OVERSIMPLIFICATIONS CAN LEAD TO FAULTY COALBED GAS RESERVOIR ANALYSIS

Jeffrey R. Levine
Consultant
Tuscaloosa, Ala.

Proper characterization of coalbed gas reservoirs needs to include analyses that account for the complexity and variability of coals.

Commonly used compositional characterization methods such as proximate analysis which are geared more or less specifically to certain technologies (e.g. coke making) are inadequate for complete coalbed reservoir characterization.

Geologic studies of coalbed reservoirs should take a "systems" approach in considering the complex interplay among all the constituent species comprising the coal. Because of the complexity and heterogeneity of coalbed gas reservoir systems, the concentration of occluded gases cannot be accurately predicted, and should be directly measured for reservoir evaluation.

The term "coal" is loosely applied to a diverse class of organic matter rich sedimentary materials. Compositionally, coal is so complex and heterogeneous that it defies simple analysis and description. Nevertheless, some relationships involving the influence of coal composition on gas production have been presented in an overly simplified fashion.

Table 1 modifies five oversimplifications that are used to describe coalbed gas reservoirs.

These oversimplified relationships,1 even if generally true, have tended to draw attention away from the variability among different types of coals. The impression created is that these systems are more uniform and predictable than they really are.

Oversimplifications have generally evolved out of studies that were either "biased" in their sampling or based upon certain simplifying assumptions. In either case, the conclusions from a few particularly influential early studies have been widely cited and widely used in coalbed reservoir evaluations.

While it can sometimes be counterproductive to focus attention on exceptions rather than rules, it is always beneficial in coal analysis to remain aware of the limits of our knowledge, and perhaps more importantly, the assumptions and simplifications that were originally used in deriving certain relationships.

SOLID SUBSTANCE

The first oversimplification is that coal is a solid substance.

Superficially, coal appears to be a solid material. Upon closer scrutiny, however, its composition and structure are much more complex. Coal is more properly viewed as a heterogeneous, polyphase system having some physical and chemical properties of both a solid and a fluid.

According to widely accepted definitions,2 3 coal is distinguished from other sedimentary materials solely by being comprised principally of organic matter.

Beyond this vague criterion, coal composition can be highly variable, both within a single coalbed and between different coalbeds. Compositional differences reflect the combined influences of:

• Variety of source materials contributing toward the coal

• Degree of initial alteration during deposition

• Degree of coalification during burial.

Compositional heterogeneity within any given coalbed ranges from the macroscopic (millimeter) scale (i.e., lithotype banding, visible to the eye) down to molecular (manometer) scale.

On a micrometer scale, coal is comprised of different macerals, visible through the optical microscope. Some coal macerals, e.g., bituminite, exudatinite, and to a certain degree, vitrinite, cannot be regarded as solids because they show clear evidence of fluid behavior during at least part of their history.

On a molecular scale, current models of coal composition regard coal as a heterogeneous mixture of constituent "phases," at least some of which would be gases or liquids at room temperature and pressure were it not for their intimate association within the coal structure.4 7

The molecular constituents of coal may be grouped into two conceptual classes:

• A matrix fraction, consisting of a three dimensionally crosslinked, macromolecular organic network

• An occluded molecular fraction of lower molecular weight constituents that are either loosely bonded to or physically entrapped within the matrix (Fig. 1).

MOISTURE

Included in the molecular fraction are inherent moisture, and a diverse suite of "hydrocarbons," beginning with methane at the low end and continuing through high molecular weight tars and asphaltenes at the high end.4 6

Although moisture is sometimes not considered to be a constituent of coal, inherent moisture constitutes an integral component of coal at all ranks. Moisture shares the same sort of inter relationship with coal structure as other so called "organic" molecular constituents.

When moisture is removed, the coal structure collapses and shrinks, and all measurable physical and chemical properties change.

In high rank coals, changes associated with dehydration are reversible by resaturating the coal with moisture,10 11 but at low rank they are partly irreversible.12

MOLECULAR FRACTION

The composition and concentration of the molecular fraction of coal is not static, but changes markedly and progressively during coalification (Fig. 2). The principal change occurring at low rank (peat through subbituminous A) is a progressive loss of water.

In the middle rank ranges (high volatile bituminous), coal enters the "oil window," through which lower molecular weight hydrocarbons are cleaved from the matrix. Most of these substances remain entrapped within the coal structure, either joining or displacing other components of the molecular fraction.

In higher rank bituminous coals, entrapped molecular constituents are either cracked to lower molecular weight fragments or repolymerized with the matrix. It is during this stage of coalification that most (vitrinite rich) coals begin to lose large amounts of hydrogen, indicating expulsion of hydrocarbon by products, especially methane.

Hydrocarbon by products that remain entrapped within the coal structure can be removed with varying degrees of difficulty. Small molecules such as methane can be easily removed merely by reducing ambient fluid (partial) pressure or by mild heating. However, higher molecular weight constituents can be removed only through strong solvent extraction or by more intensive heating (about 300 C.).

Largely because of the inherent difficulty in distinguishing molecular vs. matrix constituents, weight percentage estimates of molecular fraction present in coal (Fig. 3) are variable and somewhat ambiguous, depending on the analytical method used.6 7

Moisture contents, which are less ambiguous, decrease dramatically from over 75% in peat to as little as 1% (or less) in high volatile A bituminous coals.

The percentage of organic constituents removable by solvent extraction is more uncertain, however, and may vary widely for any given coal depending upon the solvent system used.

Pyridine extract yields, which are among the highest measured, reach a maximum of around 25 35% for some high volatile A bituminous coals, then decrease once again at higher ranks (Fig. 3). Coals exhibiting the highest pyridine yields correspond to the peak of the oil window.

The proportion of mobile components in coal may also be derived from interpretation of nuclear magnetic resonance that provides a nondestructive, in situ view of the coal structure. Although the significance of these data is debated, the proportion of so called mobile phase (in dry coal) is also interpreted to pass through a maximum at the peak of the oil window, attaining values estimated as high as 40%.6

CLASSIFICATION

In spite of its great compositional complexity, coal is usually described and classified according to extremely simplistic compositional parameters, such as "volatile matter" yield and vitrinite reflectance.

Volatile matter, as determined through "proximate analysis,"2 represents the weight fraction of coal that is vaporized during rapid pyrolysis at 950 C. The non volatile residue (exclusive of ash) is termed "fixed carbon."

On a dry, mineral matter free basis, fixed carbon or volatile matter yield may be used for ASTM rank classification.2

Although very useful, such parameters can mask the inherent heterogeneity in coal composition, particularly in terms of its behavior as a hydrocarbon source rock and reservoir. For example, proximate analysis, which is normally conducted on a bulk sample, is incapable of recognizing the diverse assemblage of truly volatile constituents entrapped within the coal.

Therefore, it would be unjustified to conclude that all coals having the same fixed carbon yields are fundamentally similar in their gas generative capabilities and reservoir behavior (Table 2). Similar limitations apply to the use of vitrinite reflectance as a compositional parameter.

The concept of coal as having a distinct, solid structure gives rise to a number of questionable conclusions regarding coals' behavior as a gas reservoir. For example, coal has typically been viewed as a porous, solid substrate upon which gaseous molecules are physically adsorbed.

In fact, neither porosity nor surface area are fixed properties of coal, but vary for different sorbates. Coal structure swells measurably during sorbate penetration, even for very weak solvents such as CO2,13 14 and CH4.15 Thus, sorbate molecules are not merely occupying vacancies within the coal structure, they are interacting with and causing a physical rearrangement of the coal.

Moreover, the molecular sieve properties of coal16 indicate that the regions within the coal structure accessible to sorbates are typically on the order of Angstrom units in dimension. The physical significance of porosity and surface area is highly questionable at this scale.17

In light of these issues, Franklin18 and Fuller19 have suggested the term accessibility rather than porosity to describe the ambiguous relationship between transient molecular species and the more stable substrate.

DRY GAS

The second oversimplification is that coalbed methane is a dry gas.

Early data from the U.S. Bureau of Mines regarding the composition of coalbed gases were based on analyses of gases released in underground mine atmospheres ("fire damp") and on a limited number of core desorption tests.20

Preliminary results indicated that gas from coal is comprised predominantly of methane, usually with insignificant amounts of other hydrocarbons. This observation contributed to describing natural gas from coal seams as "coalbed methane."

These conclusions, however, were based almost entirely on data for Carboniferous age coals, of mid to upper bituminous rank range, from the eastern and central U.S. Thus, these conclusions are not representative of coals worldwide.

OTHER CONSTITUENTS

While it is still generally true that most coals desorb mostly methane, it is now widely recognized that gas compositions can be highly variable. Scott, et al.,21 report that the proportion of C2+ hydrocarbons ranges from near zero to over 14%. They note CO2 as high as 10% or more from coalbed gas wells in the San Juan basin.

Coalbed gases in U.K. coal fields have reported up to 25% C2+ and significant N2.22

CO2 is reported for coals in parts of Australia, where its origin is believed to be igneous rather than sedimentary.23 24 CO2 also is found in parts of France and Silesia.25

Some San Juan basin coals are found with high N2 concentrations.

DESORPTION TESTS

The compositions of these desorbed gases are not fully representative of the composition of occluded gases. Desorbed gases are released at a rate that is partly related to their actual concentration and partly related to rate of diffusion (so called "chromatographic effect").26

Moreover, core desorption tests must be recognized as representing a re equilibration of coal composition to a particular set of ambient conditions (normally atmospheric pressure and "room" or "reservoir" temperature).

In describing the results of desorption testing, the term "gas yield" is always preferable to the more commonly used term "gas content."

This distinction is important in two respects. First, the coal does not contain gas per se in appreciable quantities. Rather, gases are "formed" and evolved when the physical conditions of the coal are changed. So called "gas contents" represent the quantity of gas liberated upon re equilibration of the coal to a new set of ambient pressure and temperature conditions.

Second, in the standard desorption test procedure, core samples inside the desorption canister are typically equilibrated to a pressure equivalent to ambient atmospheric pressure. Internal pressure, however, is due to whatever mixture of gases resides in the canister at the cessation of desorption testing.

For a gassy coal, which will have flushed out most of the head space gases (usually atmospheric air, entrapped at the time the canister was sealed), this is usually mostly methane. Current research at the University of Alabama shows that 1 atm partial pressure of methane can "hold" a significant amount of residual CH4 within the coal. This capacity is a consequence of the isotherm curve being very steep at low pressures.

This may be an important consideration in reservoir evaluation and may partly explain the phenomenon of "residual gas" (the gas which remains in the coal after desorption rate has fallen to very low levels, and which is liberated rapidly when the coal is pulverized in ambient air in a sealed grinding mill).

Residual gas has conventionally been interpreted solely as a consequence of slow diffusion of gas out of the coal but may, in part, represent the gases remaining in equilibrium under (approximately) 1 atm methane pressure.

Higher molecular weight substances tend to bond more strongly to the coal structure27 and diffuse more slowly.28 Thus, they are not desorbed as readily as methane.

Nitrogen diffusion rates may be even faster than methane.29 Typically, the percentage of higher hydrocarbons liberated in desorption tests increases with time.30

The percentage also increases at higher temperatures. For example, data reported by Kim31 indicate that gases desorbed from coal at 35 C. were nearly 100% methane, while gases desorbed from the same coals at 150 C. were only 30 60% methane.

This relationship is to be expected if gas transport on a molecular scale is partly an "activated" process.32 Data on actual production gases are scanty, but there is evidence of increasing CO2 content with time in some San Juan basin wells, and some evidence of increasing percentage of higher hydrocarbons with time.33

METHANE GENERATION

The third oversimplification is that all coals generate enormous quantities of methane during coalification.

Virtually all references estimating methane generated during coalification cite the same original source, a study published by Juntgen and Karweil in 1966.33

Juntgen and Karweil's study provided an excellent well reasoned, innovative treatment of the topic of gas generation and storage by coal. Nevertheless, their widely published methane generation curve, based upon mass balance approach, used a number of clearly stated simplifying assumptions that limit its applicability to all coals in all settings.

For example, in calculating their methane yields,33 they use a maturation pathway representing a vitrinite rich, hydrogen poor coal, typical of coals of Carboniferous age. Other coals, such as inertinite rich Permian Gondwana coals or perhydrous lower Tertiary coals will follow maturation paths that are significantly different. These coals may differ also in their volatile by products, both in quantity and composition.

Juntgen and Karweil also assumed that methane was the only hydrocarbon liberated during maturation. This assumption may not be true for all coals, particularly for geologically younger, oil prone coals.

Mass balance considerations dictate that if higher hydrocarbons are expelled by coal,34 35 then yields of methane must be diminished proportionately. This inevitable conclusion is borne out by recent pyrolysis experiments.

In calculating their methane generation curves, Juntgen and Karweil first assumed that the entire fraction of coal present as volatile matter at low rank was ultimately liberated as volatile by products at higher rank, but this assumption yielded a negative mass balance for water, which is quite improbable in nature.

When they made the alternative assumption that CH4 and CO2 were the only products released, they obtained positive results, but this assumption is inconsistent with data on the molecular structure of coal.16 More recent estimates of methane generation, based upon pyrolysis experiments or mass balance calculations have yielded lower, but probably more realistic values.36 These estimates have never gained the popularity of Juntgen and Karweil's original curve, perhaps because of their less impressive numbers.

As noted by Juntgen and Karweil33 and again by Meissner,37 realistic, accurate estimates of gas generation are important from the standpoint of basin modeling in order to estimate when coalbeds first become saturated with methane and thereafter, serve as source rocks for conventional gas reservoirs.

The timing and quantities of methane and other hydrocarbons generated during maturation vary for different coals and should be evaluated on a case by case basis for any particular basin. Therefore, it is inadvisable to use a single curve for all coals, or to assume that coals of a particular rank have or have not generated sufficient methane to become fully saturated.

SORPTION CAPACITY

The fourth oversimplification is that gas sorption capacity increases with coal rank.

One of the most widely reproduced diagrams in the coalbed methane literature depicts the methane content of coal as a function of depth for coals of various rank.38 Although these curves are very useful in demonstrating some important generalized relationships, they should be interpreted subject to the qualifications that:

• Not all underground coals are fully saturated with methane.

• Fixed carbon (or volatile matter) yield is not the sole compositional parameter needed to accurately predict sorption capacity.

• Gas sorption capacity does not uniformly increase with rank.

It has long been established that the accessibility of coals to gaseous sorbates generally decreases with increasing carbon content (for typical vitrinite rich coals), passing through a broad minimum around 85 90 wt % C, and increasing once again at higher ranks. This relationship has been demonstrated for N2, Kr, CO2, Xe, He, H2O, hexane, methanol, and benzene.18 39 41 Several studies have determined that methane sorption exhibits a similar trend with rank, although there is quite a bit of scatter in the data.25 42 43

The decreasing accessibility is apparently due to "plugging" of the "pore structure" by secondary hydrocarbons formed within the oil window.40

The previous relationships apply specifically to single component sorbate systems. When multicomponent systems are evaluated, competition between or among different sorbates for accessibility to the coal structure can considerably complicate the picture.44 45

Of particular relevancy to coalbed reservoirs is the competition between water and methane. Both are abundant naturally in underground coal seams.

Laboratory experiments demonstrate that the methane sorption capacity of low to medium rank coals is strongly suppressed when the coals are initially saturated with moisture, as compared with the methane capacity of the same coal on a dry basis.46

Because of lower moisture accessibilities in higher rank coals, however, methane sorption is only slightly suppressed by moisture.

When methane and water are considered together as a system, total sorbate accessibility decreases with rank. But methane capacity may exhibit a net increase for moisture saturated coals.

It has been generally assumed that all underground coal seams are saturated with moisture. However, this may not be the case.

Some coal seams in the San Juan basin contain no free water in their fracture systems47 and produce essentially no water during production.48

While it has not been proven that these coals are undersaturated with respect to their maximum moisture holding capacity, circumstantial evidence suggests that they are. In any case, it cannot be assumed a priori that all underground coal seams are water saturated.

Interrelationships of multicomponent sorbate systems in coal are quite complicated and, as yet, poorly documented. On the basis of available data, it is clear that the composition of the molecular fraction should be evaluated in its entirety.

In this regard, the suppressed accessibility of gaseous sorbates for coals in the oil window may be attributed to the increasing concentration of higher hydrocarbons in the molecular phase. The destruction and/or repolymerization of these constituents at higher rank is accompanied by an increased accessibility of gaseous sorbates.

Even at a given rank, however, the variable composition of the molecular phase may influence methane sorption. As an example, Table 2 and Fig. 4 pertain to the sorption capacity of two coalbeds in the Fruitland formation, northwestern San Juan basin.

The samples were collected from the same well within 100 ft of one another, stratigraphically. Fixed carbon yield of the two was essentially identical, and vitrinite reflectance was similar. Nevertheless, the coals differed substantially in the composition of their molecular fraction, as revealed by Rock Eval pyrolysis and solvent (methylene chloride) extract studies.

The stratigraphically lower coal contained a much higher concentration of waxy hydrocarbons. In terms of the accessibility of sorbates to the coal structure, the waxy coal exhibited equilibrium moisture capacity and CO2 surface area roughly 50% of that of the other coal.

Gas yields from the two coals, measured by desorption testing, showed roughly twice as much gas desorbed from the less waxy coal (Fig. 4a). Methane sorption capacity (Fig. 4b) was also slightly lower in the waxy coal.

Although few data are available from which to draw strong conclusions, it is clear that compositional factors other than fixed carbon yield need to be considered in understanding variability in gas yields of San Juan basin coals. Anecdotal evidence suggests that similar complications may influence hydrocarbon production from coals in other U.S. coal basins, as well.

GAS YIELDS

The fifth oversimplification is that gas yields of coal can be accurately predicted on the basis of rank and depth.

The preliminary sample set analyzed by the U.S. Bureau of Mines indicated that the coals studied were methane saturated for the most part,38 suggesting that high concentrations of methane are to be found ubiquitously in coal seams at depth. This is by no means invariably true. Many areas have been documented worldwide where gas yields from coal are anomalously low in comparison with these standard curves.

In several cases, these unexpectedly low gas concentrations are related to former burial depths having been less at some time in the geologic past than they are today, as documented in the U.K.,22 the Black Warrior basin, Silesia, and other localities.

This evidence indicates that once occluded gases escape from the coal structure, they are not readily replenished, even given tens of millions of years.

In the Black Warrior basin of Alabama, gas yields are extremely variable (Fig. 5). Large parts of the basin, particularly in the western regions, are significantly undersaturated with methane.

The low gas yields are apparently related in large measure to a buried unconformity at the top of the coal bearing Pottsville formation, which implies that the coal bearing rocks were much less deeply buried during the late Mesozoic than they are today.

McFall, et al.,49 reported that 60% of the variability in gas concentration of high volatile coals in the Black Warrior basin is related to variation in present day burial depth.

In fact, their data actually show a statistical correlation (r2) of only 14.0% (Fig. 5).

Because very little compositional data are available for these samples, it is impossible to determine what portion of the remaining variability (86%) is attributable to compositional controls and what proportion to other geologic variables.

In either case, it is clear from a variety of evidence that the simple model relating gas yields solely to variations of depth and rank (as determined via proximate analysis) is not satisfactory in many cases.

REFERENCES

1. ICF, The coalbed methane resource and the mechanisms of gas production, Gas Research Institute, 1989.

2. ASTM, American Society for Testing & Materials Standards, D 121 91, D 2796 88, D 3172, Philadelphia, 1992.

3. Bates, R.L., and Jackson, J.A., Glossary of geology, 2nd Edition, Falls Church, Va., American Geological Institute, 1980.

4. Vahrman, M., "The smaller molecules an overlooked component of coal," Chemistry in Britain, Vol. 8, 1972, pp. 16 22 and 24.

5. Given, P., "An essay on the organic geochemistry of coal," in Gorbaty, M.L, et al., eds., Coal Science, Vol. 3, New York, Academic Press, 1984, pp. 63 252 and 339 341.

6. Given, P.H., Marzec, A., Barton, W.A., Lynch, L.J. and Gerstein, B.C., "The concept of a mobile or molecular phase within the macromolecular network of coals: a debate," Fuel, Vol. 65, 1986, pp. 155 159.

7. Derbyshire, F., et al., "Molecular structure of coals: a debate," Fuel, Vol. 68, 1989, pp. 1091 1106.

8. Davidson, R.M., "Molecular structure of coal," in Gorbaty, M.L., et al., eds., Coal Science, Vol. 1, Academic Press, New York, 1982, pp. 83 160.

9. Herod, A.A., Stokes, B.J., and Radeck, D., 1991, Coal structure and mass spectrometry: volatiles in peat and coals, Fuel, Vol. 70, 1991.

10. Stansfield, E., and Gilbart, K.C., "Moisture determination for coal classification," AIME Transactions, Vol. 101, 1932, pp. 125 143.

11. Rees, O.W., Reed, F.H., and Land, G.W., "A study of the equilibration method of determining moisture in coal for classification by rank," Illinois State Geological Survey Report of Investigations, Vol. 58, 1939.

12. Selvig, W.A., and Ode, W.H., "Determination of moisture holding capacity (bed moisture) of coal for classification by rank," U.S. Bureau of Mines Report of Investigations 4968, 1953.

13. Reucroft, P.J., and Patel, K.B., "Surface area and swellability of coal," Fuel, Vol. 62, 1983, pp. 279 284.

14. Reucroft, P.J., and Patel, H., "Gas induced swelling in coal," Fuel, Vol. 65, 1986, pp. 816 820.

15. Briggs, H., and Sinha, R.P., "Expansion and contraction of coal caused respectively by the sorption and discharge of gas," Proceedings of the Royal Society of Edinburgh, Vol. 53, 1933, pp. 48 53.

16. Van Krevelen, D.W., Coal Amsterdam, Elsevier, 1981.

17. Spencer, D.H.T., and Bond, R.L., "The determination and use of specific surface values for coal," in Gould, R.F., ed., Coal science: Advances in Chemistry Series 55: Washington, D.C., American Chemical Society, 1966, pp. 724 730.

18. Franklin, R.E., "A study of the fine structure of carbonaceous solids by measurements of true and apparent densities," Transactions of the Faraday Society, Vol. 45, 1949, pp. 274 286.

19. Fuller, E.L., Jr., "Physical and chemical structure of coals: Sorption studies," in Gorbaty, M.L. and Ouchi, K., eds., Coal Structure: American Chemical Society Advances in Chemistry Series 192, 1981, pp. 293 309.

20. Kim, A.G., "The Composition of Coalbed Gas" U.S. Bureau of Mines Report of Investigation, Vol. 7762, 1973.

21. Scott, A.R., Kaiser, W.R., and Ayers, W.B., Jr., 1991, Composition, distribution, and origin of Fruitland Formation and Pictured Cliffs sandstone gases, in Schwochow, S.D., ed., Coalbed methane of western North America, Denver, Rocky Mountain Association of Geologists, pp. 93 108.

22. Creedy, D.P., 1988 Geological controls on the formation and distribution of gas in British coal measure strata: International Journal of Coal Geology, Vol, 10, pp. 1 31.

23. Smith, J.W., and Gould, K.W., "An isotopic study of the role of carbon dioxide in outbursts in coal mines," Geochemical Journal, Vol. 14, 1980, pp. 27 32.

24. Gould, K.W., Hargraves, A.J., and Smith, I.W., "Variation in the composition of seam gases issuing from coal," Bull. Proc. Australasian Institute of Mining and Metallurgy, Vol. 292, 1987, pp. 69 73.

25. Patching, T.H., "The retention and release of gas in coal a review," The Canadian Mining & Metallurgical Bulletin, Vol. 63, 1970, pp. 1302 1308.

26. Wyman R.E., "Gas resources in Elmworth coal seams," in Masters, J.A., ed., Elmworth a case study of a deep basin gas field, AAPG Memoir 38, 1984, pp. 173 187.

27. Ruppel, T.C., and Grein, C.T., and Bienstock, D., "Absorption of methane/ethane mixtures on dry coal at elevated pressure," Fuel, Vol. 51, 1972, pp. 297 303.

28. Sevenster, P.G., "Diffusion of gases through coal," Fuel, Vol. 38, 1959, pp. 403 418.

29. Creedy, D.P., and Pritchard, F.W., "Nitrogen and carbon dioxide in U.K. coal seams," International Journal of Mining, Vol. 1, 1983, pp. 71 77.

30. Kim, A.G., "Experimental studies on the origin and accumulation of coalbed gas," U.S. Bureau of Mines Rep. of Inv. 8317, 1978.

31. Nandi, S. P., and Walker Jr., P. L, "Activated diffusion of methane from coals at elevated pressures," Fuel, Vol. 54, 1975, pp. 81 86.

32. Von Schonfeldt, H., Pothini, B.R., Aul, G.N., and Henderson, R.L., "Production and utilization of coalbed methane gas in Island Creek Coal Company mines," SPE/DOE Unconventional Gas Recovery Symposium, Pittsburgh, 1982, pp. 235 238.

33. Juntgen, H., and Karweil, J., "Formation and storage of gases in bituminous coal seams, Pt. 1, Gas formation & Pt. 2. Gas storage (English translation)," Erdol und Kohle Erdgas Petrochimie, Vol. 19, 1966, pp. 251 258 and 339 344.

34. Durand, B., and Paratte, M., "Oil potential of coals," in Brooks, J., ed., Petroleum Geochemistry and Exploration of Europe, Oxford, Blackwell Scientific Publications, 1983, pp. 285 292.

35. Cook, A., and Struckmeyer, H., "The role of coal as a source rock for oil," in Glennie, R.C., ed. Second Southeastern Australia Oil Exploration Symposium, Petroleum Exploration Society of Australia, Melbourne, 1986, pp. 419 432.

36. Juntgen, J., and Klein, J., "Entstehung von Erdgas aus kohligen Sedimenten," Erdol und Kohle Erdgas Petrochemie, Erganzungsband, Vol. 1, 1975, pp. 52 69.

37. Meissner, Fred, F., "Cretaceous and Lower Tertiary coals as sources for gas accumulations in the Rocky Mountain area, Hydrocarbon source rocks of the greater Rocky Mountain region," Rocky Mountain Association of Geologists, Denver, 1984, pp. 401 430,

38. Kim, A.G., "Estimating methane content of bituminous coalbeds from adsorption data," U.S. Bureau of Mines Report of Inv. 8245, 1977.

39. Walker Jr., P.L. and Kim, K.A., "Measurement of the ultrafine surface area of coals," Fuel, Vol. 44, 1965, pp. 453 459.

40. Thomas Jr., J., and Damberger, H.H., "Internal surface area moisture content, and porosity of Illinois coals variations with rank," Illinois State Geological Survey Circular No. 493, 1976.

41. Nelson, J.R., Mahajan, O.P., and Walker, P.L., Jr., "Measurement of swelling of coals in organic liquids: a new approach," Fuel, Vol. 59, 1980, pp. 831 837.

42. Moffat, D.H., and Weale, K.E., "Sorption by coal of methane at high pressures" Fuel, Vol. 34, 1955, pp. 449 462.

43. Schwarzer, R.R., Variations in the quantity of methane adsorbed by selected coals as a function of coal petrology and coat chemistry, Draft Final Report, U.S. Dept. of Energy Contract No. DE AC21 80MC14219, 1983.

44. Ruthven, D.M., Principles of adsorption and adsorption processes, New York, John Wiley Interscience, 1984.

45. Puri, R., and Yee, D. "Enhanced coalbed methane recovery," SPE Proceedings, New Orleans, Sept. 23 26, 1990, pp. 193 202.

46. Joubert. J.I., Grein, C.T., and Bienstock, D., 1974, "Effect of moisture on the methane capacity of American coals," Fuel, Vol. 53, 1974, pp. 186 191.

47. Decker, A.D., and Seccombe, J.C., "Geologic parameters controlling natural gas production from a single deeply buried coal reservoir in the Piceance basin, Mesa County, Colorado," Paper No. SPE 15221, SPE Unconventional Gas Technology Symposium, Louisville, Ky., 1986, pp. 163 172.

48. Boyce, B.C., and Levine, J.R., "Reservoir heterogeneity and gas production characteristics of oil rich Fruitland coals, San Juan basin, Colorado," Geological Society of America Abstracts with Programs, Vol. 23, 1991, P. A39.

49. McFall, K., Wicks, D.E., and Kuuskraa, V.A., "A geological assessment of natural gas from coal seams in the Warrior basin, Alabama," Gas Research Institute, GRI 86/0272, 1986.

Copyright 1992 Oil & Gas Journal. All Rights Reserved.