EXPLORATION Cambrian potential indicated in Kentucky Rome trough

Feb. 19, 1996
David C. Harris, James A. Drahovzal Kentucky Geological Survey Lexington A recent gas discovery in the Rome trough has Appalachian basin operators re-evaluating the deep Cambrian potential of eastern Kentucky. The Rome trough has seen sporadic exploration since the late 1940s, with very limited commercial success. A new exploration phase began in mid-1994 with completion of the Carson Associates 1 Kazee well in Elliott County, Ky. (Fig. 1 [35638 bytes]).
David C. Harris, James A. Drahovzal
Kentucky Geological Survey
Lexington

A recent gas discovery in the Rome trough has Appalachian basin operators re-evaluating the deep Cambrian potential of eastern Kentucky.

The Rome trough has seen sporadic exploration since the late 1940s, with very limited commercial success. A new exploration phase began in mid-1994 with completion of the Carson Associates 1 Kazee well in Elliott County, Ky. (Fig. 1 [35638 bytes]).

This well blew out and initially flowed 11 MMcfd of gas from a zone in the upper Conasauga Group/Rome formation at 6,258-70 ft. Some formation damage is reported to have occurred after the well was killed, although the well was producing about 500 Mcfd in mid-1995.

Carson Associates has recently permitted an offset well, the 33 Lawson Heirs, to 7,500 ft in the Rome formation.

Encouraged by Carson's Kazee well, Blue Ridge Group completed a second Elliott County Rome test in August 1995. The 1 Greene was drilled on a separate structure, and results have not yet been announced. Logs from this well have not been released, but early reports indicate that the Conasauga/Rome interval may be wet.

These two wells are characteristic of a frustrating exploration history in the Rome trough that is marked by numerous gas and oil shows but rare commercial production. Hydrocarbon shows in the trough date to 1947, when gas and oil shows were reported from the Rome formation in the South Central 1 Hall well in Powell County, Ky.1 2

A well in Boyd County, Ky., Inland 529 White, was completed as a Rome oil well in 1967 and produced about 30,000 bbl of oil. This well also produced about 90 Mcfd of gas from the same zone.

In 1980 the Lancaster 1 Lee well in Garrard County reportedly flowed 750 Mcfd of low-Btu gas from Rome formation sandstones but was never commercially produced. Numerous other wells have reported shows of oil or gas.

The first pre-Knox gas well in the Rome trough was the Exxon 1 McCoy in Jackson County, W. Va. An initial open flow of 9.2 MMcfd and sustained production of 5.6 MMcfd was reported from a sandstone in the Conasauga Group at 14,350-360 ft. This well produced dry gas for about 6 months and had a total cumulative production of 427 MMcf before an increasing water cut forced plugging. The McCoy well holds the record for deepest production in the Appalachian basin.

In 1986, a second commercial gas well was reported from the Rome trough. Ashland 1 Williams, in Johnson County, Ky., was completed in the Conasauga Group (Rome formation of Kentucky). Initial open flow was 1.055 MMcfd from 6,250-6,350 ft in a fractured shale interval. A slight show of condensate was also mentioned in the completion report. This well was still producing in 1995, but production data are not available. Neither the McCoy nor the Williams well was offset, and both remain single-well pools.

No drilling to pre-Knox objectives occurred in the trough for the next 8 years. In 1994, Carson Associates successfully completed the 1 Kazee well as a Conasauga/ Rome producer in Elliott County, Ky. This well tested a structure interpreted from recently reprocessed seismic data originally acquired by Exxon during the 1970s. Future exploration will rely heavily on enhanced seismic interpretation of complex structures in the graben.

Stratigraphy

Stratigraphic units below the Cambro-Ordovician Knox Group on the stable shelf in northeastern Kentucky consist of (in descending order) the Conasauga formation, the Rome formation, and the Mt. Simon sandstone (Fig. 2 [39406 bytes]). These units together are 300-600 ft thick in northernmost Kentucky, but they abruptly thicken across a series of extensional growth faults (the Kentucky River fault system) into a northeast-southwest trending rift basin, the Rome trough.3-7

In the trough, the pre-Knox section includes older and dramatically thickened units, with as much as 10,000 ft of pre-Knox sediments in some areas8 17 (Fig. 3 [81155 bytes]). Rocks as old as Early Cambrian are in the trough and have no equivalents to the north on the shelf. The Upper Cambrian Mt. Simon sandstone is not present south of the northern boundary fault and is probably equivalent to part of the upper Rome/ Conasauga sequence. The expanded Cambrian section in the Rome trough is largely conformable and predominantly marine in origin.9 A type log for the Rome trough sequence is shown in Fig. 4.[81155 bytes]

The pre-Knox sequence in the Rome trough in Kentucky consists of the Conasauga Group, Rome formation, Shady/Tomstown dolomite, and a basal sandstone zone (Fig. 2 [39406 bytes]). The Conasauga Group comprises a 2,400-5,500 ft thick sequence of shale, limestone, dolomite, and siltstone that conformably overlies the Rome formation in the Rome trough. The Rome formation is a 1,000-2,900 ft thick sequence of sandstone, siltstone, shale, limestone, and dolomite that overlies the Shady/Tomstown dolomite. Depositional environments of the Conasauga Group and Rome formation were interpreted by Donaldson et al.,10 who studied cores from two West Virginia wells. They described carbonate and clastic facies deposited in tidal flat, tidal channel, and shallow subtidal marine environments. In Kentucky, turbidite sandstones and potential for fan delta and basin-floor fan deposits were interpreted by Webb6 and Drahovzal.18 19

Confined to the Rome trough and the deeper Appalachian basin to the southeast is a carbonate unit, the Shady/Tomstown dolo- mite, which underlies the Rome formation and overlies a basal sandstone.1 9 The Shady/Tomstown dolomite interval consists of limestone in some areas, and is considered to be Early Cambrian in age.

The oldest sedimentary unit in the trough is a sandstone, informally named the basal sandstone, which un- conformably overlies middle Proterozoic crystalline basement rocks of the Grenville province.1 6 7 This sandstone unit ranges from 20-650 ft thick and is commonly arkosic. Previously, this basal sandstone in the Rome trough was correlated with the Mt. Simon sandstone on the craton. The Mt. Simon appears to be a much younger unit, and probably correlates with the uppermost part of the Rome formation in the trough.9

Structure

The Rome trough is the most significant structure affecting pre-Knox rocks in the Appalachian basin. The trough is part of a larger Cambrian extensional feature, extending northeastward across West Virginia and Pennsylvania into southern New York and is associated with the similar-age Rough Creek graben and Reelfoot rift farther west in western Kentucky and Tennessee, eastern Missouri, Arkansas, and northwestern Mississippi1-15 (Fig. 5 [40927 bytes]). The trough, first recognized by Woodward,3 is bounded in Kentucky by the Kentucky River fault system and the Irvine-Paint Creek fault system on the north, and the Rockcastle fault system on the south1 5 6 16 (Fig. 1 [35638 bytes]). The southern bounding fault system is more discontinuous than the northern zone.

The Rome trough has been interpreted as a failed Cambrian continental rift basin that developed approximately 250 miles cratonward of the central area of Iapetan rifting. Maximum subsidence within the trough occurred during the Middle Cambrian, based on paleontological data and the thickness of the Rome formation. Cambrian depocenters are located along the northern rift margin in Kentucky and the southern margin in West Virginia, indicating that the rift is characterized by half-graben structures of alternate polarity.

A recently published structural map of the Precambrian basement surface in eastern Kentucky17 illustrates the regional configuration for the trough in Kentucky (Fig. 6 [43144 bytes]). In eastern Kentucky, structural relief on the top of the Precambrian Grenville basement rocks is greater than 13,000 ft from the northern boundary to the deepest part of the Rome trough. Relief along the southern boundary is generally only about 7,000-8,000 ft and locally much less. Farther south, basement deepens into the Appalachian basin. The resulting geometry is therefore a flat-bottomed graben bounded by a high northern shoulder that steps down across a series of down-to-the-south normal faults and a lower south-bounding shoulder that deepens into the graben across a few down-to-the-north faults. The graben generally deepens to the east, ranging from about -5,000 to -8,000 ft along the Grenville front to more than -17,000 ft at the Kentucky-West Virginia border.

The dominant normal faults are oriented east-northeast (Fig. 1 [35638 bytes]). Two north-south-oriented faults bound the Floyd County channel in southeastern Kentucky, separating the Pike County uplift from the Perry County uplift. The latter uplift, together with the Rockcastle River uplift, lies south of the Rockcastle River fault system. North of the Floyd County channel is a known north-south surface structure, the Paint Creek uplift. It is along this uplift and its flanks that the two recent wells were drilled. Current publicly available data neither confirm nor deny an uplift in the basement rocks along this axis. Reverse faulting in the basal part of the pre-Knox section is interpreted to represent a Cambrian wrench component that resulted in local transpressional structures.17

Reservoir items

Potential reservoir facies in the Rome trough consist of sandstones, carbonates, and fractured shales. Sandstones and fractured shales have been responsible for most of the production to date, but the dolostone intervals of the Conasauga Group and Shady/Tomstown dolomite may have reservoir potential. Rome formation sandstones have quite variable reservoir characteristics, and are difficult to predict. These sandstones are typically fine to very fine grained, micaceous, and glauconitic. Coarser grained facies may occur in proximity to major border faults, where fan-delta deposits are likely.6 Limited seismic data may indicate such deposits as well as basin-floor fans7 18 19 (Fig. 7 [38290 bytes]). Porosity data for the three commercial pre-Knox wells in the Rome trough are limited. The Ashland 1 Williams well produces from a fractured shale, but the borehole is washed out over this interval, invalidating the porosity logs. The Exxon 1 McCoy produced from a 10 ft thick sandstone that averaged 11% porosity. The producing zone in the Carson Associates 1 Kazee well was not logged.

Trapping mechanisms

Structural traps are the primary target in and around the Rome trough, where basement-controlled normal faults influenced deposition and created potential structural traps during initial rifting and later reactivation. The three commercial gas wells that have been reported all appear to be structurally influenced.

The Ashland 1 Williams well in Johnson County, Ky., produces from a fractured shale interval in the Con- asauga Group (Rome formation of Kentucky). This well is near the Irvine-Paint Creek fault zone in the Rome trough, and fracturing is thought to be related to proximity to this fault. The Exxon 1 McCoy, discovery well for Belgrove field in Jackson County, W. Va., produced for about 6 months from a probable fault-related four-way closure in the Rome trough. Data are limited for the recent Carson Associates 1 Kazee well in Elliott County, Ky., but it appears to be a fault block trap.

Stratigraphic traps, al- though much higher risk, also represent potential reservoirs. Stratigraphic trapping could exist in both sandstones (including turbidite fans) and carbonates in the Conasauga, Rome, and Shady/Tomstown intervals due to depositional pinchout, erosional truncation, and diagenetic variability. Potentially sand-rich fan deposits may exist in both the shallower block south of the Kentucky River fault system (Fig. 7 [38290 bytes]) and the deep block south of the Irvine-Paint Creek fault zone18 19 (Fig. 1 [35638 bytes]). These inferred fans represent po- tential reservoirs that in some areas may be discontinuous in their distribution. Careful analysis of existing and new seismic data will be required to map these potential reservoir facies. Such facies, especially where combined with structure, could result in significant hydrocarbon traps.

Source rocks

Potential source rocks for pre-Knox hydrocarbons in the Rome trough are not well constrained. Stratigraphic separation of the pre-Knox interval from the Knox unconformity makes Upper Ordovician shales an unlikely source in these areas. Oil produced from the Rome-Conasauga interval in eastern Kentucky is distinguished by high gravity (41-54), unlike mostly lower gravity oils derived from post-Knox source rocks. This suggests that both oil and gas in pre-Knox reservoirs were generated from pre-Knox source rocks at higher thermal maturities. Ryder et al.20 analyzed 22 shale samples from the Rome and Cona- sauga interval in three wells in the Rome trough of West Virginia and an outcrop section in Tennessee. Total organic carbon (TOC) values of these samples range from 0.05 to 0.59%, and they are considered to have low to marginal source potential.

Ryder et al. also calculated production indices (PI) for samples with TOC greater than 0.5% using pyrolytic yields (S1 and S2). Average PI values for the pre-Knox samples range from 0.4 to 0.6, indicating that the interval sampled is in the gas generation window. Richer Cambrian source rocks may occur elsewhere in the basin and may have reached relatively high thermal maturity, generating the characteristic gas, condensate, and high-gravity oil found in pre-Knox reservoirs.

Gas composition

Gas composition varies significantly within the trough. Gas from the Exxon 1 McCoy well had an almost pure methane composition, while gas from the Ashland 1 Williams has heavier components, consistent with reports of some produced condensate. Both wells had Btu values of over 1,000.

At least two wells in the western part of the Rome trough in Kentucky have encountered significant shows of low-Btu, non-combustible gas. The wells are located in Garrard County and yielded Rome gas with an average methane content of 15% and an average nitrogen content of 78%. These wells were of interest as a possible source of helium, which averaged an unusually high content of 1.6%. The origin of these high-nitrogen gases is not known, but Garrard County is located near the Grenville front, a major tectonic suture between the Precambrian Grenville province and the Kewee- nawan(?) East Continent rift basin.21 A deep basement origin for some of this low-Btu gas is possible. The risk of low-Btu gas appears confined to the western Rome trough, since gas produced farther east (Johnson and Elliott Counties, Ky., and Jackson County, W. Va.) is of commercial quality.

Summary

Additional details of the Cambrian pre-Knox play in the Appalachian basin can be found in Harris and Baranoski.22 Recent activity in the Rome trough has proven the presence of commercial accumulations of gas in Cambrian pre-Knox reservoirs.

We feel additional discoveries will be made, but success in the trough will require acquisition and interpretation of high-quality seismic data to delineate complex structures and reservoir facies. In addition, detailed stratigraphic and diagenetic studies will be required to allow prediction of reservoir zones.

The Rome trough remains largely a geologic puzzle. Integration of regional stratigraphy with the structural history will provide the key to successful exploration in this play. The Kentucky Geological Survey is in- volved in some aspects of this work and welcomes industry involvement in this effort.

Acknowledgments

A portion of this work was completed in preparation of a chapter on the pre-Knox play for "The Atlas of Major Appalachian Basin Gas Plays," and funded by the U.S. Department of Energy, Morgantown Energy Technology Center under contract number DE-FC21-91MC28176. This project also funded part of the preparation of "Preliminary Map of the Structure of the Precambrian Surface in eastern Kentucky" (Kentucky Geological Survey, Map and Chart Series 8, 1995). We also acknowledge review and editing of the manuscript by M.C. Noger and Meg Smath. Thanks to Mike Murphy and Collie Rulo for preparation of the figures. TerraSciences, Inc. provided software for generation of log plots, maps, and cross sections.

References

1. McGuire, W.H., and Howell, P., Oil and gas possibilities of the Cambrian and Lower Ordovician in Kentucky, Lexington, Ky., Spindletop Research Center, 1963.

2. Weaver, O.D., and McGuire, W.H., Cambro-Ordovician potential of the Rome trough of eastern Kentucky, OGJ, Nov. 14, 1977, pp. 250-255.

3. Woodward, H.P., Preliminary subsurface study of southeastern Appalachian Interior plateau, AAPG Bull., Vol. 45, No. 10, 1961, pp. 1,634-55.

4. Silberman, J.D., Cambro-Ordovician structural and stratigraphic relationships of a portion of the Rome trough, in Hutcheson, D.W., ed., Proceedings of the technical sessions, Kentucky Oil and Gas Association 34th and 35th annual meetings, 1970-71, Kentucky Geological Survey, Ser. 10, Special Pub. 21, 1972, pp. 35-45.

5. Silberman, J.D., Exploration along the northwestern margin of the Rome trough, in Luther, M.K., ed., Proceedings of the technical sessions, Kentucky Oil and Gas Association 38th annual meeting, June 6-7, 1974, Kentucky Geological Survey, Ser. 11, Special Pub. 3, 1981, pp. 20-30.

6. Webb, E.J., Cambrian sedimentation and structural evolution of the Rome trough in Kentucky, Ph.D thesis, University of Cincinnati, 1980.

7. Sutton, E.M., Deep exploration in eastern Kentucky by the SCLAW Group during the early seventies, in Luther, M.K., ed., Proceedings of the technical sessions, Kentucky Oil & Gas Association 38th annual meeting, June 6-7, 1974, Kentucky Geological Survey, Ser. 11, Special Pub. 3, 1981, pp. 31-44.

8. Ryder, R.T., Harris, A.G., and Repetski, J.E., Stratigraphic framework of Cambrian and Ordovician rocks in the Central Appalachian basin from Medina County, Ohio, through southwestern and south-central Pennsylvania to Hampshire County, W. Va., USGS Bull. 1839-K, 1992.

9. Ryder, R.T., Stratigraphic framework of Cambrian and Ordovician rocks in the Central Appalachian basin from Morrow County, Ohio, to Pendleton County, W. Va., USGS Bull. 1839-G, 1992.

10. Donaldson, A.C., Heald, M.T., and Warshauer, S.M., Cambrian rocks of the Rome trough in West Virginia: Cores from Mingo and Wayne Counties, in Smosna, Richard, organizer, A walk through the Paleozoic of the Appalachian basin: AAPG Eastern Section Meeting Core Workshop, Charleston, W. Va., Sept. 13, 1988, pp. 6-18.

11. Wagner, W.R., Growth faults in Cambrian and Lower Ordovician rocks of western Pennsylvania, AAPG Bull., Vol. 60, No. 3, 1976, pp. 414-427.

12. Beardsley, R.W., and Cable, M.S., Overview of the evolution of the Appalachian basin, Northeastern Geology, Vol. 5, 1983, pp. 137-145.

13. Harper, J.A., Effects of recurrent tectonic patterns on the occurrence and development of oil and gas resources in western Pennsylvania, Northeastern Geology, Vol. 11, 1989, pp. 225-245.

14. Thomas, W.A., The Appalachian-Ouachita rifted margin of southeastern North America, GSA Bull., Vol. 103, 1991, pp. 415-431.

15. Potter, C.J., and Drahovzal, J.A., The regional configuration of the Cambrian Reelfoot-Rough Creek-Rome rift system (abs.), in Ridgley, J.L., Drahovzal, J.A., Keith, B.D., and Kolata, D.R., eds., Proceedings of the Illinois Basin Energy and Mineral Resources Workshop, Sept. 12-13, 1994, Evansville, Ind., Kentucky Geological Survey Open-File Report OF-94-12 (Illinois State Geological Survey Open-File Report 94-4; Indiana Geological Survey Open-File Report 94-12; USGS Open-File Report 94-298), pp. 34-35.

16. Cable, M.S., and Beardsley, R.W., Structural controls on Late Cambrian and Early Ordovician carbonate sedimentation in eastern Kentucky, American Journal of Science, Vol. 284, 1984, pp. 797-823.

17. Drahovzal, J.A., and Noger, M.C., Preliminary map of the structure of the Precambrian surface in eastern Kentucky, Kentucky Geological Survey, Ser. 11, Map and Chart Series 8, 1995.

18. Drahovzal, J.A., Basin-floor fan complexes in Cambrian rift basins of Kentucky (abs.), AAPG annual convention, June 12-15, 1994, Denver, Vol. 3, p. 139.

19. Drahovzal, J.A., The structure, stratigraphy, and future hydrocarbon potential of the Rome trough in Kentucky (abs.), 25th Annual Appalachian Petroleum Geology Symposium, Morgantown, W. Va., I.C. White Memorial Fund Publication 6, 1994, pp. 33-38.

20. Ryder, R.T., Burruss, R.C., and Hatch, J.R., Geochemistry of selected oil and source rock samples from Cambrian and Ordovician strata, Ohio-West Virginia-Tennessee, part of the Appalachian basin, USGS Open-File Report 91-434, 1991.

21. Drahovzal, J.A., Harris, D.C., Wickstrom L.H., Walker, Dan, Baranoski, M.T., Keith, Brian, and Furer, L.C., The East Continent Rift Basin: A new discovery, Kentucky Geological Survey, ser. 11, Special Pub. 18, 1992.

22. Harris, D.C., and Baranoski, M.T., Play Cpk: Cambrian pre-Knox Group Play, in Patchen, D.G., ed., Atlas of major Appalachian basin gas plays, West Virginia University, in press.

The Authors

David C. Harris is a geologist in the Petroleum and Stratigraphy Section at the Kentucky Geological Survey, University of Kentucky, Lexington. Before joining the survey in 1990, he spent 6 years with BP Exploration and Sohio, where he worked as a carbonate sedimentologist, and 2 years with Mobil Exploration and Producing Southeast as an exploration geologist. He has an MS degree in geology from State University of New York and a BS degree in geology from the College of William and Mary.
James A. Drahovzal is geologist and head of the Petroleum and Stratigraphy Section at the Kentucky Geological Survey, Lexington, where he has been for 7 years. He spent 10 years in petroleum geology research with Gulf Research & Development Co. and ARCO Research and Technology Services. Earlier he spent 12 years with the Geological Survey of Alabama, where he worked in Appalachian and Paleozoic geology. He holds a PhD degree in geology from the University of Iowa.

Copyright 1996 Oil & Gas Journal. All Rights Reserved.