Flexibility, efficiency to characterize gas-processing technologies

An example of the new generation of gas plant, Amoco Production Co.'s Jayhawk plant in the Hugoton field, southwest Kansas, started up in late 1998. The plant's fractionation train and de-ethanizer and de-propanizer towers (shown here) have given the plant the flexibility characteristic of more recent design philosophies. (Photograph by Douglas R. Childs, ABB Randall, Houston)

These technologies have arisen to allow a much more rigorous optimization of several variables: operational flexibility, CO₂ tolerance, and capital and operating cost.

The selection of an optimum process will depend on conditions and composition of the inlet gas, cost of fuel and energy, product specifications, and relative product values.

Increased attention to thermodynamic efficiency and better use of refrigeration and recompression since 1980 have led to development of enhanced NGL recovery processes that are mostly proprietary and available for license.

These processes are equally applicable to the retrofit of existing less efficient plants. In many cases, the economic return on retrofit investments is much more attractive than new plants.

What follows is a review of the major technologies that will likely be in the forefront for new and retrofitted plants in the next century.

Ethane recovery economics

Recovery of NGL components in gas not only may be required for hydrocarbon dew point control in a natural gas stream, but also yields a source of revenue. Lighter NGLs, such as C₂, C₃, and C₄s, can be sold as fuel or feedstock to refineries and chemical plants, while the heavier portion can be used as gasoline-blending stock

The price difference between selling NGL as a liquid and as fuel, commonly referred to as the "shrinkage value," often dictates the recovery level desired by the gas processors.

Advances in gas processing technology in conjunction with improvements in equipment design and efficiency have attracted the processors to invest in gas plants with high liquid recoveries in recent years. ^{1 2}

In addition to their simplicity and lower recompression requirements, new processes are often designed with great flexibility and easy transition between ethane rejection and ethane recovery, allowing the gas processor to respond quickly to the cyclic ethane liquid market, a phenomenon that seems to occur frequently in today's market.

NGL recovery: technical trends

In the 1950s and 1960s, a simple oil-absorption process was commonly utilized to recover propane in the 25 to 50% ranges. This process used light oil to absorb NGL components from the feed-gas stream in an absorption column, typically operated at the ambient temperature and a pressure close to the sales-gas pipeline pressure.

The late 1980s saw introduction of a proprietary selective solvent process based on the refrigerated lean-oil-absorption process in which a lighter oil used as the physical solvent is selectively tailored for recovery of desirable hydrocarbons from a gas stream.³

This process also provides the flexibility of recovering only selective hydrocarbons economically desirable: recovering or rejecting ethane, for example.

Beginning around 1960, as the demand for ethane increased, new processes were sought to recover ethane more efficiently. For ethane recovery up to 50% and LPG recovery up to 90%, a simple propane refrigeration system provides refrigeration at temperatures to -40° F.

For higher recovery, 80% ethane recovery, for example, a cascade refrigeration cycle would be required to provide refrigeration at a lower temperature to approximately -90° F.

The use of turbo-expanders in gas-processing plants began in the early 1960s as a result of technological improvements in the manufacture of practical mechanical designs suitable for continuous operation in a variety of operating conditions.

By 1970, most new gas processing plants were being designed to incorporate the particular advantages of an expander producing useable work normally used partially to recompress residue or export gas. The gas processing industry is now dominated by turbo-expander plants for NGL recovery.

Coincident with development of reliable turbo-expanders was development of compact, efficient, and relatively inexpensive plate-fin or brazed aluminum heat exchangers.

The combination of these two pieces of equipment, together with the availability of accurate thermophysical properties and thus reliable process simulation software, has allowed process designers to achieve significant improvements in the cost and flexibility of turbo-expander plants.⁴

Fig. 1 illustrates the flow for a 1970-vintage turbo-expander process for ethane recovery.

The inlet gas is first prepared for processing by removal of the water and other contaminants. Cooling the inlet gas is accomplished by use of cold residue gas in gas-gas exchangers and cold tower internal liquid in the side exchanger.

Propane refrigeration is often required to help in condensing heavy components for a rich gas. Liquid condensed from the inlet gas is separated and fed to the tower for further fractionation after being flashed to the tower pressure.

The remaining non-condensed vapor portion is subject to turbo-expansion to the top section of the de-methanizer, with the cold liquids acting as the top reflux to enhance recovery of heavier hydrocarbon components.

Cold residue gas from the top of the de-methanizer is first used to cool the inlet gas. Then it is compressed to the sales-gas pressure by the expander-compressor driven by the expander and recompressor the driver of which may be a gas engine, gas turbine, steam turbine, or electric motor.

Hot oil or hot residue gas from the discharge of the recompressor provides heat for the de-methanizer bottom reboiler.

Because the de-methanizer thus operated acts mainly as a stripping column, the expander discharge vapor leaving the column overhead and not subject to rectification still contains a significant amount of heavy components. These components could be further recovered if they were brought to a lower temperature or subjected to a rectification step.

Operating the de-methanizer at a lower pressure or employing two expanders in series to provide a higher expansion ratio and thereby additional refrigeration could achieve the lower temperature option. The increase in recompression horsepower requirements, however, is not cost effective.⁴

During the 1970s, it was recognized that the recompression, generally accounting for 25-50% of a gas plant's cost, could be reduced by increasing the operating pressure of the de-methanizer. The limitation of this concept is illustrated in Fig. 2.

This figure shows that as the pressure is increased toward the critical pressure, the relative volatility (vertical dashed lines) decreases, making fractionation more difficult. Moreover, higher de-methanizer pressure reduces the expansion ratio across the expander, and the temperature profile in the column rises, making the heat integration to the process more difficult.

To compensate for the reduction in relative volatility, ongoing efforts have mostly concentrated on improving the reflux to de-methanizer rectification. For instance, liquid recovery could be improved by addition of a reflux condenser to the de-methainzer.⁵ Refrigeration for the reflux condenser is provided by the expander discharge vapor after being separated from condensed liquid.

Also proposed was a method to preheat the partially condensed liquid then expand it to a second separator at an intermediate pressure to yield a vapor stream preferably consisting of lighter hydrocarbons.⁶ This leaner stream returns to the de-methanizer top as an enhanced reflux after being condensed and sub-cooled.

The most effective approach, perhaps, is the split-vapor process developed independently for a project in Australia and also by Ortloff Corp., Midland, Tex.

This process, often referred as the "gas sub-cooled process" (GSP), uses a small portion of the non-condensed vapor as the top reflux to the de-methanizer, after substantial condensation and sub-cooling. The main portion, typically in a range of 65-70%, is subjected to turbo-expansion as usual.

In spite of less flow being expanded via the expander, this higher and colder reflux flow permits an improved ethane recovery even at a higher column pressure, thereby reducing recompression horsepower requirements. It also provides an advantage in reducing the risk of CO₂ freezing in the de-methanizer.

Propane-recovery processes

Due to equilibrium constraints, the propane recovery achievable in the GSP process during ethane rejection is ultimately limited by propane content in the top reflux to the de-methanizer.

For high propane recovery, a two-column scheme (Fig. 3) is used. In this process, the overhead vapor from the second column (de-ethanizer, for example) is condensed and recycled to the top of the de-methanizer.

This scheme is typically employed in a two-column arrangement with the de-methanizer comprising only the rectification section like an absorber.⁹ The absorber bottom liquids are pumped to the top of the de-ethanizer. The top reflux thus derived is leaner in propane, thereby enhancing recovery efficiency of propane and heavier hydrocarbons.

A two-column scheme, improved mainly in two areas, was also proposed.^{10 11}

The first involves preheating the cold absorber-bottom liquid before it enters the de-ethanizer for a better utilization of its cold refrigeration. The second improvement provides a reflux condenser for the de-ethanizer, minimizing propane and heavier components' escape from the volatile overhead vapor. The absorber top reflux becomes much leaner in propane and consequently enhances the overall recovery efficiency.

In another cryogenic absorption process, the absorber is located upstream of the expander with the expander acting as a partial condenser to provide reflux for the absorber.¹²

Propane recovery in excess of 98% can be achieved cost effectively by use of any process described here.

The introduction of essentially pure ethane stream as the overhead reflux theoretically permits 100% propane recovery.

An improvement to the two-column scheme thermally links both columns via a side reboiler-overhead condenser and introducing an enhanced stripping section to the de-methanizer.¹³ The provision of the stripper section allows undesirable light components to be stripped off the liquids feeding to the de-ethanizer.

As a result, loads and energy requirements for the de-ethanizer can be reduced substantially. In addition, the overhead vapor from the de-ethanizer is essentially pure ethane and free of propane, permitting further enhancement in propane recovery efficiency. Propane recovery in excess of 99% can be achieved efficiently even with a relatively lower expansion ratio.

Another advantage is that the second tower, originally designed and operated as a de-ethanizer, can now be used to produce the ethane product when the plant is switched to ethane recovery operation.

Ethane-recovery processes

As noted earlier, the recovery level attainable in the split-vapor process is ultimately limited by the composition of the vapor stream used for the top reflux to the de-methanizer. The use of a leaner reflux is an attempt to overcome the previous deficiency.

In 1985, Randall Corp., Houston, added a separator onto the GSP reflux stream marginally to improve NGL recovery.¹⁴ Recycling a portion of residue gas as the top reflux to the de-methanizer (Fig. 4) has been introduced in some processes to enhance ethane recovery.

Because the refrigeration available from the residue gas is not cold enough to liquefy the slip recycle stream, a compressor operating at cryogenic temperatures will be required to raise the recycle stream pressure to permit condensation by the main cold residue gas.

The cold residue-gas recycle (CRR) process was introduced to the rectification section to achieve high ethane recovery in a typical cryogenic expansion process.¹⁵ The CRR scheme was incorporated into the original GSP design to improve ethane-recovery efficiency. ¹⁶

In this scheme, the sub-cooled split-vapor stream is introduced to the middle portion of the rectification section, providing the bulk reflux to the expander discharge vapor. The small amount of ethane remaining in the up-flowing vapor subsequently contacts with the ethane-free top reflux and is retained in the down-flowing liquid.

As a result, a smaller recycle flow and thus less compression requirement are required for a given recovery level. Very high ethane recovery, in excess of 98%, is economically achievable with the CRR process; however, the cryogenic compressor can be very expensive.

Alternatively, warm residue gas recycle (WRR) taken from the residue-gas compressor was employed and patented.^{17 18} The compression requirements for the recycle residue gas are accomplished by the main residue-gas compressor, thereby eliminating the need of a dedicated booster compressor and reducing capital investment in most cases.

A recycle compressor may be required, however, when the residue-gas pressure is too low for liquefaction or when an optimal pressure for the recycle residue gas is desired. Very high ethane recovery is also attainable with the WRR scheme.

It is clear that the recovery of hydrocarbon liquids theoretically governed by the phase equilibrium at the column top stage can be enhanced by the combination of a higher, colder, or leaner top reflux.

Recently, IPSI developed and subsequently applied for patent on an alternative approach of obtaining an ethane-free reflux stream to achieve a higher recovery.¹⁹ This lean-reflux process (LRP) eliminates the additional compression required in the residue-gas recycle processes by using a small absorber to condition the split-vapor stream.

Table 1 demonstrates an example comparison of various cryogenic expansion processes to recover up to 95% ethane.

Enhanced NGL-recovery processes

It is interesting to note that all of the above processes focus on improving the reflux stream by making it colder or leaner in ethane or propane. In 1997, IPSI developed and subsequently received a US patent on a process enhancement that focuses on the bottom of the de-methanizer or de-ethanizer.²⁰

This scheme (Fig. 5) utilizes a slipstream from or near the bottom of the tower as a mixed refrigerant stream. The stream is used to cool the inlet gas, thereby reducing or eliminating the need for propane refrigeration.

The vapor portion is recycled back to the bottom of the tower where it serves as a stripping gas, which provides the following enhancements to the tower operation:

• Increases the critical pressure, thereby enhancing the relative volatility at the same pressure; or increases the pressure while maintaining the same relative volatility.
• Reduces or eliminates the need for external reboiler heat, thereby saving fuel and refrigeration.
• Lowers the temperature profile in the tower, thereby enhancing the ability to cool the inlet gas via side reboilers and resulting in reduced heating and refrigeration loads.

Table 2 illustrates an example of the savings that can be obtained with the enhancement vapor process. It is noteworthy that the simplicity of the self-refrigeration scheme can be used to enhance the operating efficiency and reduce capital of any of the aforementioned processes.

Moreover, it can be configured into a simple add-on skid particularly suitable for retrofitting plants to enhance recovery or increase plant capacity without adding more residue gas compression.

References

1. Mallett, M. K., " Largest US gas processing plant begins operations," OGJ, Jan. 19, 1987, p. 33.
2. Lynch, J. T., Pitman, R. N., Slack, J. W., Kruger, K. M., and Hulsey, B. J., "Retrofitting the Williams Energy Services Ignacio Plant for High Throughput and Recovery," presented at the 78th Annual GPA Convention, Mar. 1-3, 1999.
3. Mehra, Y. R., "Mehra Process Flexibility Improves Gas Processing Margins," presented at the 66th Annual GPA Convention, Mar. 16-18 1987.
4. Elliot, D. G., Chen, J. J., Brown, T. S., Sloan, E. D., and Kidnay, A. J., "The Economic Impact of Fluid Properties Research on Expander Plants," Fluid Phase Equilibria, Vol. 116 (1996), pp. 27-38.
5. Gulsby, J. G., US Patent No. 4,251,249 (1981), Randall.
6. Vijayaraghavan, B., and Ostaszewski, R. J., US Patent No. 5,566,554 (1996), KTI Fish.
7. White-Stevens, D. T. and Elliot, D. G., "Massive Cooper Basin Liquids Project in Australia Meets Design and Startup Targets," OGJ, Jan. 20, 1986, p. 58.
8. Campbell, R. E., and Wilkinson, J. D., US Patent No. 4,157,904 (1979), Ortloff.
9. Buck, L. L., US Patent No. 4,617,039 (1986), Pro-Quip.
10. Paradowski, H., Castel, J. H., and Parfait, H. B., US Patent No. 4,690,702 (1987), Technip.
11. Campbell, R. E., Wilkinson, J. D., and Hudson, H. M., US Patent No. 5,771,712 (1998), Elcor.
12. Sorensen, J. N., US Patent No. 5,685,170 (1997), McDermott (Canada).
13. Yao, J., Chen, J. J., Lee, R. J., and Elliot, D. G., "Improved Propane Recovery Methods," US Patent Application No. 09/209,931 (1998).
14. Huebel, R. R., US Patent No. 4,519,824 (1985), Randall.
15. Montgomery, G. J., US Patent No. 4,851,020 (1989), McDermott.
16. Campbell, R. E., Wilkinson, J. D., and Hudson, H. M., US Patent No. 4,889,545 (1989), Ortloff.
17. Aghili, H. K., US Patent No. 4,687,499 (1987), McDermott.
18. Campbell, R. E., Wilkinson, J. D., and Hudson, H. M., US Patent No. 5,568,737 (1996), Ortloff.
19. Lee, R. J., Yao, J., Chen, R. J. J., and Elliot, D. G., "Lean Reflux Process for High Recovery of Ethane and Heavier Components." Patent Pending.
20. Yao, J., Chen, J. J., and Elliot, D. G., "Enhanced NGL Recovery Processes," US Patent Application No. 08/987,183.

The Authors

Douglas G. Elliot is president, chief operating officer, and co-founder of IPSI LLC, a unit of Bechtel Corp., Houston. He holds a BS from Oregon State University and MS and PhD degrees from the University of Houston, all in chemical engineering. He is a Fellow of the American Institute of Chemical Engineers and a Bechtel Fellow, the highest honor Bechtel offers for technical achievement. Elliot has been a member of the Gas Processors Association research steering committee since 1972.

Jame Yao is vice president of IPSI LLC. He holds a BS from National Taiwan University and MS and PhD degrees from Purdue University, all in chemical engineering. He is a member of the AIChE.

R. J. Lee is a senior process engineer with IPSI LLC. He graduated from National Taiwan University in 1981 with a BS in chemical engineering and holds MS and PhD degrees in chemical engineering from Purdue University, Indiana.