Possible Continuous-Type (Unconventional) Gas Accumulation in Lower Silurian "Clinton" Sands, Medina Group, and the Tuscarora Sandstone in the Appalachian Basin: A Progress Report of 1995 project activities

Robert T. Ryder, Kerry L. Aggen, Robert D. Hettinger, Ben E. Law, John J. Miller, Vito F. Nuccio, William J. Perry, Jr., Stephen E. Prensky, John R. SanFilipo, and Craig J. Wandrey

Open-File Report 96-42

SANDSTONE CHARACTER AND DEPOSITIONAL SEQUENCES ALONG SELECTED TRANSECTS

Four transects were chosen to characterize the stratigraphic framework of the "Clinton" sands, Medina Group, Tuscarora Sandstone and adjoining units. Adjoining units include the Upper Ordovician Queenston Shale, Lower Silurian Clinton Group, and Lower and Upper Silurian Lockport Group. The Medina, Clinton, and Lockport Groups constitute the Niagaran Provincial Series. Three of the transects, B-B' (New York and Pennsylvania), C-C' (Pennsylvania and Ohio), and D-D' (Ohio and West Virginia), are between 120 and 185 mi long and trend northwest-southeast, approximately subparallel to the depositional dip of the Lower Silurian sandstone system (fig.2). A 425-mi-long, northeast-southwest trending transect, A-A' (New York, Pennsylvania, and Ohio), is oriented approximately subparallel to the depositional strike of the Lower Silurian sandstone system and connects the three northwest-southeast trending transects (fig.2). All transects are constructed with geophysical well logs, primarily of the gamma ray/density or gamma ray/neutron type. About 35 to 40 wells are used for each of the northwest-southeast transects and about 75 to 85 wells are used for the northeast-southwest transect. Regional sections shown in Cate(1965), Knight(1969), Piotrowski(1981), and numerous M.S. theses assisted correlations between many of the control points. In addition to detailed correlations of reservoir sandstone units and the overlying Reynales/Irondequoit/Packer Shell/Dayton carbonate units, the transects contain information such as production status of wells, perforated interval(s), type and amount of gas and(or) fluids produced, formation pressure, and bottom-hole temperature.

Transect B-B'

Transect B-B' extends roughly 40 miles across Chautauqua County, New York (fig. 3 ). The 80- to 90-mi-long Pennsylvania part of transect B-B' has not yet been completed. The N35°W-S35°E trend of the transect is oriented nearly perpendicular to paleoshoreline trends as interpreted from Lower Silurian Medina Group sandstones and the Tuscarora Sandstone by Cotter (1983). Rocks of the Medina Group dip southeastward across the transect and occur at depths of 2,260 to 2,420 ft at the northwest end of the transect and 4,480 to 4,640 ft at the southeast end. Geophysical logs from 43 drill holes, spaced approximately 1 mi apart, are used to correlate the Medina Group and adjoining strata. The drill holes are identified and located in table 1.

Type and reference sections for the Niagaran Provincial Series are described from exposures along the Niagara Escarpment, located 55-120 mi north and northeast of transect B-B'. Brett and others (1995) give an excellent historical stratigraphic overview of these rocks and provide stratigraphic revisions based on their own research. Sequence stratigraphic interpretations for the Niagaran Series are provided by Brett and others (1990). Stratigraphic nomenclature, lithologic descriptions, and rock ages of the Niagaran Series, as used in this discussion, are based primarily on Brett and others (1990, 1995) and are summarized in table 2. Additional lithologic descriptions of the Medina Group are provided by Laughrey (1984) from drill-hole core about 40 mi west of the transect in northwestern Pennsylvania.

Preliminary correlations of the Medina, Clinton, and lower Lockport Groups along transect B-B' are shown in figure 4. Correlated formations are the Whirlpool Sandstone, Power Glen Shale, and Grimsby Formation of the Medina Group; the Reynales Limestone, Williamson Shale, Irondequoit Limestone, and Rochester Shale of the Clinton Group; and undivided strata of the Lockport Group. The Devils Hole Sandstone, Thorold Sandstone, Cambria Shale, Kodak Sandstone, Neahga Shale, and Rockway Dolomite are not recognized along the transect. The undivided Lockport Group contains strata that are coeval to the Gasport, Goat Island, Ermosa, and Guelph Dolomites.

Three laterally continuous depositional units are recognized in the siliciclastic strata that comprise the Medina Group along transect B-B' (fig. 5). The basal unit is 15-25 ft thick and consists of a fining-upward succession in the Whirlpool Sandstone and the lower part of the Power Glen Shale. The middle unit is 35-65 ft thick and consists of coarsening-upward successions in the upper part of the Power Glen Shale and lower part of the Grimsby Formation. The upper unit comprises the remaining 75 to 100 ft of the Medina Group and consists of laterally discontinuous units of sandstone, siltstone, and mudrock interbedded in coarsening- and fining-upward successions. In the Niagara region, these three units of the Medina Group are interpreted to contain shelf margin, transgressive, and highstand deposits within Sequence 1 (Brett and others, 1990)(table 2). However, stacking patterns observed in the Medina Group along transect B-B' suggest alternative sequence stratigraphic interpretations whereby the lower and middle units, respectively, comprise the transgressive and highstand systems tracts of Sequence I and the upper unit is within the transgressive systems tract of Sequence II (table 2). These interpretations are based on similarities of these rocks to sequence stratigraphic models described from Upper Cretaceous strata in southern Utah by Shanley and McCabe (1991) and Hettinger and others (1994).

The lower unit overlies the Cherokee unconformity (Brett and others, 1990) and fines upward from sandstone to siltstone and mudrock (fig. 5). Coeval rocks in the Niagaran Escarpment region demonstrate a significant basinward facies shift as well as a deepening upward facies succession through fluvial, shoreface, and offshore marine deposits. Fluvial deposits are placed in a shelf margin systems tract and the shoreface and offshore deposits are placed in a transgressive systems tract by Brett and others (1990). Similar interpretations are reasonable for strata within the lower unit of transect B-B'. However, all three facies may have been deposited within a transgressive systems tract during an overall retrogradational event whereby shallow incised valleys are backfilled with fluvial deposits during a rise in sea level. In both examples, overlying shoreface and offshore marine strata were deposited during a marine transgression in response to a rise in sea level.

Coarsening-upward successions in the middle unit are 20-40 ft thick and pass upward through mudrock, siltstone, and sandstone. Sandstones are 10-20 ft thick and persist at least 10 mi in a northwest direction before they split, thin, and grade into siltstone and mudrock (fig. 5). Similar overlying successions forwardstep to the northwest and downlap onto the lower unit. These upward-coarsening and forward stepping characteristics suggest that the middle unit is a highstand deposit comprised of progradational shoreface parasequences. Interpretations of offshore and barrier bar deposits in coeval rocks in Pennsylvania (upper part of the Cabot Head Shale and lower part of the Grimsby Sandstone) by Laughrey (1984) support a shoreface interpretation. Correlations from the middle unit suggest that the upper part of the Power Glen Shale and Devils Hole Sandstone in the Niagara region are progradational rather than retrogradational as interpreted by Brett and others (1990). If this scenario is correct, the maximum flooding surface in the Niagara region may be a 2-ft-thick condensed interval at the top of the Whirlpool Sandstone (see Brett and others, 1990, fig. 4) rather than the base of the Artpark Phosphate Bed as interpreted by Brett and others, 1990).

The upper unit contains laterally discontinuous sandstone, siltstone, and mudrock interbedded in fining- and coarsening-upward successions. Siltstone and mudrock dominate the upper part of the unit. The basal contact is scoured into underlying shoreface strata and scours are as much as 35 ft deep and several miles across (see drill holes 21 through 32, fig. 4). The upper unit is interpretated to be fluvial and tidal in origin based on geophysical log signatures and core descriptions of the Grimsby Sandstone which contain braided river deposits in its middle part and tidal flat deposits in its upper part (Laughrey (1984). We interpret the basinward shift of braided river over nearshore strata to represent an unconformable facies relationship; this interpretation is further supported by the widespread erosional surface that separates the two facies along the transect. Strata in the upper unit is further interpreted to be within a transgressive systems tract based on its deposition over an erosion surface and retrogradational stacking patterns. Deepening-upward successions are characterized by braided river and superimposed tidal flat deposits in the Grimsby Sandstone (Laughrey, 1984), inner shelf muds of the Neahga Shale, and offshore carbonates of the Reynales Limestone (Brett and others, 1990). In this interpretation, broad valleys were cut into emergent highstand deposits of Sequence I and backfilled by fluvial strata during the initial stages of sea level rise that resulted in the deposition of Sequence II of Brett and others (1990). The fluvial deposits were subsequently covered by tidal deposits as the sea advanced. Continued rise in sea level resulted in marine transgression that cut a ravinement surface and deposited the pebble and cobble lag of the Densmore Creek Phosphate Bed followed by deposition of inner shelf muds and offshore carbonates. The top of the transgressive systems tract is interpreted to be eroded by the unconformity at the base of Sequence III following interpretations of Brett and others (1990).

Properly interpreted sequence boundaries, systems tracts, and depositional facies may be critical for understanding reservoir and production variability in the Clinton/Medina regional gas accumulation.

For example, thick valleyfill sandstones of fluvial and(or) estuarine origin deposited in a transgressive systems tract may be an important control of production "sweet spots". This investigation of transect B-B' provides an alternative sequence stratigraphic interpretation for the Medina Group and lower part of the Clinton Group in New York State; however, core studies and additional detailed correlations are needed to substantiate the interpreted unconformity near the base of the Grimsby Formation and the tidal and fluvial origin of Grimsby Formation strata above the unconformity. Moreover, perforated gas-bearing zones and their initial and ultimate gas yield must be identified for each well of transect B-B' to determine which sequences, systems tracts, and depositional facies have the most favorable gas reservoirs.

Transect C-C'

A preliminary east-west cross section, several miles north of transect C-C', was prepared from seven drill holes used in deWitt and others (1975)(figs. 6, 7). This section shows the general stratigraphic relationships of the Lower Silurian Medina and Clinton Groups prior to studies by Piotrowski (1981), Laughrey (1984), and Zagorski (1991). For example, except for cross sections by Cate (1961) where a combined Irondequoit-Reynales lime was identified, the Irondequoit (Limestone) has only recently been recognized as a separate unit in northwestern Pennsylvania (Laughrey, 1984).

Twenty wells (table 3) along transect C-C' have been correlated with the Mark Resources Shaffer #1 well in the Cooperstown gas field (Zagorski, 1991)(fig. 7). Three of these wells are shown on figure 8: the Transamerica Acker #1 (API 37-039-20077), the Mark Resources Shaffer #1 (API 37-121-42719) and, to the east, the Quaker State Fee #1-H (API 37-053-21250). Only the Benedum Trees J. Kardosh #1 drill hole (P67; API 37-039-20007) is common to the two cross sections (fig. 6). The Reynales Limestone of the earlier preliminary cross section has been reclassified into three carbonate units in transect C-C'. In ascending order, these carbonates are named, respectively, 1) the Irondequoit Limestone, 2) an unnamed limestone unit, and 3) the Reynales Limestone as restricted by Zagorski (1991). These latter correlations are shown in figure 8 and are used throughout the correlation of logs along transect C-C'. Care must be taken in the correlation of the carbonate intervals above the Medina Group sandstones. It would appear that the Dayton Limestone (Packer Shell) has been confused with the Reynales Limestone in western Pennsylvania.

Commonly, the bulk density of the basal part of the Reynales Limestone in the western part of transect C-C' is high with respect to the bulk densities of the younger carbonates. The most likely explanation is that the basal unit of the Reynales Limestone is dolomite rather than limestone.

A preliminary examination of bulk density values within the sandstone intervals of the Whirlpool Sandstone, Cabot Head Shale, and the Grimsby Formation along transect C-C' indicates that the Whirlpool Sandstone commonly has very low porosity. Porosity values calculated for sandstones in the Cabot Head Shale and the Grimsby Formation commonly range from 3-4%. These porosity values calculated from well logs, however, are considerably lower than the 5-8% porosity values measured from core in nearby wells (Laughrey, 1984).

Transect D-D'

Lack of funds for geophysical logs in Ohio prevented the completion of transect D-D' in FY1995. This transect should be completed in FY1996.

Transect A-A'

During FY95 geophysical logs from about 100 drill holes were correlated across transect A-A', thus, providing us with a reasonably good understanding of the distribution of Lower and lower Upper Silurian strata across the length of Clinton/Medina gas play 6728. The transect is incomplete across several 10- to 20-mi-wide gaps where well logs have not been purchased. For this report, seven drill holes were selected to characterize the preliminary stratigraphic correlations and framework along transect A-A' (figs. 9, 10). Stratigraphic nomenclature for the Medina and Clinton Groups in New York follows Brett and others (1995). In most cases, the New York terminology for the Medina and Clinton Groups can be extended into Pennsylvania (Piotrowski, 1981; Pees, 1983; Laughrey, 1984). In Ohio, a combination of commonly used driller's terms and formal nomenclature are applied to strata equivalent to the Medina and Clinton Groups (Pepper and others, 1953; DeBrosse and Vohwinkle, 1974; Gray and others, 1985).

One important aspect of the transect is that it shows regional

thickness variations of the Medina Group in New York and Pennsylvania and of equivalent strata (Medina sand, Cabot Head Shale, "Clinton" sands, upper tongue of the Cabot Head Shale) in Ohio (fig.10). The maximum thickness of the Medina Group and equivalent strata along transect A-A' is located in the vicinity of Mercer County, PA (drill hole 4) and Columbiana County, OH (drill hole 5)(figs. 9, 10). The 210- to 220-ft-thick Medina Group and equivalent strata there constitute part of the Canton embayment, a depocenter recognized by Knight (1969) consisting of Medina Group siliciclastic units, lowermost Clinton Group carbonate units, and equivalent strata. From this depocenter, the Medina Group thins northward to about 96 ft in Erie County, NY (drill hole 1) and southward to about 156 ft in Washington County, OH (drill hole 7) (figs. 9, 10).

A second important aspect of the transect is that it shows the net thickness and geometry of sandstone in the Medina Group, "Clinton" sands, and Medina sand. Net sandstone thicknesses in the Medina Group and equivalent strata in the seven selected drill holes are greatest in Mercer County, PA (drill hole 4, 96 ft), Chautauqua County, NY (drill hole 2, 90 ft), and Columbiana County, OH (drill hole 5, 85 ft)(figs. 9, 10). The Chautauqua County, NY site, located nearly 100 mi north of the Canton embayment, represents a subsidiary sandstone depocenter in the "Clinton" sand and Medina Group sequence. The net sandstone thickness of the Medina Group interval thins to 64 ft in Erie County, NY (drill hole 1) at the north end of the transect and to 29 ft in Washington County, OH (drill hole 7) at the south end (figs. 9, 10).

Basal sandstone units in the Medina Group and equivalent units (Whirlpool Sandstone and Medina sand) are characterized by an 8- to 20-ft-thick, well-defined low ("clean") gamma ray log response that gradually changes upward to a higher response (higher clay content)(fig.10). This log shape has been interpreted by Metzger (1981) and Laughrey (1984) to represent sublittoral sheet sandstones. Based on outcrop studies, Middleton and others (1987) suggest that the lower part of the Whirlpool Sandstone has been deposited in a braided fluvial environment. Sandstone units in the Devils Hole Sandstone, upper part of the Cabot Head (Power Glen) Shale, lower part of the Grimsby Formation, and lower part of the "Clinton" sands are 10 to 30 ft thick and commonly have blocky to upward-decreasing ("cleaner") gamma ray log responses (fig. 10). These log shapes have been interpreted in previous investigations to represent littoral (barrier bars and tidal deltas), distributary mouth bar, distributary channel, tidal-dominated delta, tidal ridge, and offshore marine bar sandstones (Metzger, 1981; Laughrey, 1984; Keltch, 1985). Directly overlying these shallow-marine sandstones are 2- to 14-ft- thick sandstone units in the upper part of the Grimsby Formation and "Clinton" sands that have spike-shaped to upward-increasing (higher clay content) gamma ray log responses (fig.10). These log shapes have been interpreted in earlier investigations to represent braided fluvial channel, meander point bar, and tidal channel sandstones (Metzger, 1981; Laughrey, 1984; Keltch, 1985).

In this report, the Whirlpool Sandstone and Medina sand are interpreted as shoreface or sublittoral sheet sandstones, possibly with a lowermost braided fluvial component, deposited in a transgressive systems tract that unconformably overlies the Upper Ordovician Queenston Shale. This unconformity is recognized as the Cherokee unconformity by Brett and others (1990). Following the preliminary interpretation of the middle unit in transect B-B' (fig. 5), sandstones in the lower part of the Grimsby Formation and "Clinton" sands with blocky and upward-decreasing gamma ray responses are identified here as shoreface sandstones deposited in a highstand systems tract (fig.10). These shoreface sandstones constitute parts of progradational parasequences that successively overlap one another toward the northwest, pinch out seaward into offshore marine shale of the Cabot Head and Power Glen Shales, and appear to downlap across the underlying transgressive systems tract. Although transect A-A' is largely oriented normal to depositional strike, it has a slight dip-oriented component that cuts obliquely across the paleoshoreline from north to south in a seaward direction and exhibits locally the stacking and westward progradational character of the parasequences (figs. 9, 10). The Devils Hole Sandstone of Brett and others (1990, 1995) and informally named sandstones in the Cabot Head Shale (Pees, 1983; Laughrey, 1984; Zagorski, 1991) are interpreted here to be part of the same progradationally stacked parasequences. Following the interpretation of the upper unit in transect B-B' (fig. 5), sandstones with spike-shaped and upward-increasing gamma ray responses that overlie the shoreface sandstones are identified here as estuarine (tidal) and fluvial sandstones in a transgressive systems tract deposited unconformably across a highstand systems tract (fig.10). The unconformable contact between the proposed highstand and transgressive systems tracts is subjective because of the variety of gamma ray log signatures involved. This boundary may mark a previously unrecognized sequence boundary (unconformity) caused by a eustatic fall in sea-level (see discussion of transect B-B').

A third important aspect of transect A-A' is that it shows the regional distribution of carbonate units in the Clinton Group (Reynales Limestone, Irondequoit Limestone, Dayton Limestone, Packer Shell of drillers). These units are commonly used as a datum horizon for stratigraphic cross sections (transects) through the "Clinton" sands and Medina Group sandstones or the contoured horizon on structural contour maps. figure10 shows that, although these carbonate units are regionally extensive, long-distance correlations between them can be erroneous because of pinch outs and thickness changes in the intervening shale beds. The Irondequoit and Reynales Limestones extend across most of transect A-A' and the latter unit is used as the datum horizon (fig.10). However, at the north end of the transect in New York State they are separated from one another by 2 to 5 ft of shale whereas at the south end in Ohio they are separated by a 70-ft-thick shale and a 10-ft-thick unnamed limestone unit. This unnamed limestone extends as far north as Crawford County, PA. The Dayton Limestone (Packer Shell of drillers) is located stratigraphically about 45 ft beneath the Reynales Limestone in southern Ohio (drill hole 7, Washington County, OH) and extends as far north as northwestern Pennsylvania (drill hole 4, Mercer County, PA) where it underlies the Reynales Limestone by less than 10 ft (figs. 9, 10).

Created by the EERT WWW Staff.