Study compares C2-recovery for conventional turboexpander, GSP

Dec. 8, 2008
Among the various ethane-recovery processes, the turboexpander process is the most widely used.

Among the various ethane-recovery processes, the turboexpander process is the most widely used. Proponents of the gas subcooled process (GSP), however, claim it achieves higher recovery than the conventional turboexpander process.

Several new processes based on modifications to the GSP have appeared: cold residual recycle (CRR), IPSI LLC’s Enhanced NGL Recovery, Fluor’s Recycle Split Vapor (RSV), and RSV with enrichment (RSVE).

This article presents results of a comparison of the two processes for ethane recovery: the GSP and the conventional turboexpander process. To perform a fair comparison, we selected four different feeds, ranging from lean to rich gas, with C2+ contents ranging 6-30%.

Also, comparing the two processes using only one choice for the design (also called “decision”) variables (i.e., those independent variables over which the engineer has some control for each process) does not yield a fair comparison.1 We therefore determined the design variables for both processes and changed them in order to achieve the maximum ethane recovery for each process.

We employed the HYSYS optimizer tool for each level of the demethanizer pressure. The results were confirmed by sensitivity analysis. In it, the design variables were changed manually in each simulation run while covering the entire range of allowable values in the design variables to ensure that for each process, the optimum reached is a global one (“the best for all allowable values of the decision variables”) and not only local (“a point from which no small, allowable change in decision variables in any direction will improve the objective function”).1 The pressure range in the demethanizer covers the full typical range: 100 to 450 psia.

The results obtained show higher ethane recovery with the GSP under two conditions: lean feed and low demethanizer pressure. In all other cases, conventional turboexpansion yields either higher or equivalent ethane recovery.

Choosing ethane recovery processes, therefore, must be carefully considered. Feed compositions and demethanizer pressure can favor one process or the other.

Ethane extraction

Several NGL extraction methods have been proposed. Ethane, a valuable petrochemical feedstock, is the most volatile NGL component and the most difficult to recover. Recovery of a large fraction of C2 guarantees recovery of significantly larger fractions of C3+. References 2-6 review and discuss the different options and provide the advantages and disadvantages.

The Joule Thompson valve expansion requires high feed-gas pressure and can be the method of choice for small feed-gas rates and moderate ethane recovery.5 If the gas pressure is not high enough, refrigeration7 can be added to enhance recovery.5

Obtaining high ethane recovery requires a turboexpander process. Typically, a turboexpander is used in combination with JT expansion and propane refrigeration. Cascade refrigeration is complex and requires high compression cost.2 The use of mixed refrigerant is another alternative2 8 and is commonly used in LNG processes but much less in NGL recovery.2

In contrast to the previous processes in which recovery is obtained by cooling through expansion or mechanical refrigeration, oil absorption and refrigerated lean-oil absorption have been used to recover NGL. Although the pressure drop of the gas stream is minor, the process is expensive in terms of equipment and energy requirements and hard to operate.3 5 Mehra and Gaskin discuss guidelines for choosing cryogenics or absorption.9

McKee summarizes the evolution in design for the “old” generation of turboexpander process.10 The simple plant consists of turboexpansion. Reboiling with inlet gas is used to recover some level of refrigeration. In the cold liquid separator process, the liquid from the higher pressure separator is expanded and the cold liquid is used for feed-gas refrigeration. The other options include the use of side reboiler, refrigeration, and two stages of expansion.10

Chebbi et al. compared five different turboexpander ethane-recovery processes, using the HYSYS optimizer tool to determine the maximum possible ethane recovery.11 The study showed that more complex processing scheme may yield the same or less ethane recovery.

The next generation of ethane recovery processes uses the idea of a reflux to condense more ethane in the demethanizer. The residue recycle (RR) process cools a fraction of the recompressed residue gas, then flashes it by JT expansion, and then sends it to the demethanizer as a reflux.5

Click here to enlarge image

The gas subcooled process (GSP) splits the gas stream leaving the cold separator into two streams, one of them feeding the turboexpander and the second one subcooled by the demethanizer overhead stream, flashed in a valve, and then sent to the demethanizer as a reflux.5 12

The CRR process has one addition to the GSP: A fraction of the demethanizer overhead stream is compressed, partially condensed, expanded through a valve, and then sent back to the demethanizer as additional reflux.5 12 The CRR process is reported to achieve high ethane recovery.12 The cryogenic compressor cost, however, may be prohibitive.4

The IPSI process is another modification of the GSP.4 5 A side stream flows from the demethanizer column, is expanded through a valve, used as a mixed refrigerant to cool a fraction of the feed gas, then returns to the demethanizer (and used for stripping) after separation and compression.4 5 Combining the GSP with the liquid subcooled process (LSP) provides higher recoveries than the GSP and LSP used individually, according to simulation for eight different feeds.13

The RSV process is another modification of the GSP.12 Sending a small fraction of the recompressed residue gas to the demethanizer after condensation and subcooling and then to JT expansion yields an additional reflux in the demethanizer. A modification of the RSV process is the RSVE (recycle split-vapor with enrichment) process.12 Mixing with a portion of the gas leaving the cold separator occurs before condensation and subcooling, therefore requiring a lower capital cost.

Maximizing recovery

We considered four different feeds: A to D. The table (previous page) shows their compositions.14

The feeds range from lean gas to rich gas with C2+ content ranging 6-30%. The four demethanizer pressures considered are 100, 215, 335, and 450 psia. In all cases, the feed gas is at 100° F. and 882 psia (60 atm). The residue gas is recompressed to 882 psia and, in the NGL stream, the molar ratio of C1 to C2 is set at 0.02.2 In all cases the feed-gas rate is 10,980 lb-mole/hr.

Click here to enlarge image

In the conventional process (Fig. 1), the feed is first cooled while providing the reboiler duty, then cooled further by heat exchange with the residue gas, followed by heat exchange with propane in a chiller to reach a temperature of –31° F.

After separation from the liquid, the gas from the separator is cooled by heat exchange with the overhead stream from the demethanizer. It is then separated from the liquid in the cold separator, expanded in a turboexpander, and sent to the demethanizer.

The liquids from the two separators undergo JT expansion and flow to the demethanizer at lower levels. The turboexpander provides part of the power needed to recompress the residue gas. The recompressor provides the other portion.

The temperatures after cooling by heat exchange with the residue gas from the demethanizer are design variables that are changed in the optimization process to maximize ethane recovery for different demethanizer pressures. The constraints are conditions preventing temperature cross (a situation in which “the cold-fluid outlet temperature” would be “higher than that of the warm fluid” in these two heat exchangers).2

Click here to enlarge image

In the gas subcooled process (Fig. 2), the gas undergoes cooling as in the previous case, and after chilling and separation, the entire gas stream does not flow to a turboexpander. Rather, it is split, with a portion condensed and subcooled by the overhead stream from the demethanizer then expanded through a valve to provide a cold reflux after feeding the column.

The design variables include those mentioned for the conventional turboexpander process as well as the flow ratio in the splitter splitting the gas stream leaving the separator and sending a portion to the subcooler. The constraints are the same as in the previous process.

Feed; pressure

For HYSYS simulation, we selected the Peng-Robinson equation of state as the thermodynamic model. After ethane-recovery maximization for each case, we compared results.

Click here to enlarge image

Figs. 3a to 3d show the maximum percent C2 recoveries for the four feeds (A to D) at four different demethanizer pressures: 100, 215, 335, and 450 psia. It is clear that the GSP provides higher C2 recoveries in the cases 100 psia for feeds A and B and at 215 psia for Case A. Ethane recoveries are about the same for the GSP and the conventional turboexpander process for Feed B at 215 psia.

Click here to enlarge image

The effect of pressure on the maximum ethane recovery appears in Figs. 4a to 4d for feeds A to D, respectively. As expected, decreasing the demethanizer pressure enhances C2 recovery for Feeds A to D.

The trend for the variation of maximum C2 recovery with the demethanizer pressure depends on the feed selected. For Feeds A and B, at low pressure, the GSP allows a higher maximum recovery than that of the conventional turboexpander process, whereas the maximum ethane recovery with the GSP is less at higher pressure.

For feeds C and D, the maximum ethane recovery with the GSP is less than for the conventional turboexpander process at all pressures.

Optimization results show that the GSP provides higher maximum ethane recoveries, provided that two conditions are satisfied: The demethanizer pressure is low and the feed gas is lean. In all the other cases, the conventional turboexpander process provides a higher or an equivalent maximum ethane recovery.

Simulation results show that the temperature of the feed gas after the first heat exchange with the overhead stream from the demethanizer does not affect ethane recovery, which is expected since the next cooling is by refrigeration that brings the temperature to –31° F. no matter the temperature after cooling with the overhead gas stream.

The second cooling with the overhead stream has an effect on ethane recovery. A lower temperature enhances condensation of ethane and ethane recovery, which is expected. The maximum ethane recovery is obtained with the lowest possible temperature without temperature cross in the heat exchanger in question. The HYSYS optimizer tool confirms these results.

Acknowledgment

This work was supported by the American University of Sharjah under grant No. FRG07-040.

References

  1. Turton, R., Bailie, R.C., Whiting, W.B., and Shaeiwitz, J.A., Analysis, Synthesis and Design of Chemical Processes, New Jersey: Prentice Hall, 2003.
  2. Manning, F.S., and Thompson, R.E., Oilfield Processing of Petroleum, Vol. 1; Tulsa: PennWell Publishing Co., 1991.
  3. Arnold, K., and Stewart, M., Surface Production operations, Vol. 2; Houston: Gulf Publishing Co., 1999.
  4. Lee, R.J., Yao, J., and Elliot, D.G., “Flexibility, efficiency to characterize gas-processing technologies,” Oil & Gas Journal, Dec. 13, 1999, p. 90.
  5. GPSA Engineering Data Book, Sec. 16, 12th Ed., Tulsa: Gas Processors Suppliers Association, 2004.
  6. Kidnay, A.J., and Parrish, W.R., Fundamentals of Natural Gas Processing, Boca Raton, Fla.: Taylor and Francis, 2006.
  7. Russel, T.H., “Straight refrigeration still offers processing flexibility,” Oil & Gas Journal, Jan. 24, 1977, p. 66.
  8. MacKenzie, D.H., and Donnelly, S.T., “Mixed refrigerants proven efficient in natural-gas-liquids recovery process,” Oil & Gas Journal, Mar. 4, 1985, p. 116.
  9. Mehra, Y.R., and Gaskin, T.K., “Guidelines offered for choosing cryogenics or absorption for gas processing,” Oil & Gas Journal, Mar 1, 1999, p. 62.
  10. McKee, R.L., “Evolution in Design,” Proceedings of the 56th Annual GPA Convention, Dallas, Mar. 21-23, 1977, pp. 123-25.
  11. Chebbi, R., Al-Qaydi, A.S., Al-Amery, A.O., Al-Zaabi, N.S., and Al-Mansouri, H.A., “Simulation study compares ethane recovery in turboexpander processes,” Oil & Gas Journal, Jan. 26, 2004, p. 64.
  12. Pitman, R.N., Hudson, H.M., Wilkinson, J.D., and Cuellar, K.T., “Next Generation Processes for NGL/LPG Recovery,” presented to the 77th Annual GPA Convention, Dallas, Mar. 16, 1998.
  13. Jibril, K.L., Al-Humaizi, A.I., Idriss, A.A., and Ibrahim, A.A., “Simulation study determines optimum turboexpander process for NGL recovery,” Oil & Gas Journal, Mar. 6, 2006, p. 58.
  14. Bandoni, J.A., Eliceche, A.M., Mabe, G.D.B., and Brignole, E.A., Comp. Chem. Eng., 1989, Vol. 13, p. 587.

The authors

Click here to enlarge image

Rachid Chebbi ([email protected]) is professor of chemical engineering at the American University of Sharjah, Sharjah, UAE, which he joined in 2006. Before that, he was a member of the faculties of the United Arab Emirates University and the University of Qatar. He also worked for the Tunisian Ministry of Economy, Shell Tunisie, and Entreprise Tunisienne d’Activités Pétrolières. Chebbi received his Diplôme d’ingénieur from the Ecole Centrale de Paris and holds MS (1984) and PhD (1991) degrees in chemical engineering from the Colorado School of Mines, Golden.

Click here to enlarge image

Khamis Al Mazroui ([email protected]) is a process engineer with BP Sharjah Ltd. He holds a BS (2008) in chemical engineering from the American University of Sharjah.

Click here to enlarge image

Nabil Abdel Jabbar (nabdel [email protected]) is associate professor of chemical engineering at the American University of Sharjah and is on leave from Jordan University of Science and Technology. For 1998-99, he worked as a senior project engineer at AspenTech, Houston, in advanced control and optimization. Abdel-Jabbar holds a PhD (1996) in chemical engineering from the University of Michigan, Ann Arbor, and MS (1990) and BS (1986) degrees in chemical engineering from Kuwait University.