Characteristics of discrete and basin-centered parts of the Lower Silurian regional oil and gas accumulation, Appalachian basin: Preliminary results from a data set of 25 oil and gas fields

U.S. Geological Survey Open-File Report 98-216

COMPARISON OF SELECTED OIL AND GAS FIELD CHARACTERISTICS

Depth to Production and Hydrocarbon Types

A plot of depth to production vs. hydrocarbon type for the 25-field data set shows several clusters of data points (fig. 4). The depth to production for the oil and gas fields in the data set ranges from about 2,000 ft along the western margin of the discrete part of the accumulation in east-central Ohio (Homer, Appendix L, and Lancaster/Sugar Grove, Appendix O, gas fields) to as much as 6,200 ft near the eastern margin of the Clinton/Medina part of the basin-centered part of the accumulation in northwestern Pennsylvania (Oil Creek pool of the Cooperstown gas field, Appendix J).

Hydrocarbon types produced in the data set range from oil and associated gas to nonassociated gas. Intermediate categories of hydrocarbon type in the data set are 1) gas and associated oil and 2) gas and local associated oil. These general classes of hydrocarbon type are roughly determined by their gas-to-oil ratio (GOR) (fig. 4). For example, a GOR of 20,000 (cu ft of gas): 1 (barrel of oil) or greater defines a gas well or field. Few GORs are available for the fields in the data set so their location as plotted on the horizontal axis of figure 4 is very approximate. Several of the fields noted as having nonassociated gas as their hydrocarbon type also produce some condensate.

Most of the data fall in two slightly overlapping clusters (fig. 4). One of the data clusters consists primarily of fields in the discrete part of the regional accumulation that range in depth from about 2,800 to 5,100 ft and whose hydrocarbon types are either oil and associated gas or gas and associated oil. The second data cluster consists of fields in the basin-centered part of the accumulation that range in depth from about 4,300 ft to 6,200 ft and whose hydrocarbon types are either nonassociated gas or gas with local associated oil. The overlap of these data clusters is caused by gas and associated oil at about 4,000 ft in the Conneaut gas field (Indian Springs and Kastle pools; Appendices G, H) (fig. 2). Perhaps this overlap indicates that the Conneaut field has been misidentified as belonging to the basin-centered part of the accumulation. In this region of northwestern Pennsylvania, the discrete- basin-centered boundary of Ryder (1995) could be moved eastward to include the Conneaut gas field with the discrete part of the regional accumulation. Although obvious overlap occurs, figure 4 suggests that a depth of about 5,000 ft and a GOR of about 100,000:1 are threshold values to distinguish discrete verses basin-centered accumulations. Moreover, the addition of equivalent vitrinite reflectance (%R_o) isograds from Middle Ordovician strata about 1,000 to 1,500 ft below the Lower Silurian (Wandrey and others (1997) to figure 4 suggests that %R_o = 1.10 may represent another threshold value for discriminating discrete versus basin-centered accumulations. Equivalent %R_o isograds may be significant for discriminating between discrete and basin-centered accumulation because they identify areas of peak gas generation and, thus, favorable regions for basin-centered accumulation. Law and Dickinson (1985) report that in Rocky Mountain basin examples of basin-centered accumulation, %R_o = 0.94 corresponds to the top of active overpressuring, a temperature of 180° F (82° C), and the top of active gas generation.

Several small anomalous data clusters on figure 4 indicate that nonassociated gas can be present at shallow depths (2,000 to 3,000 ft) in the discrete and basin-centered parts of the accumulation. Shallow nonassociated gas in the Homer and Lancaster/Sugar Grove gas fields (fig. 2) characterizes a band of fields along the western margin of the discrete part of the accumulation that extends from northern Kentucky to Ontario, Canada. Moreover, shallow nonassociated gas in the Lakeshore gas field characterizes the basin-centered part of the accumulation (Ryder, 1995) in most of western New York State and adjoining Ontario, Canada (fig. 2). These large areas of shallow nonassociated gas that seem out of place with respect to their apparent thermal maturation history must be explained in petroleum generation and migration models that are proposed for the regional accumulation (Nuccio and others, 1997).

Structural Setting and Natural Fractures

The dominant structure associated with the Lower Silurian hydrocarbon accumulation is a gentle southeast-dipping homocline (<1° ) that forms the northwest flank of the Appalachian basin (Knight, 1969; Boswell and others, 1993). In the 25-field data set, this dominant north- to northeast-striking structural trend commonly has a subtle overprint of structural terraces, anticlinal noses, faults, and fractures (table 2). Minor closure is noted on several of the anticlinal noses. Moreover, several of the oil and gas fields are closely associated with probable basement-controlled, cross-strike structures such as the Cambridge arch (Baranoski, 1993; Root and Martin, 1995) and the Tyrone-Mt. Union lineament (Rodgers and Anderson, 1984). The conspicuous absence of structural closure in the fields of the data set supports Knight’s (1969) interpretation that structure has a negligible effect on entrapment in the regional accumulation. However, structure is considered important for controlling: 1) the segregation of oil, gas, and water in the reservoir (Knight, 1969), 2) conduits for hydrocarbon migration, 3) zones of preferential reservoir drainage (Bush and others, 1987), and 4) the distribution and character of naturally fractured reservoirs (Core, 1986).

Natural fractures are documented or inferred in over one-half of the fields in the data set (table 2). Many of the fractures accompany regional structures cited in table 1 and probably share a common origin with them. In the Best gas field (Appendix C) and the Carbon Hill oil field (Appendix D), measured fracture orientations in the "Clinton" reservoir are dominated by northwest- and northeast-trending sets. Outcrops of Devonian strata near the Athens gas field (Appendix B) show similar fracture orientations. Very likely, the northwest-southeast oriented fracture sets were caused by compressive stresses during the Alleghanian orogeny (Engelder and Geiser, 1980; Engelder, 1985), whereas the northeast-southwest oriented sets were caused by contemporary compressive stresses on the crust of the eastern and midcontent regions of the U.S. (Zoback and Zoback, 1980; Engelder, 1982). Most petroleum-industry geologists recognize the importance of natural fractures for creating high-yield oil and gas wells in the Clinton and Medina sandstone reservoirs (Sitler, 1969; Alexander and others, 1985; Core, 1986). In wells stimulated by hydrofracturing, these natural fractures improve the permeability of the reservoir and the subsequent drainage of hydrocarbons into the wellbore. Also, Zagorski (1991) suggests that natural fractures are important conduits for migrating formation water. According to Zagorski, this water has dissolved chemically unstable grains, such as feldspar, to form highly productive gas-bearing zones with secondary porosity. Although concentrations of natural fractures are very important for predicting production sweet spots, they are widely distributed across both the discrete and basin-centered parts of the regional accumulation and, thus, have little discriminatory value.

Trap Types

Although most oil and gas fields in the Lower Silurian regional accumulation are identified as stratigraphic-trap fields (Knight, 1969; McCormac and others, 1996), many of them do not demonstrate a well-defined updip pinchout of the sandstone reservoir(s) (table 3). Entrapment of oil and (or) gas in these situations may be explained by several mechanisms: 1) subtle updip changes in depositional and (or) diagenetic facies and 2) high-water saturation in low- to moderate-permeability rocks (water block).

Subtle updip changes in reservoir character, for depositional and (or) diagenetic reasons, have been documented as the cause of stratigraphic entrapment of oil and gas in many shallow-marine and coastal terrigenous-clastic sequences (Harms, 1966; Reinert and Davies, 1976; Wood and Hopkins, 1992). Depositional and (or) diagenetic changes may create subtle increases in capillary pressure in a given reservoir that, in turn, form a barrier to updip oil and gas migration (Berg, 1975; Schowalter, 1979; Vavra and others, 1992). The height of the hydrocarbon column is a direct measure of the effectiveness of the trap (Berg, 1975). That is, the thicker the hydrocarbon column the greater the differential capillary pressure between the reservoir and trap facies. Berg (1975) reports that permeable and water-bearing facies may form an effective barrier to hydrocarbon migration. Moreover, capillary pressures required to trap natural gas are higher than those required to trap oil. Undoubtedly, the majority of the stratigraphic traps in the Lower Silurian regional accumulation are controlled by similar factors.

The Athens gas field is one of a small number of Clinton/Medina fields that have been comprehensively studied for depositional and diagenetic facies variability and their effect on hydrocarbon entrapment (Laughrey, 1984). Laughrey reports that there is no obvious updip pinchout of the reservoir sandstones in the field but they do show an updip loss of porosity due to increased cementation. He proposes that gas may originally have been trapped in a paleostructure and cementation occurred at an associated gas-water contact beneath it. In a later episode of tilting that formed the present-day homocline, gas was kept in place by the pre-tilt zone of cementation. Gas is produced from the trapping facies but at lower initial rates than is produced from the reservoir facies (Laughrey, 1984).

Depositional facies of the Senecaville gas field (Appendix W) were studied in detail by Keltch and others (1990). They show that the field is a stratigraphic-trap accumulation caused by the updip pinchout of reservoir sandstone of distributary mouth bar and distributary channel origin. Four time slices through the 185-ft-thick "Clinton" sands-Cabot Head Shale interval show marked thickness variations of individual sandstone reservoirs ranging from 2 to 24 ft.

Except for the East Canton/Magnolia oil field (Appendix K; fig. 2) and its large associated gas cap, hydrocarbon columns in the "Clinton" sands of Ohio are relatively thin, perhaps averaging 50 ft thick or less. Such relatively thin hydrocarbon columns suggest that the trapping facies have capillary pressures that only slightly exceed those of the reservoir facies and, consequently, updip leakage of oil and gas has been a common condition of the regional accumulation. Schowalter (1979) postulated a stratigraphic-leak differential-entrapment model for such a setting where oil is trapped downdip from the gas. The model operates on the premise that the traps along a migration path are a series of displacement-pressure barriers that will hold a certain hydrocarbon column and leak gas preferentially to oil updip through the barrier before the trap is filled to its stratigraphic spillpoint (Schowalter, 1979). This model might account for the large gas fields that rim the western margin of the discrete part of the Lower Silurian regional accumulation in Ohio and Ontario, Canada. However, other origins for the gas also should be considered such as late-stage gas exsolution from oil during Mesozoic uplift and erosion of the Appalachian basin (R. C. Burruss, oral communication, September 1997).

High water saturation in low-permeability rocks was recognized by Masters (1979) as a dominant trapping mechanism for the deep basin Elmworth gas field in Lower Cretaceous strata of western Canada. This trapping mechanism differs from that accompanying conventional reservoirs because gas is overlain by water rather than the reverse. This mechanism of entrapment (commonly referred to as water block), is probably caused by moderate- to low-permeability rocks with high water saturation where relative permeability of gas to water is reduced to essentially zero (Masters, 1979; Price, 1995). Gies (1984) suggests that the Elmworth deep basin (basin-centered) gas accumulation of western Canada is in a dynamic state of updip migration.

Davis (1984) and Zagorski (1988, 1991) applied the basin-centered (deep basin) concept to the Lower Silurian regional accumulation of the Appalachian basin. In Ohio, Davis interprets the eastern margin of the water-block trap to conform approximately with the -3,500 ft subsea structure contour at the top of the "Clinton" sands. In the area east of this contour line, which includes Best (Appendix C), East Canton/Magnolia (Appendix K), and Claysville fields (Appendix E), the "Clinton" sands are considered to be part of a 2,000-ft-thick gas column with no associated formation water (fig. 5). According to Davis (1984), west of this line the "Clinton" sands and their hydrocarbon accumulations are associated with formation water. Zagorski (1991) extends the line of Davis (1984), and consequently the water-block mechanism of entrapment, into northwestern Pennsylvania along the updip margins of Sharon Deep (Appendix X), Kantz Corners (Appendix M), and Cooperstown (Appendix I) fields (fig. 5). There, the boundary between the water-block trap and the downdip gas accumulation is somewhat irregular as a result of its interaction with cross-cutting lineaments (Zagorski, 1996).

Of the 25 oil and gas fields in the data set, only the Cooperstown, Kantz Corners, and Sharon Deep fields—having a water-block mechanism of entrapment—can be clearly identified as belonging to the basin-centered part of the accumulation (table 3). The remainder of the fields in the data set, particularly those without a recognizable updip pinchout of the reservoir or a significant updip decrease in net sandstone thickness, have no distinguishing characteristics to classify them as either discrete or basin-centered accumulation. However, those fields characterized by an updip pinchout or decrease in thickness of the reservoir have a greater likelihood of being associated with the discrete part of the regional accumulation. Additional details of depositional and diagenetic facies variations and their associated capillary pressures are required to determine whether or not the traps are strictly stratigraphic. Data concerning the volume of produced water from these fields—to be discussed in a following section of the report—will provide a better understanding of the nature of entrapment. In particular, the produced water data are required to evaluate the disparity between the discrete and basin-centered accumulation boundaries identified by Davis (1984)-Zagorski (1991) and Ryder (1995).

Reservoir Porosity and Permeability

A plot of average reservoir porosity (F _ave) vs. average reservoir permeability (K_ave) for the 25-field data set shows an expected direct relation between these variables (fig. 6). Although tentative, this plot suggests two distinct reservoir types in the Lower Silurian regional accumulation.; one type with K_ave > 0.1 mD, F _ave > 10%, and another with K_ave < 0.1 mD, F_ave < 10%. Furthermore, K_ave = 0.1 mD is the threshold value used by the Federal Energy Regulatory Commission (FERC) to designate a tight (low-permeability) gas formation (Dutton and others, 1993), implying geologic significance to the regulatory limit.

Judging from figure 6, average reservoir permeability seems to be a reasonable first approximation for identifying discrete and basin-centered parts of the regional accumulation. Of the 12 fields in the data set whose reservoirs have an average permeability of 0.1 mD or less, 8 fields are located in the basin-centered part of the regional accumulation as defined by Ryder (1995). In contrast, those fields in the data set whose reservoirs have an average permeability of greater than 0.1 mD, 9 of 13 are located in the discrete part of the accumulation. Although slightly different fields are involved, a 0.1 mD threshold value of average reservoir permeability has about the same degree of success for differentiating water saturated (discrete) versus gas saturated (basin-centered) parts of the regional accumulation as defined by the boundary of Davis (1984) and Zagorski (1991).

The imperfect match between average permeability and accumulation type is further shown on the Grimsby-interval permeability map by Boswell and others (1993). This map suggests that highly irregular, northwest-oriented tongues having 0.3 to >1.0 mD permeability extend from the discrete part of the accumulation, where they predominate, into the basin-centered part of the accumulation where they eventually pinch out. In contrast, similarly oriented irregular tongues of lower permeability (<0.3 mD) are predominant in the basin-centered part of the accumulation but extend westward, tens of miles, into the discrete part of the accumulation.

Reservoir Water Saturation and Volume/Salinity of Produced Waters

Law and Dickinson (1985) postulate that most reservoir water accompanying active basin-centered accumulation is at irreducible saturation levels. However, they add that mobile water may re-enter the accumulation after temperature and pressure reduction following basin uplift and erosion. Capillary pressure curves that show irreducible water saturation of a given reservoir rock are normally unavailable for the Lower Silurian accumulation and, thus, could not be used to help discriminate between its discrete and basin-centered parts. A plot of average water saturation (S_w) vs. depth to production shows a general trend of decreasing S_w with depth (fig. 7). Although the trend of the plot is consistent with basin-centered accumulation, the plot appears to have little value for differentiating discrete versus basin-centered parts of the accumulation as defined by Ryder (1995). Part of the reason for the poor discriminatory value of the plot is because reservoir permeability has not been accounted for as it would have been had irreducible water saturation been used.

Castle and Byrnes (in press) report that most water saturation in the Cooperstown field (basin-centered part; Appendix I) is at irreducible levels. Here, irreducible water saturation (S_wi) ranges from 10 to 80 percent and varies inversely with porosity. Medina Group sandstones with porosity in the 6 to 8 percent range and S_wi < 20 % contain the majority of the gas storage capacity in the reservoir whereas those with porosity less than 3 percent and S_wi > 40% are non-pay.

Produced water has been reported for 20 fields in the data set (table 4) and, very likely, the remaining 5 fields will be reported as such when data are available. Of the 20 fields in the data set that produce water, 11 of them have information regarding the volume of produced water (table 4). Information ranges from incomplete reports that indicate "all wells produce some water" or "water production is low" to detailed reports that permit calculations of water production (table 4 fig.7). Water production per well is summarized in this report as barrels of water (BW) per million cu ft of gas (MMCFG). As suggested by the plot of water saturation (fig. 7), fields classified with the basin-centered part of the regional accumulation have lower volumes of produced water per well than fields classified with the discrete part of the accumulation (table 4). Four fields in the data set affiliated with the basin-centered part of the accumulation have an average yield per well that ranges from 9 to 13 BW/MMCFG (table 4; fig. 7). Water yields in this range or less may be a diagnostic feature of the basin-centered part of the accumulation. A water yield of 1,200 BW/MMCFG for the Logan field (Appendix Q) (fig.7) suggests high water yields for the discrete part of the accumulation.

Produced water from the Lower Silurian regional accumulation is classified as a brine [>35,000 mg/l total dissolved solids (TDS)] with sodium and chloride as dominant constituents and calcium and magnesium as major components (Stith, 1979; Breen and others, 1985; Lowry and others, 1988; Siegel and others, 1990; Rose and Dresel, 1990; Sanders, 1991). Sodium, calcium, and chlorine account for approximately 97% of the TDS (Sanders, 1991). Potassium and bromide are present in the brines as relatively concentrated minor components (Breen and others, 1985). These Na-rich brines in the Lower Silurian regional accumulation probably originated from the interaction of migrating connate water with beds of halite in the Upper Silurian Salina Group (Lowry and others,1988; Rose and Dresel, 1990; Siegel and others, 1990)(see fig. 3 for the stratigraphic position of the Salina Group). Also, Lowry and others (1988) recognize a second type of brine (Ca-rich) that they imply may have originated beneath the Salina Group in the Appalachian basin. This second type of brine may represent water expelled during basin-centered gas generation and accumulation in a manner proposed by Law and Dickinson (1985).

Brine salinity expressed as TDS (in mg/l or ppm) or as total concentration of common constituent elements (in ppm) is available for 14 of the fields in the data set. Salinity of the produced waters in the data set ranges from 147,000 to 327,000 TDS but is largely confined to a range of 200,000 to 300,000 TDS (fig. 7). In the data set used in this report there does not appear to be a correlation between salinity and depth to production or salinity and hydrocarbon type (fig. 7). However, larger data sets in Ohio suggest a regional eastward to southeastward decrease in Na-rich brine (Lowry and others, 1988) and divalent metal chlorides (Sanders, 1991). Sanders (1991) suggests that the southeastward decrease in divalent metal chlorides reflects an earlier stage of compaction-driven water flow toward the margin of the basin. Alternately, these waters characterized by a southeastward decrease in divalent metal chlorides could have been expelled during basin-centered gas generation and accumulation in a manner proposed by Law and Dickinson (1985).

Data in this report (table 4; fig. 7) reveal the important fact that water (brine) is produced from both discrete and basin-centered parts of the regional accumulation. Moreover, the data set suggests that fields in the basin-centered part of the accumulation produce less water than fields in the discrete part. Thus, although an oversimplification, the concept proposed by Davis (1984)—whereby the Clinton/Medina hydrocarbon accumulation is subdivided into water-bearing and water-deficient compartments—appears to be correct. Obviously more data are required to better quantify water production per well in terms of volume with respect to each MMCF of gas produced. Judging from the small data set gathered to date, the basin-centered part of the regional accumulation seems to be characterized by an average water production per well of 13 BW/MMCFG or less.

Segregation of Gas and Fluids in Reservoirs

Of the few fields in the data set where fluid contacts were evaluated, none have recognizable oil-water or gas-water contacts. However, by analogy to the Arabia gas field in Lawrence County, Ohio (fig. 2), studied by Zagorski (1996), a gas-water contact could be present at the Homer and Lancaster/Sugar Grove gas fields. Moreover, probable oil-water and gas-water contacts—although not reported as such—appear to be present in the Onondaga oil field (Harkness, 1935) and locally in the Norfolk field (MacDougall, 1973) in Ontario, Canada (fig. 2). These proposed oil-water and gas-water contacts in the Canadian part of the regional accumulation may be very localized because they have not been reported in published investigations (MacDougall, 1973; Cochrane and Bailey Geological Services Ltd., 1986).

Referring to the regional Clinton accumulation in general, Lockett (1929) reports that, "gas occurs in the higher parts of the sandstone reservoir and, where the reservoir is relatively continuous, considerable oil has accumulated in its lower parts." He did not mention oil-water or gas-water contacts. This published comment by Lockett (1929) also seems to apply to oil-bearing fields in the data set such as 1) Best gas field (Seibert, 1987) and East Canton/Magnolia oil field (Sitler, 1969) where oil most commonly occupies a structurally low position with respect to gas and 2) Ravenna and Mantua/ Shalersville (Wilson, 1988) gas fields where oil seems to be associated with the thickest sandstone reservoirs. Exceptions such as the Lenox gas field (Munsart, 1975) where oil occurs in the structurally highest parts of the field and the Northeast Salem gas field (Seibert, 1987) where local oil lies updip or lateral to gas suggest special circumstances caused by paleostructure or marked stratigraphic variability within the reservoir.

Although water is most commonly located in the structurally lowest part of the reservoir, such as in the Lenox gas field (Munsart, 1975), it can appear anywhere. For example, in the East Canton/Magnolia oil field, the structurally highest part of the field produces the most water (Schrider and others, 1969). Moreover, long gas-to-water transition zones are reported in the Athens gas field (Laughrey, 1984) and in the Norfolk gas field (MacDougall, 1973). Finally, as discussed by Zagorski (1991), a zone of high water saturation is located updip of the Cooperstown and Sharon Deep gas fields and appears to serve as the trap. Even in this situation, the boundary between gas and water is very transitional and most water-bearing units contain some associated gas.

Normally, classical discrete (conventional) accumulations show well-defined gas-oil-water, oil-water, or gas-water contacts due to the differential buoyancy of the fluid and gas involved. These well-defined oil-gas-water contacts are obviously missing in the discrete part of the regional accumulation in central and eastern Ohio, as defined by Ryder (1995). The absence of fluid-gas segregation begs the important question as to whether or not any of the oil and gas fields in central and east-central Ohio should be classified as discrete accumulations. A major cause of the poor oil-gas-water segregation observed here and the associated classification dilemma is pervasive reservoir heterogeneity. Significant contributors to reservoir heterogeneity are relatively thin, intertonguing, lenticular sandstone bodies; marked depositional and diagenetic facies variability; and relatively low porosity and permeability. Another factor contributing to the classification dilemma is human nature. Commonly, we overly simplify a natural system into two end members when in reality it is far more complex (T. S. Dyman, written communication, March 1998). Only those fields observed or surmised to have gas-water contacts along the western margin of the accumulation are recognized here as true discrete (conventional) accumulations. The remainder of the discrete part of the accumulation as defined by Ryder (1995) is neither a discrete or basin-centered type and, thus, following Zagorski (1996), is tentatively classified as a hybrid type of accumulation.

Segregation of gas and water by a different mechanism occurs at the broad transitional contact between the discrete and basin-centered parts of the regional accumulation (Davis, 1984; Zagorski, 1988, 1991). By analogy to basin-centered accumulations in western North America (Gies, 1984; Law and Dickinson, 1985), the zone of high water saturation (water-block trap) located updip from the zone of high gas saturation consists, in large part, of mobile pore water that was displaced from deeper in the basin by large-scale gas generation. The continuous gas phase and irreducible pore water left behind constitutes the basin-centered gas accumulation. This mechanism can only operate in low- permeability rocks (Gies, 1984; Law and Dickinson, 1985; Price, 1995). The interface between downdip gas and updip water fluctuates according to rates of gas generation and accumulation vs. rates of gas loss. Law and Dickinson (1985) and Spencer (1987) suggest that gas generation and its net accumulation coincides with an overpressured phase in the basin that occurs during maximum burial and high heat flow whereas gas loss coincides with an underpressured phase that occurs during regional uplift and erosion. According to Gies (1984), the updip limit of high gas saturation is controlled by a regional facies change to sandstone reservoirs of higher permeability. This facies change to more permeable rocks permits buoyant forces to become dominant and, thus, permit gas to move updip at a rate that exceeds the rate of gas influx from lower permeability rocks in the deeper part of the basin.

Reservoir Pressure and Bottom-hole Temperature

A plot of reservoir pressure vs. depth to production for the data set shows that most of the Clinton/Medina reservoirs are underpressured with respect to a normal hydrostatic gradient for salt water (fig. 8). The plot shows considerable scatter because 8 fields include pressures calculated by Thomas (1993) in addition to those reported from other sources. Most pressures in the Thomas (1993) data set are higher than those from other sources, commonly by several hundred psi, because they represent bottom-hole pressure rather than wellhead pressure. However, in older fields, such as Homer and Lancaster/Sugar Grove, pressures calculated by Thomas (1993) tend to be lower than those reported from other sources because they were based on wells drilled after the original pressures had been greatly reduced.

The overall trend of the plot is toward increasing underpressuring with depth (fig. 8), a trend also recognized by Thomas (1993). Moreover, data points on the plot are grouped into two clusters. One cluster consists of fields very close to being normally pressured and that define a trend with a gradient of approximately 0.35 to 0.40 psi/ft (fig. 8). All 5 fields in this cluster are from the discrete part of the regional accumulation and 4 of the 5 fields are located near the western margin of the accumulation where probable gas-water contacts occur. In contrast, the second cluster consists of fields that are markedly underpressured and that define a trend with a gradient of approximately 0.10 to 0.25 psi/ft (fig. 8). The second cluster is dominated by fields from the basin-centered part of the accumulation. Although overlap occurs, these two populations of pressure/depth gradients on figure 8 —and also shown by Thomas (1993)— may define useful criteria for differentiating discrete from basin-centered parts of the accumulation.

The origin of the underpressuring is unresolved. According to Thomas (1993), no single existing hypothesis is sufficient to explain all the observed pressure trends. Most promising of the hypotheses appear to be basin hydrodynamics, deep basin gas saturation, epeirogenic movement, and horizontal flow (Thomas, 1993). Post-orogenic uplift of the regional accumulation with subsequent cooling and slow leakage of the hydrocarbons after an earlier phase of overpressuring offers another plausible explanation of the regional underpressuring (Law and Dickinson, 1985; C. W. Spencer, oral communication, May 1997).

Although the Lower Silurian regional accumulation has been modified by uplift and erosion, bottom-hole temperature may discriminate between discrete and basin-centered accumulation because it identifies areas of peak gas generation that were most favorable to basin-centered accumulation. A plot of bottom-hole temperature (corrected) vs. depth to production for the data set approximates a straight line that defines a regional temperature gradient of about 1.3° F/100 ft (23.7° C/km) (fig. 9). This gradient is slightly lower than the recognized norm of about 1.65° F/100 ft (30° C/km) but it is in general agreement with geothermal gradients shown for the Appalachian basin (American Association of Petroleum Geologists and USGS, 1976). A threshold temperature of about 125° F (52° C) separates most of the fields associated with the discrete and basin-centered parts of the regional accumulation (fig. 9). Moreover, as suggested in figure 4, a vitrinite reflectance ~1.10 may be a plausible threshold value for discriminating discrete from basin-centered parts of the Lower Silurian regional accumulation. The 126° to 162° F (52° to 72° C) temperatures of fields in the basin-centered part of the accumulation (fig. 9) suggest significant cooling since the 180° to 200° F (82°-94° C) temperatures probably achieved during its active phase (Law and Dickinson, 1985; Spencer, 1989).

Ultimate Gas Recovery Per Well

Gas production expressed as estimated ultimate recovery (EUR) per well is available for 10 fields in the data set (table 5). These EURs are shown in table 5as median (F₅₀) or mean (´ ) values depending on how they are reported in the literature or on probability plots used in this investigation (fig. 10). In fields such as Yorktown/Clay (Appendix Y) where production records are available for just the best 1 or 2 wells, the EUR is expressed as the 5^th percentile (F₅) (table 5). EURs are most valuable where they are plotted as a probability distribution of at least 25 wells (fig. 10). However, only 5 of the 10 fields with mean or median EURs have been calculated from a probability plot; the remainder represent a "best guess" based on the experience of the reporting individual.

Judging from this preliminary data set, the fields from the basin-centered part of the accumulation have slightly greater median EURs than fields from the discrete part (fig. 11). Fields from the basin-centered part of the accumulation have median (or mean) EURs that range from 98 to 400 MMCFG per well whereas fields from the discrete part have median (or mean) EURs that range from 84 to 450 MMCFG per well. Far more production data are required from discrete and basin-centered parts of the accumulation before their EUR per well distributions are established. Also, additional production data are required to better understand the geologic causes of production sweet spots such as Ravenna (F₅₀=200 MMCFG) and Cooperstown (´ =400 MMCFG) fields (table 5, fig. 11), although it is probably significant that these fields adjoin fault zones and (or) surface lineaments.

Created by the EERT WWW Staff.