TN295 M No. 9174 LIBRARY OF CONGRESS □0001040315 * 5\ ^ U' * ^ u •a5 ^*> * v 1 AsaTj#N «v 'o, . gy v « •» «» < , y|yfi «v ^ . e^ i> v < *» -5 -* % <8" v *i^* '^ a? * ^ °* '"^" 9 a?' '^ '^■o« w ."#ftl^^ "^o^ •^SS: '^o^ ;jflH^ a - '^^ :<^S3;*. '**o« ° " ° A r^»* -^ ^ °." . •_^.> > 4 \-^iX / (i ^-> y\-^k\ <-°.>^ °- ^ ••■ "oV" ^o< .^^ V ^ / ^ ^ ^ •••■•• <^ V ° N ° !v v . 8 i^:* TL f ,° V ,C^!. ^°o ,^' .*^>"^ f .° V >l^L% "°o ^ . & ^:% ^ rP v -v .": V** •»•- %/ #k-, \^ .*». \/ /• v % ^;• , .** A G ^ '" • * * A <. ^ «.' cv * Bureau of Mines Information Circular/1988 Review of Membrane Technology for Methane Recovery From Mining Operations By F. Garcia and J. Cervik UNITED STATES DEPARTMENT OF THE INTERIOR Information Circular 9174 Review of Membrane Technology for Methane Recovery From Mining Operations By F. Garcia and J. Cervik UNITED STATES DEPARTMENT OF THE INTERIOR Donald Paul Hodel, Secretary BUREAU OF MINES David S. Brown, Acting Director MS Library of Congress Cataloging in Publication Data: Garcia, F. (Fred) Membrane technology and methane recovery from mining operations. (Information circular ; 9174) Bibliography: p. 6 Supt. of Docs, no.: I 28.27: 9174. 1. Coalbed methane. 2. Membranes (Technology). I. Cervik, Joseph. II. Title. III. Series: Information circular (United States. Bureau of Mines) ; 9174. -JFN295.U4- 622 s [622 '.334] 87-600370 [TN305] CONTENTS Page Abstract 1 Introduction 2 History of membrane technology 2 Gas transport through membranes 3 Possibilities for applying membrane technology to mining 5 Conclusions 6 References 6 ILLUSTRATIONS 1. Hollow-fiber membrane 3 2. Spiral-wound membrane 3 TABLE 1. Flammability limits of natural gas-air mixtures 5 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT cm centimeter lb/in 2 pound (force) per square inch ft foot lb/in 2 (ga) pound (force) per ft 3 cubic foot square inch, gauge gal gallon m meter hp horsepower m 3 cubic meter in inch pet percent kPa kilopascal St short ton kW kilowatt vol pet volume percent lb pound REVIEW OF MEMBRANE TECHNOLOGY FOR METHANE RECOVERY FROM MINING OPERATIONS By F. Garcia 1 and J. Cervik 2 ABSTRACT Recent advances in the commercial separation of gases using membranes have renewed interest in the possibility of applying this technology to the recovery of methane (CH4) from mining operations. This Bureau of Mines report briefly reviews the history of the development of membranes for gas separation, the theory of how they work, and their application to the separation of methane from air and associated problems. However, methane-air mixtures are difficult to separate with membranes because the pertinent gas couples, 02 - N2, O2-CH4, and N2-CH4, have poor separation characteristics, as indicated by their separation factors of about 3 or less. Even if these separation factors were substantially higher, there is doubt that methane could be recovered economically from the low concentrations in mine ventilation exhaust (2 vol pet or less). The exhaust pressures are not sufficient for adequate separation. The power cost of compressing these mixtures would far exceed the value of the methane recovered. New discoveries could make separation of gob hole methane-air mixtures practical. These mixtures have much higher concentrations of methane (from 30 to 100 vol pet); however, for safety reasons, treatment would be limited to gob gas with 60 vol pet CH4 or more. 'Mining engineer. ^Supervisory geophysicist. Pittsburgh Research Center, Bureau of Mines, Pittsburgh, PA. INTRODUCTION A Bureau of Mines study showed that in 1985 coal mines in the United States emitted 304 million ft (8.6 million m ) of methane daily (1). This reflects an increase of about 48 million ft (1.4 million m ) since 1980 (2). Factors con- tributing to this increase were the open- ing of new mines in deeper and gassier coal beds in the last ten years and these mines are now larger mines. Ventilation is the primary method of controlling methane in coal mines. At an active coal face, the methane must be diluted with air to 1 vol pet CH4 or less for safety reasons. In return airways and bleeders, the maximum allowable methane concentration is 2 vol pet. The methane in gas vented from coal mines cannot be recovered economically with present technology such as distillation, absorption, and adsorption (3_). Cryogenic distillation has always been the system of choice for large-scale sep- aration of gases as well as liquids (4^). However, advances in membrane technology in the past 6 yr have spurred interest in and wide industrial use of membranes for gas separation. HISTORY OF MEMBRANE TECHNOLOGY A simple approach to separating a gas mixture is to construct a barrier that permits molecules of one kind to pass through it while excluding others. Such a barrier in the form of a membrane was first reported in 1831 (5). Thirty-five years later, the mechanism of permeation through a membrane was discussed and dem- onstrated by using a rubber membrane to separate a gas mixture. Polymer membranes were introduced in the late 1940's. These were composite membranes that consisted of a very porous but inert substrate covered by. a polymer layer that separated components. Certain uses of these membranes are well estab- lished, and many improvements in the per- formance of these membranes have been achieved (6^). For example, large-scale water-desalination plants can process more than 600 million gal (2.3 million m ) of water daily; the dairy industry uses membrane technology to process whey proteins; and hemodialyses, a standard treatment for patients with kidney fail- ure, depends upon membrane technology. The possibility of using membranes for industrial gas separation became evident in the 1950 's with the development of new polymeric materials. Many industrial gas separation processes were examined to determine if the use of membranes could — _ -"Underlined numbers in parentheses re- fer to items in the list of references at the end of this report. be applied to them. These processes in- cluded separation of O2 from air, He from natural gas, H2 from coal-hydrogenation tail gas and refinery gas, NH4 from mix- tures containing N2 and H2, and CO2 from various gas mixtures. The use of mem- branes in these industrial processes was limited severely by low permeation rates through the membrane and the poor membrane durability under operational conditions. A major breakthrough in membrane tech- nology occurred in 1960 with the develop- ment of asymmetric membranes (]_)• These membranes are porous throughout , but have a thin, relatively dense skin near one surface, which generally accounts for a very small fraction (0.1 to 1 pet) of the total membrane thickness (4^). These mem- branes have proportionally higher perme- ation rates than the dense membranes of equivalent thickness because the effec- tive separating layer (the dense skin) is so thin. However, they contain pores that are substantially larger than gas molecules, and as a result, they make poor gas separators. In the mid-1970's, a new process de- veloped by Monsanto Co., St. Louis, MO, overcame the problem of surface porosity in asymmetric membranes by applying a high-permeability coating to the porous membrane. The coating plugged surface pores and also served to protect the substrate from damage due to abrasion and normal handling (]_)» Because of the Nonpermeative gas outlet Fiber bundle plug 4- or 8-in diam by 10 ft Feed stream of mixed gases Hollow-fiber membrane ASME code carbon steel shell Not to scale Permeative gas outlet FIGURE 1 .—Hollow-fiber membrane. coating, the gas-separating layer can be made thin without concern regarding pore problems and consequently, gas flow rates through the substrate are 1,000 to 10,000 times faster than through other types of membranes. The Monsanto membrane 4 can be manu- factured in flat sheets or as a hollow- fiber membrane (8). The two membrane types have different configurations. Hollow-fiber membranes are slender, spun Feed Permeative at low pressure Residual gas KEY 1 Feed channel 2 Membrane Feed channel Membrane -Permeative channel FIGURE 2.— Spiral-wound membrane. filaments several hundred microns in diameter. Typically, they are packaged in parallel in a 4- or 8-in-diam (10.2- or 20.3-cm) steel tube (module) into which the gas mixture is forced. The gas to be separated permeates from the out- side to the inside of the hollow fibers (or vice versa) and is collected at one end of the tube (fig. 1). To obtain a spiral-wound configuration, several flat sheets are separated by spacers to create a turbulent flow path for the feed gas, and then rolled up on a central tube and inserted into a steel shell (module) (fig. 2). Both hollow-fiber and spiral- wound modules are arranged into various interconnected banks to constitute a sep- aration system. Monsanto-type members are used commer- cially for the separation of H2 and He from gases such as N2, CO, and CH4 (4_). Potential applications for these mem- branes include H2 recovery from purge gases in ammonia synthesis and the recov- ery and recycling of CO 2 in enhanced oil recovery processes. GAS TRANSPORT THROUGH MEMBRANES Gas transport through a membrane is controlled by Fick's law of diffusion and Henry's law relating solubility of a gas in the polymeric membrane (4, 9). By Fick's law, diffusion through the mem- brane is ^Reference tc specific products does not imply endorsement by the Bureau of Mines. Q = DaAACa ) (1) where Q = flow of component a through membrane , D a = diffusion coefticient for component a, AC a = concentration difference across the membrane, A = membrane surface area, and L = membrane thickness. Henry's law relates the concentration of gas a to the partial pressure of gas a in contact with the polymer: Ca - S a P a , (2) where S a = solubility constant and P a = partial pressure in contact with membrane. Substituting equation 2 into equation 1 yields Q= KaAAPa, (3) Li where K a = S a D a = permeability coefficient. An equation of this form can be written for each component in the gas stream. Equation 3 shows that where the partial pressure differentials for two gases in a mixture are the same, the ratio of the flow rates of each gas through the mem- brane may be expressed as Qa Qb Ka Kb (4) where a = selectivity or separation factor. The separation factor indicates the abil- ity of the polymer to separate two gases in a given mixture. Equation 3 shows that high gas flows through a membrane can be obtained by in- creasing the permeation coefficient (K), or increasing the surface area of the membrane (A), or increasing gas pressure (P). The problems associated with the control of the physical and separation properties (K) of polymers can be almost as great as those associated with mak- ing working membranes (^0. Consequently, commercial membrane separation systems are based on available polymers developed for other applications. Increasing the surface area of a membrane increases the size and cost of the system and at some point, which is specific for each appli- cation, use of additional surface area makes the system uneconomical. Gas flow through a membrane requires a driving force; this force is represented in equation 3 by the partial pressure dif- ferential across the membrane (AP a ). Gas flow through the membrane can be in- creased by increasing gas pressure. How- ever, compression consumes energy; and the increased cost of pressure vessels, compressors , and energy associated with compression makes this approach unattrac- tive. In addition, higher operating pressures require a membrane of much greater strength. Thus, practical and economic problems limit the degree to which the permeation coefficient, mem- brane surface area, and gas pressure can be changed. Equation 3 shows that gas flow rates through a membrane are inversely propor- tional to the thickness (L) of the mem- brane. Thus, the development of thinner polymer membranes made possible the pro- duction of high-gas-flow systems that satisfy the demands of the commercial gas separating industry. For example, the overall thickness of the Monsanto mem- brane ranges from 1 by 10~ 3 to 10 by 10~ 3 in (2.5 by 10~ 3 to 25 by 10" 3 cm). Its dense skin, which actually accomplishes the gas separation, is 0.004 by 10" 3 to 0.04 by 10" 3 in (0.01 by 10" 3 to 0.1 by 10" 3 cm) thick (4_). The selectivity or separation factor (a in equation 4) should be at least 20 and often must be more than 40 for the gases to be separated (4^). The separation fac- tors for various gas couples follow: H 2 (vapor)-CH 4 200-400 H 2 -CH 4 40-55 CO 2 -CH 4 20-30 H 2 S-C 3 H 8 75-110 He-CH 4 60-85 O2-N2 4-5 The O2-N2 separation factor is of an order of magnitude less than the separa- tion factors of the other gas couples because of the small differences between O2 and N2 molecular size and solubility (10) . This low separation factor makes O2 enrichment difficult to accomplish. POSSIBILITIES FOR APPLYING MEMBRANE TECHNOLOGY TO MINING Methane is exhausted from coal mines through the ventilation system and through surface gob holes. In some mines, long horizontal holes are drilled into the coalbed to drain methane. These holes are connected to an underground pipeline that transports the gas to the surface. Because the methane content of the drained gas is 90 vol pet or greater, it usually requires no remedial treatment and can be compressed and pumped into a commercial gas transmission pipeline. However, the gas exhausted through the ventilation system and produced from gob requires remedial treatment before it can be pumped into commercial pipelines. The main gas components of a CH4~air mixture are O2, N2, and CH4. The fol- lowing tabulation gives the permeation rates of these gases at 750 lb/in 2 (5,170 kPa), in standard cubic feet per hour per square foot times 100 lb/in : CH 4 0.18 N 2 . 0.16 2 0.59 The permeation rates for N2 and CH 4 are practically the same, while the rate for O2 is almost three times greater than either CH4 or N2 (11). Consequently, the use of membranes to upgrade coal mine exhaust ventilation systems does not ap- pear to be feasible with existing mem- brane systems. Even if a membrane system existed that would separate the methane from the exhaust ventilation gas with 100 vol pet efficiency, only 2 ft 3 (0.06 m 3 ) of methane at most would be obtained for every 100 ft 3 (2.8 m 3 ) of exhaust gas treated. Because membrane systems oper- ate at pressures of 2,000 lb/in 2 (13,790 kPa) or more (4), the cost of energy re- quired to sufficiently compress the ex- haust gas far exceeds the value of the CH 4 recovered. For example, if the ex- haust gas is compressed to 2,000 lb/in , and 1 million ft 3 (0.028 million m 3 ) of gas were treated daily, a 700-hp (522 kW) compressor would be required. Electrical power to operate the compressor would cost about $525 daily, while the value of the recovered methane would be only about $130 based on a sale price of $6.50 per 1,000 ft 3 ($6.50 per 28.8 m 3 ), a price that is higher than you can now get. Gob gas is generally mixed with mine air, and its composition varies from nearly 100 vol pet CH 4 when production first starts to 30 vol pet over a period of months. Because the separation fac- tors for the gas couples in the mixture are less than about 3, membrane tech- nology does not appear applicable to upgrade gob gas to a saleable product. There are safety factors to consider when compressing gob gas. Methane-air mix- tures are explosive in the range from 5 to 15 vol pet CH4 at atmospheric condi- tions. Methane is almost always the major constituent of natural gas. Conse- quently, methane and natural gas have similar limits of flammability (table 1). At pressures of 2,000 lb/in the lower limit of flammability decreases to 3.60 vol pet and the upper limit increases to 59.0 vol pet. Thus, if appropriate TABLE 1. - Flammability limits of natural gas-air mixtures Pressure, lb/in 2 (ga) (atmospheric) 500 1,000 2,000 3,000 e Estimated. Source: Jones (12, p. 7). Limits, vol pet natural gas 14.20 44.2 52.9 59.0 e 60.0 60142 138 membrane technology were available, only gob gas that contained more than 60 vol pet CH4 could be treated. The re- mainder would have to be vented. The gases N 2 , 2 , and CH 4 rend to have low permeabilities in most polymers and are therefore difficult to separate. New discoveries will be needed to make such separations practical with membranes. Many new developments in both membranes and the process design of new applica- tions are expected in the next several years (4). CONCLUSIONS The present state of membrane technol- ogy precludes the use of membranes for separation of methane from exhaust venti- lation or gob gas. Because methane concentration in ex- haust ventilation is 2 vol pet or less, it is doubtful that this methane could be recovered economically even if membrane technology were available. The power cost of gas compression far exceeds the value of the methane that could be recov- ered. For gob gas, treatment would be limited to methane concentrations of 60 vol pet or more because of the explo- sive nature of compressed methane-air mixtures. Membrane technology for gas separation is still developing. New membranes yet to be developed could make upgrading of gob gas feasible. REFERENCES 1. Grau, R. H. An Overview of Methane Liberations From U.S. Coal Mines in the Last 15 Years. Paper in Proceedings of 3rd U.S. Mine Ventilation Symposium. Soc. Min. Eng. AIME, 1987, pp. 251-255. 2. Irani, M. C, J. H. Jansky, P. W. Jeran, and G. L. Hassett. Methane Emis- sion From U.S. Coal Mines in 1975, A Sur- vey. BuMines IC 8733, 1977, 55 pp. 3. Skow, M. L. , A. G. Kim, and M. Deul. Creating a Safer Environment in U.S. Coal Mines. The Bureau of Mines Methane Control Program, 1964-79. A Bu- reau of Mines Impact Report. BuMines Spec. Publ. , 1981, 50 pp. 4. Henis, J. M. S., and M. K. Tripodi. The Developing Technology of Gas Separat- ing Membrane. Science, v. 220, No. 4592, 1983, pp. 11-17. 5. Lacey, R. , and S. Loeb (ed.). Industrial Processing With Membranes. Wiley, 1978, pp. 279-339. 6. Fox, J. L. Membrane Development Slowed by Weak Economy. Chem. & Eng. News, v. 62, No. 52, Nov. 8, 1982, pp. 7- 12. 7. Rosenzweig, M. D. Unique Membrane System Spurs Gas Separation. Chem. Eng. (N.Y.), v. 88, No. 24, Nov. 1981, pp. 62- 66. 8. Parkinson, G. , S. Ushio, and R. Lewald. Membranes Widen Roles in Gas Separations. Chem. Eng. (N.Y.), v. 91, No. 8, Apr. 16, 1984, pp. 14-19. 9. Maclean, D. L. , D. J. Stookey, and T. R. Metzger. Fundamentals of Gas Per- meation. Hydrocarbon Process., v. 62, No. 8, 1983, pp. 47-51. 10. Schell, W. J., and C. D. Houston. Membrane Gas Separations for Chemical Processes and Energy Applications. ACS, 1983, pp. 125-143. 11. Schell, W. J. Membrane Use/Tech- nology Growing. Hydrocarbon Process., v. 62, No. 8, 1983, 43 pp. 12. Jones, G. W. , R. E. Kennedy, and I. Spolan. Effect of High Pressures on the Flammability of Natural Gas-Air- Nitrogen Mixtures. BuMines RI 4557, 1949, 16 pp. U.S. GOVERNMENT PRINTING OFFICE: 1988 - 547-000/80.022 INT.-BU.0F MINES, PGH., PA. 28655 J.S. Department of the Interior Sureau of Mines— Prod, and Distr. Cochrane Mill Road P.O. Box 18070 Pittsburgh. Pa. 15236 AN EQUAL OPPORTUNITY EMPLOYER OFFICIAL BUSINESS PENALTY FOR PRIVATE USE, S300 [ ~"~J Do not wish to receive this material, please remove from your mailing list* l~~J Address change. Please correct as indicated* 4 O SUV ■ /\ w" / v ^ •! /\ : "\ ^ • ** ' *♦ •• c .v.' / "V*. o>* \'$ / ,5^ ,^V ~A" \(fi^^*. '"^rtf o«^M0l'- ^|*