plant that is technically feasible, economical, safe, and reliable (OGJ, June 30, 1997, p. 66). Presented in that article were descriptions of the general design basis, hull modeling and testing, topsides and storage layouts, and LNG offloading.
Kris Yost, J. Raul Lopez, Jin Mok
Mobil E&P Technology Co.
Dallas
Use of a large-scale floating liquefied natural gas (LNG) plant is an economical alternative to an onshore plant for producing from an offshore field.

Mobil Technology Co., Dallas, has advanced a design for such a plant that is technically feasible, economical, safe, and reliable (OGJ, June 30, 1997, p. 66). Presented in that article were descriptions of the general design basis, hull modeling and testing, topsides and storage layouts, and LNG offloading.

But such a design also presents challenges for designing topsides equipment in an offshore environment and for including flexibility and safety. These are covered in this second article.

Mobil's floating LNG plant design calls for a square concrete barge with a moonpool in the center. It is designed to produce 6 million tons/year of LNG with up to 55,000 b/d of condensate from 1 bcfd of raw feed gas (Fig. 1 [51,312 bytes]).

Economical, flexible

Although additional costs are incurred in a floating LNG plant for items such as hull and mooring, these are typically more than offset by elimination of a harbor, a pipeline to shore, a field production platform, and site preparation (Fig. 2 [27,570 bytes]).

Also, a large economic benefit is realized by reduction by 1 year of project time from front-end engineering to first production. This is possible because the floating LNG plant can be built in a harbor with a skilled labor base utilizing modular construction techniques.

The floating LNG plant can be moved from one field to another, allowing a small, marginal field to enjoy the same economy-of-scale benefits as a large field. Building flexibility into the generic plant design is important because the different fields may result in a variety of feed conditions, compositions, and environments.

The cost of providing the flexibility is minimal when planned into the original plant design. Safety is also an important design consideration because of both the normal risks associated with an LNG plant and the added concerns associated with the close proximity of equipment and personnel in an offshore environment.

With the exception of the liquefaction section, the LNG plant design discussed here applies equally to any floating production storage and offloading (FPSO) vessel conditioning gas for pipeline sales or for LPG recovery.

Fig. 3 [28,603 bytes] shows the LNG process for the floating plant.

Inlet systems

The floating LNG plant is designed to be located above the gas-production field. Small diameter flow lines from each subsea wellhead template flow directly to the plant's inlet manifold.

These relatively short and small flow lines preclude slugging as a major problem. Additionally, oversized inlet separators can control any potential slugging.

Once liquids are separated, the feed gas is cooled from a maximum temperature of 170° F. to 95° F., in preparation for gas treating. The exchangers are designed to handle a variety of temperatures and gas-to-liquids ratios (GOR).

Besides GOR and temperature, a third feed-gas variable, which can change during the plant's life, is pressure. The required process inlet pressure is 800 psig. Higher inlet pressures are choked down to this level. Lower gathering pressures, however, may be required during the field's depletion stage.

In preparation for this condition, crane-accessible space is built into the barge for future booster compression to be added for compressing the inlet gas from pressures as low as 300 psig. This retrofit should be relatively easy because it will be incorporated into the original plant design.

Condensate stabilization

The liquids from the inlet separators are sent to a condensate stabilizer tower where the light ends off the top will be compressed and recombined with the inlet gas ahead of the gas-treating section.

The stabilizer's bottom product is required to meet a maximum Reid vapor specification of 10 psia before being cooled and sent to storage. The overhead stream has no specification, but as much of the C5+ as possible should be removed.

Random and structured packing is preferred for all tower designs in the floating plant to minimize the effects of swaying and permanent tilt. The condensate stabilizer in particular has a high liquid-to-vapor ratio which allows random and structured packing to provide the additional benefit of a smaller tower diameter.

Proprietary distributors will compensate for motion and help ensure proper distribution. A small loss of efficiency is still expected for packed columns during major storm conditions but is addressed by increasing the total packing height and the number of beds in the column.

A ballasting system was also incorporated into the barge design to help minimize permanent tilt, even when offloading.

The condensate stabilizer may require the largest turndown of any tower because of the potential of different feed compositions and production rates. Feeds with higher GORs would result in smaller volumes going to the stabilizer.

Solutions include installing different sized distributors, which can be done in the field, and/or changing the tower's reflux ratio.

Gas treating

Without a specific field and gas composition, the plant is designed to handle inlet-gas compositions containing up to 15% CO 2 and 100 ppm (vol) H 2S. Gas feeding the liquefaction section must be treated to 50 ppm (vol) CO 2 and 5 ppm (vol) H 2S maximum.

Some of the criteria for selecting a treating process are flexibility, proven technology, ease of maintenance, and availability of consumables at remote locations.

An important objective in the selection for this application is minimizing the heat load required for regeneration of the solvent. Gas treating consumes the majority of the utility heat available from the plant's waste-heat recovery system.

The treating process selected is an amine process. Although other processes are available that require less heat, they do not meet all the requirements.

The amine process is flexible in treating any range of acid gas up to the 15% design limit. This is also proven technology used throughout the world and is less difficult to maintain than other processes considered.

Amine is a heat-intensive process that ordinarily requires more heat than available from the waste-heat recovery system for the 15% CO2 base case. After evaluation of various amine solvents, the BASF aMDEA solvent was found to be the most heat efficient; however, the standard process still requires more heat than available.

To solve this problem, the heat load was reduced further through the use of a split flow or two-stage design (Fig. 4 [25,473 bytes]). In this process, the circulation of a semi-lean solution, which utilizes less heat for regeneration, achieves bulk CO2 removal.

Fine CO2 removal then takes place by the circulation of a normally regenerated lean amine. Other standard methods were also used to minimize heat load such as maximizing cross exchanger area and minimizing regeneration pressure.

After establishment of a base case design, the next step was to test its flexibility with CO2-concentration levels of 2% to the maximum of 15%.

A flexible inlet gas-treating system is a key to a generic floating LNG plant. It provides the ability to move the plant to a different location and to process new gas streams when the original reservoirs are depleted.

At lower CO2 levels, where the heat load would be reduced, other criteria were considered for flexibility. The first was conversion to a standard, one-stage process. The BASF aMDEA helped determine that a single-stage configuration could be applied for CO2 concentrations up to 10%.

A standard process configuration is desirable because it reduces the complexity of the operation. It also increases lean amine circulation rates, at lower CO2 levels, so that equipment, especially towers, operate closer to their design parameters.

The conversion to a standard one-stage configuration would be accomplished by the following (Fig. 4):

  • Low-pressure flash becomes top half of stripper
  • Shutdown semi-lean amine line
  • High-pressure flash bottoms bypass low-pressure flash to lean/rich exchanger then return to top of low-pressure flash
  • Low-pressure flash bottoms bypass booster pump and lean/rich exchanger directly to stripper top.
The ability to convert to a standard amine configuration also allows consideration of a conversion to a different solvent. Some reasons other solvents may be desirable for use include cost savings, better availability, and ease of operation.

While other solvents could be used at lower CO2 levels, including the BASF aMDEA, DEA was used for this application. Given the available heat, a one-stage DEA system can be used to treat gas containing up to approximately 8% CO2.

Table 1 [12,238 bytes] lists key parameters in operating an amine train with inlet CO2 concentrations ranging from 15% to 2%. The natural-gas rate is 400 MMscfd. The process configuration and the solvent are changed at lower CO2 levels.

Major equipment

The final step in this analysis is to verify that all equipment functions properly throughout the wide range of operating conditions. The major pieces of equipment of an amine plant are its towers, heat exchangers, and pumps.

Each of these can present its own problems:

  • Towers. Three of five towers that make up the two-stage configuration have liquid rates which include both lean and semi-lean amines. These are the bulk absorber, the high-pressure flash, and the low-pressure flash.
When these towers are operated on a one-stage configuration, their liquid flow rates are reduced considerably. The limiting factors for liquid turndown in these cases are liquid distributors. If a wide range of liquid flow rates were required, the distributors would need to be modified.

Designing liquid distributors for a reasonable range of operation and then replacing them when operating conditions change drastically is a solution. As a result, distributors are designed for a 60% turndown from the base case. This is good for processing gas containing 8-15% CO2.

If the CO2 content of the gas goes to less than 8%, liquid distributors for three towers will be replaced while the other two towers remain unchanged.

  • Heat exchangers. All heat exchangers for a large amine plant require multiple shells in parallel and/or series.
If conditions were to change from the base case, exchanger shells could be added or bypassed. If all possible scenarios were identified during the design process, the proper valves and space requirements could be included.

In this application, the base-case design includes enough heat-exchanger surface area to operate for the entire range of this study without adding new shells. The only exception is the reflux condenser for approximately a 4-8% DEA process.

Operating in this range requires the appropriate number of shells be added. In other cases, either the same number of shells is utilized or some shells require bypassing.

  • Pumps. The large pumps involved in this process are the lean and semi-lean amine circulating pumps; only the lean-amine pumps require a wide range of operation.
The amine circulating pumps in this case are high flow, high head; therefore centrifugal pumps could be used for this application. These pumps provide extensive flexibility. To ensure that large pump motors are operated efficiently, three 50% pumps would be selected for lean-amine circulation.

LNG liquefaction options

Safety is an important criterion in designing the liquefaction section for the offshore environment. Most major LNG plants use a propane loop to pre-cool the treated gas and then use a mixed-component refrigerant (MR) to finish the liquefaction process.

In event of a leak, propane carries a higher risk for a potential explosion than the other refrigeration components. Minimizing its use provides less risk in the offshore environment.

Therefore, pure propane chilling loops were eliminated and only a single loop MR refrigeration process was used. Although the MR refrigerant contains some propane, the total amount of propane used is greatly reduced, and propane storage minimized.

The single-loop process has fewer pieces of equipment and is less complex, resulting in a simpler operation.

Process simulations estimate an efficiency loss of less than 0.1%, when comparing product BTU output to raw gas feed BTU value. The actual efficiency loss can vary and depends on the feed composition, as well as other design and operating parameters.

Another design consideration is the use of a spiral-wound heat exchanger vs. cold boxes.

Spiral-wound heat exchangers are not designed to contain MR pressure once the process has been down for an extended period. The light components in the MR begin to vent as pressure builds, resulting in a change in MR composition. The composition must then be adjusted after startup.

Cold boxes can be designed to contain settle-out pressure during extended plant outages and do not ordinarily require any re-blending of components. The downside to cold boxes, however, is size limitation. In this design, one spiral-wound heat exchanger could replace at least nine cold boxes.

For added safety, the floating LNG plant will not produce and store NGL other than a small quantity required for MR make up. Because of this, the fractionation system is considerably smaller than required by typical LNG plants and will only need to operate twice a year to replenish normal MR losses.

Flare system

The flare system nearly always represents a challenge on any offshore project but especially for a large floating LNG plant.

The emergency flare for this case needs to be sized for 1 bcfd. This includes the CO2 and N2 impurities in the feed gas. This volume rules out use of a vertical flare or cantilevered flare because of radiation effects and noise.

The required flare-stack height and flame size also present hazards to helicopter operations. Shielding working areas and equipment is an alternative but would be difficult because of the size of the floating LNG plant.

The solution for a large floating LNG plant is to construct a flare system away from the vessel, eliminating radiation and noise concerns. The flare line is a U-tube running underwater to prevent navigational hazards. The remote flare gas must have enough pressure to hydraulically overcome any liquids that may accumulate in the U-tube flare line.

The fuel-gas system, tank boil-off, LNG offloading, and other minor sources also produce smaller emergency flare volumes. These volumes have lower pressure and volume and are unable to utilize the remote flare.

Additionally, parts of these volumes are cryogenic and affect metallurgy design. Two onboard enclosed flares are designed to handle these smaller volumes exclusively.

Utilities

Most utility systems on the floating LNG plant are similar to the conventional land-based LNG plants. A few utility systems, however, have been designed to improve on safety, reliability, and energy efficiency:
  • Cooling system. The cooling system selected for the production barge is a closed-loop, tempered-water system, cooled by once-through seawater. Multiple pumps and heat exchangers will provide adequate turndown for the flexibility required by the floating LNG plant.
Tempered water for the closed-loop system is produced from desalinated seawater sent through a softening unit to remove remaining minerals and ions. It is pumped from a surge tank to the plate-and-frame seawater/tempered water exchangers, followed by the various process exchangers before returning back to the surge tank.

Once-through seawater cooling for all process exchangers was considered. For routine cleaning, however, this option requires a spare exchanger or extra capacity for each of the process coolers. In addition, more expensive metallurgy would have been required for the entire system.

Air coolers were also considered but proved to be too large in footprint area. In addition, with the generic barge design and the ambient temperature unknown, ambient air temperature may not be low enough for proper cooling.

  • Heating system. For added safety, a design objective of the floating LNG plant is to supply all of the heating duty requirements without the use of fired heaters. To achieve this, waste heat is recovered from the exhaust of the gas-fired turbines. This has the added benefit of increased plant efficiency.
The waste-heat recovery system uses two separate hot-oil systems. A single hot-oil system was considered but eliminated because the temperature required for the condensate stabilizer reboiler would lead to amine degradation in amine regenerator reboilers.

Splitting the system allows high enough oil temperatures for the condensate stabilizer reboiler, while supplying cooler oil to the remaining process reboilers.

Acknowledgments

The authors would like to thank Mobil E & P Technology Co. for granting permission to publish this article and the members of the Floating LNG Plant Team for developing this technology.

The Authors

Kris Yost is a process engineer in the surface engineering department of Mobil E&P Technology Co. in Dallas. He has 18 years' experience in the oil and gas industry. He holds a BS in chemical engineering from the University of Oklahoma and an MBA from Oklahoma City University.
J.R. Lopez is a process engineer in the surface engineering department of Mobil E&P Technology Co. in Dallas. He has 10 years' experience in gas processing and NGL plant operations. He holds a BS in chemical engineering from Stanford University, Palo Alto, Calif.
Jin T. Mok is a facilities engineer in the surface engineering department at Mobil E&P Technology Co. in Dallas. He has 7 years' of experience in facilities design and gas plant operations. He holds a BS in mechanical engineering from Drexel University, Philadelphia.

Copyright 1998 Oil & Gas Journal. All Rights Reserved.