Anthony L. Lee, Harlan L. Feldkirchner
Institute of Gas Technology
Chicago
Jorge P. Gamez, Howard S. Meyer
Gas Research Institute
Chicago
A pilot field study Of C02 removal from natural gas by a polymeric membrane system has found that a single pass membrane process can reduce the C02 concentration from 5 mole % in the feed to less than 2 mole % in the sales gas and H20 from 67 lb/MMscf to less than 7 lb/MMscf.
The methane loss through the membrane permeate (acid gas) was approximately 10%, and the total operating cost was $0.14/Mscf.
The study was funded by the Gas Research Institute, Chicago, and took place at a 500 Mscfd plant in Trinity County, Tex. It was implemented cooperatively with Dallas Production Inc., Grace Membrane Systems, and the Institute of Gas Technology, Chicago.
OUT OF SPEC GAS
In the past, conventional natural gas processing has included primarily absorption processes that use either chemical or physical solvents (for example, amines or glycol based systems) to remove H2S, C02, and H,O.
A recent study performed for GRI on the composition of subquality components of natural gas in the U.S. Lower 48 states indicates that approximately, 15% of nonassociated gas production contains C02 in concentrations exceeding the typical pipeline specification of mole %.
Although polymeric membrane systems are available for C02 removal, dehydration, and separation of small amounts of H2S, available field operating data on these systems have been limited. And there are no publicly available multihydrocarbon data on membrane based C02 separation systems' performances.
Proper field assessment of the capabilities of commercially, available membranes will permit a definition of the role that CO, membrane separation systems can have.
Field testing for this project occurred at a small, commercial size, spiral wound membrane system on a gas well.
Because C02 and H20 are more permeable than methane, ethane, and higher hydrocarbons, an efficient separation can be achieved. Reductions Of C02 contents from the 25 mole % level to a 2 3 mole % pipeline specification are achievable.
A recent economic study showed that hybrid membrane/amine systems may be more economic in some feed gas flow rate and C02-concentration ranges.
TEXAS UNIT
With no rotating parts or solvents required, membrane systems are simple to operate. Wide turndown ratios are also achievable by varying the number of membrane elements within the limitations of the element housings.
This provides the versatility to process natural gas with varying compositions at a wide range of flow rates.
A Grace Membrane cellulose acetate membrane unit, designed to treat 500 Mscfd of natural gas at 750 psig containing 6 mole % C02 and 67 lb/MMscf H20, was installed at a Dallas Production site in Trinity County, Tex.
A schematic of the installed unit is shown in Fig. 1; a photograph of the unit's gas metering system is shown in Fig. 2.
During the 573 day operating period, the unit processed 94.7 MMscf of natural gas to a sales gas that met pipeline specifications. There was no apparent deterioration in the membrane material or in its performance, and the data obtained were consistent with what was expected.
To yield information on membrane performance with gases containing higher concentrations of C02 (up to 25 mole %), a liquid C02 source was temporarily installed and additional tests were conducted.
These tests included the following data ranges: feed C02 concentrations from 3.5 to 24.9 mole %, pressures from 350 to 700 psig, temperatures from 102 to 1220 F., and feed gas flow rates from 30 to 240 Mscfd. The produced sales gas met specifications continuously.
Table 1 gives typical feed, sales, and vent gas compositions for three different levels of feed gas C02 content for a standard cellulose acetate membrane module with five elements present.
These are typical of the data obtained and show that, even with a feed gas C02 content of more than 20 mole %, it was possible to meet the local pipeline specification of under 2 mole
STAGE CUTS
Fig. 3 gives typical plots of stage cut (ratio of the species in the vent flow rate to the species in the feed flow rate) for ethane, methane, and C02 as a function of membrane area (given as number of membrane elements present). These results are typical of those obtained during the test period.
The following conclusions can be drawn for Constant-membrane area conditions:
C02-stage cuts:
- Increase nearly linearly with increasing total pressure
- Decrease rapidly with increasing feed gas flow rate
- Generally increase with increasing feed gas C02 content
- Increase more rapidly with pressure and are typically several times higher than methane stage cuts
- vary less with membrane area than methane or ethane stage cuts.
Methane stage cuts:
- Increase linearly with increasing total pressure
- Decrease rapidly with increasing feed gas flow rate
- Generally increase with increasing feed gas C02 content
- Increase more rapidly with pressure and are typically about 50% higher than ethane stage cuts.
Ethane stage cuts:
- Increase linearly with increasing total pressure, but the rate of increase is less than that for C02 and methane
- Decrease with increasing feed gas flow rate
- Generally increase with increasing feed gas C02 content.
PROCESS ECONOMICS
Processing costs were compared for amine based plants and single stage membrane systems. Following are the processing costs included in the evaluation on a per year basis:
- Operating expenses: Utilities, labor, maintenance, plant overhead, and amine/membrane replacement. (The cost of dehydration is not included for the amine systems.)
- Lost product gas: Hydrocarbon loss in the acid gas and in the fuel used.
- Capital charge: Taken as 10 year depreciation.
These criteria were used to provide consistency with capital and operating costs that were found in the literature 1 4 and used for economic comparisons. Table 2 provides the ranges of estimated costs for single stage membrane and for DEA (diethanol amine) treatment.
Processing costs for membrane systems decrease with increasing feed C02 concentrations, whereas those for amine systems do not. This is mainly due to their differences in operating expenses.
Table 3 shows plant capital costs for amine and membrane systems.
COST COMPARISONS
This project is providing an expanded data base that will facilitate economic evaluation of membrane based natural gas processing systems and will allow comparison of membrane system capital and operating costs with conventional amine-treatment costs.
The results of this work will also help identify specific economic niches where membrane technology can be used to process natural gas. The successful application of membrane systems will help move natural gas to the consumer at lower prices.
For this application evaluation, as stated, it was found that a single pass membrane process was sufficient to reduce the C02 concentration from 5 mole % in the feed to less than 2 mole % in the sales gas and H20 from 67 lb/MMscf to less than 7 lb/MMscf.
The methane loss through the membrane permeate (acid gas) was approximately 10%, and the total operating cost was $0.14/Mscf.
It should be noted that, among membrane systems, a single stage membrane system incurs the highest processing costs because of higher hydrocarbon losses, and the system that was evaluated did not minimize hydrocarbon loss.
The shrinkage associated with the membrane unit can be used as a fuel source.
ACKNOWLEDGMENT
The authors acknowledge the support of GRI and the participation of GRI's K. Woodcock, the cooperation and assistance of Dallas Production's D. Smith, D. Hamilton, and D. Sheetz, and the technical assistance of Grace Membrane's L. Mologne, D. Mirdadian, and W. Detloff.
REFERENCES
- Babcock, R. E., Spillman, R. W., Goddin, C. S. and Cooley, T. E., "Natural Gas Cleanup: A Comparison of Membrane and Amine Treatment Processes." Presented to the AIChE Spring National Meeting, New Orleans, March 1988.
- Cooley, T. E., and Detloff, W. L., "Field Tests Show Membrane Processing Attractive," Chemical Engineering Progress, Vol. 81, No. 10 (1985), pp. 45 50.
- Spillman, R. W., Barrett, M. G., and Cooley, T. E., "Gas Membrane Process Optimization," Presented to the AICHE Spring National Meeting, New Orleans, March 1988.
- Spillman, R. W., "Economics of Gas Separation Membranes," Chemical Engineering Progress, Vol. 85, No. 1 (1989), pp. 41 62.
