WATTENBERG FIELD, DENVER BASIN, COLORADO
Robert J. Weimer, Stephen A. Sonnenberg, and Genevieve B.C. Young (AAPG Studies in Geology 24, Geology of Tight Gas Reserviors, 1986, p. 143-164)
R. A. Matuszczak (Memoir 24, North American Oil and Gas Fields, 1976, p. 136-144)
Search and Discovery Article #20001 (1999)
Adapted for online presentation from articles, with this same title and by authors given above, published in the respective AAPG Publications noted above. The entire paper by Weimer et al. (1986) is presented here; a key illustration (Figure M4) from the Matuszczak paper is included.
In these papers attributes are given of a giant gas field in the deep part of a foredeep basin (Bally type 221), where basinal location and type of entrapment have in the past presented problems to conventional thought.
Similar settings have been described for accumulation, among others, in the Alberta basin (e.g., Masters, AAPG Memoir 38, Elworth -- Case Study of a Deep Basin Gas Field, p. viii-34) and in San Juan basin (e.g., Berry in Levorsen, Petroleum of Geology, Second Edition, 1967, p. 406). With this adaptation presented here, the question is raised again: Are most or all basins of this type characterized by this type of accumulation?
TABLE OF CONTENTS
The most important mineral resource activity in Colorado during the past decade has been the discovery and development of the Wattenberg and adjacent petroleum fields. Located north of Denver across the axis of the Denver basin, the Wattenberg is estimated to have reserves of 1.3 trillion cubic feet (tcf) in the tight J (Muddy) Sandstone (delta front) reservoir over an area of 600,000 acres at depths of 7,600 to 8,400 ft (2,310 to 2,560 m). Net pay thickness varies from 10 to 50 ft (3 to 15 m), porosity ranges from 8 to 12%, and permeability varies from 0.05 to 0.005 millidarcys (md) (Matuszczak, 1973, 1976). The preliminary prospect map (Figure M4) shows an area of no water recovery from the J Sandstone corresponding in a general way to Wattenberg field.
Drilling for J gas has resulted in multiple pays in overlying strata. The Spindle field, situated in the southwest portion of the Wattenberg field, produces from two marine-bar complexes (Hygiene and Terry) in the middle portion of the Pierre Shale. In 1981 and 1982, the Codell Sandstone, approximately 500 ft (152 m) stratigraphically above the J, was developed as a new producing horizon of oil and gas. More than 100 discoveries have been made within and marginal to the outlined Wattenberg field area. The Codell is a tight bioturbated marine-shelf sandstone generally without a central-bar facies. Net pay thickness ranges from 3 to 25 ft (0.9 to 7.6 m). Porosities determined from logs range from 8 to 24%, but the average core porosity is from 10 to 12% and permeabilities are less than 0.5 md. Because of rapid decline in production and economic uncertainties, potential reserves from the Codell are unknown. All petroleum accumulations in the Wattenberg area are regarded as stratigraphic traps, although unconformities and paleostructure have played a subtle but detectable role.
Variation in thickness and reservoir quality is related to original environmental facies and paleostructure that locally influenced unconformities, fracturing, and diagenesis.
The most important mineral activity in Colorado during the past decade has been the discovery and development of petroleum fields over an area of 1 million acres in the deeper portion of the Denver basin (Figure 1 and Figure 2; see animation of Figures 2, 5-12, 17-20, and 25). The activity first started in 1970 in the Wattenberg gas field with drilling to the J (Muddy) Sandstone at depths of 7,600 to 8,400 ft (2,300 to 2,500 m). As the development proceeded, oil and gas production was established at the Spindle field from the Terry and Hygiene sandstones of the Upper Cretaceous Pierre Shale at depths of 3,000 to 5,000 ft (915 to 1,524 m).
In 1981 and 1982, more than 100 petroleum discoveries were made in the Codell Sandstone within the producing areas of the Wattenberg field and also along the north and west margins (Figure 2; see animation of Figures 2, 5-12, 17-20, and 25). In early development at Wattenberg, the Codell was overlooked as a reservoir and bypassed in drilling and completion. The Codell may have significant hydrocarbon reserves, but the economics of the play are uncertain. It has been largely ignored because it is difficult to recognize on logs and is also a tight, low-permeability reservoir. Most production in the area is from stratigraphic traps, but subtle paleostructure and unconformities play an important role. Production straddles the synclinal axis of the Denver basin (Figure 2; see animation of Figures 2, 5-12, 17-20, and 25).
This paper discusses the geology of the low-permeability (tight) petroleum reservoirs of the J and Codell sandstones.
The stratigraphic section for the Cretaceous of the Denver basin is well known from outcrop, core, and electric log data (Figure 3). The lower 900 ft (275 m) of the Cretaceous stratigraphic section, drilled in the general Wattenberg area, contain mainly marine shale, siltstone, sandstone, and limestone. However, the Dakota group in the lower one-fourth of the interval is both marine and nonmarine in origin. Although outcrop and core data have assisted in stratigraphic studies, the main data base is from mechanical well logs. The electric log patterns for the mapable formations (or members) are illustrated by Figure 4, and the thickness and distribution of six units are illustrated by isopach maps (Figure 5 , Figure 6, Figure 7, Figure 8, Figure 9, Figure 10; see animation of Figures 2, 5-12, 17-20, and 25). More than 400 wells were analyzed in an area 60 miles (96.5 km) long and 40 miles (64 km) wide. Detailed thickness studies of these units were undertaken to determine if paleostructural movement influenced deposition and final distribution. Regional unconformities which influence thicknesses have been identified within the J Sandstone, the base of the Codell Sandstone, and the base and top of the Niobrara Formation.
The J Sandstone is the main petroleum-productive reservoir in the Denver basin. Over the Wattenberg field area, the J ranges in thickness from less than 75 to more than 150 ft (23 to 46 m) (Figure 5; see animation of Figures 2, 5-12, 17-20, and 25). Thickest sections are to the northeast and to the southwest of the Wattenberg.
Two types of sandstone bodies comprising the J Sandstone in the Denver basin were described by MacKenzie (1965) from outcrops along the west margin of the Denver basin. He formally named them the Fort Collins and Horsetooth members of the Muddy Sandstone, which is also called the J Sandstone in the subsurface (Figure 4).
The older Fort Collins Member is a very fine- to fine-grained sandstone containing numerous trace fossils and is interpreted to be delta-front, shoreline, and marine-bar sandstones that were deposited during rapid regression of the shoreline of the Skull Creek sea. Sandstones of the younger Horsetooth Member are fine to medium grained, well sorted, and cross stratified, and they contain carbonized wood fragments. The sandstones are intercalated with siltstone and shale and are interpreted to be channels deposited as part of valley-fill deposits. Valleys of an extensive drainage system were incised into the Fort Collins Sandstone, or the underlying marine Skull Creek Shale, and contain a fill varying in origin from alluvial plain to transitional deposits. The first description of valley-fill deposits and associated oil-productive sandstones in the Denver basin (Nebraska portion) was by Harms (1966).
The main productive sandstone in the Wattenberg field is from the Fort Collins Member (Figure 4), interpreted to be mainly delta-front sandstone. Surrounding the Wattenberg field are channel sandstone complexes, which are interpreted to be valley-fill deposits of the Horsetooth Member (Figure 4). Prevailing opinion has been that transport in the lower J channel sandstones was to the west-northwest (Haun, 1963; Matuszczak, 1976). Directional features from outcrop observations and a review of the channel incisement patterns indicate transport was to the southeast and east. Although these channel sandstones initially had greater porosities and permeabilities than the adjacent delta-front sandstones, diagenetic changes have occluded most of the porosity in some subsurface areas. The updip seal of the trap on the north and east flank of the Wattenberg field is thought to be by diagenetic changes or because impermeable freshwater shales of the valley-fill deposits rest on the surface of erosion between the Fort Collins and Horsetooth members.
From 400 to 500 ft (12 to 150 m) of dominantly black shale between the J Sandstone and the Niobrara Formation have been mapped in outcrop as the Benton Shale along the west flank of the basin. Based on characteristic log patterns, the Benton can be subdivided in the subsurface into four formations (Figure 4) which, in ascending order, are the Mowry Shale, the Granerous Shale, the Greenhorn Formation, and the Carlile Shale. Regional thickness patterns of these formations throughout the Denver basin were published by Weimer (1978). The Benton Group is thought to contain the source beds for the oil and gas found in Lower Cretaceous sandstones in the Denver basin (Clayton and Swetland, 1980).
One of the most distinctive Early Cretaceous formations in Wyoming and northernmost Colorado is the siliceous Mowry Shale. The unit thins southward into Colorado and pinches out just southeast of the Wattenberg field (Haun, 1963; Rojas, 1980). Across the field area the thickness varies from southeast to northwest from 6 to 25 ft (2 to 8 m) (Figure 6; see animation of Figures 2, 5-12, 17-20, and 25). The lithology is interbedded shale, siltstone, and very fine-grained sandstone in repetitive layers that are 1 to 2 in. (2.5 to 5 cm) thick. The thin sandstone beds commonly have sharp bases with underlying shales and grade upward to silt and then black shale. The combination of these lithologies gives a characteristic pattern on electric logs with a slightly higher resistivity than the overlying marine Graneros Shale. The black shale layers commonly have abundant fish scales.
In some outcrop sections a thin lenticular bed (< 1 ft; 0.3 m) of fine- to coarse-grained conglomeratic sandstone is observed at the top of the J Sandstone (MacKenzie, 1965). Although included in the J, the layer is a relict or palimpsest sandstone genetically related to the Mowry transgression (Rojas, 1980). This record of sediment reworking during the transgression of the sea over a surface on which either the Horsetooth or Fort Collins Member of the J was exposed, and the widespread thin nature of the Mowry, indicate rapid and uniform water deepening during the transgression. No significant structural movement occurred, although locally 20% thinning might suggest minor movement of fault blocks (Figure 6; see animation of Figures 2, 5-12, 17-20, and 25).
Gray bentonitic marine shale is the dominant lithology of the Graneros Shale. The unit shows a regional northward thickening (Weimer, 1978, p. 215) and across the Wattenberg area varies from less than 140 ft (43 m) along the south margin to 175 ft (54 m) along the northwest margin (Figure 7; see animation of Figures 2, 5-12, 17-20, and 25). Thin areas with thicknesses less than 150 ft (45.7 m) occur in T2S and 3N, R65-67W. Inasmuch as this minor thinning occurs in the same area where the upper Niobrara Limestone is truncated (Figure 11; see animation of Figures 2, 5-12, 17-20, and 25), paleostructural movement of basement fault blocks may have occurred during Graneros deposition.
An important stratigraphic unit within the Graneros in the Denver basin is the D Sandstone. This sandstone unit is present along the east side of the Wattenberg area (Figure 7; see animation of Figures 2, 5-12, 17-20, and 25) where it occurs 20 to 30 ft (6.1 to 9.1 m) above the base of the Graneros and varies in thickness from a wedge edge to 20 ft (6 m). To the northwest the D Sandstone changes to a siltstone (interpreted from logs) which occurs in a northeast-trending band 12 to 20 mi (19.3 to 32.2 km) wide (Figure 7; see animation of Figures 2, 5-12, 17-20, and 25). The siltstone changes to shale along the western margin of the area. In the Wattenberg area the D is interpreted as a marine sandstone, but to the east, petroleum production is found in fluvial channel sandstones. Where the D Sandstone is present, the lower part of the Graneros Shale is called the Huntsman Shale.
The Greenhorn Formation consists of thin limestones and dark-gray to black organic-rich shales. One of the most widespread units in the Great Plains and Rocky Mountain region, the Greenhorn is easily subdivided in the Wattenberg area into lower and upper intervals of limestone and calcareous shale and a middle interval of shale (Figure 4). Widespread marker beds of bentonite and thin limestone allow for easy correlation within the formation.
The contact with the Graneros is placed at the X bentonite; the contact with the overlying Carlile is placed at the top of the first limestone with a high resistivity reading on mechanical logs.
Like the Graneros, the Greenhorn has a regional northward and northwestward thickening across the Denver basin (Weimer, 1978, p. 216). In the Wattenberg field, the Greenhorn is 200 ft (60.9 m) in the south-central area and over 250 ft (76.2 m) in the northwest (Figure 8; see animation of Figures 2, 5-12, 17-20, and 25). A broad north-trending axis of thinning may reflect minor paleostructural movement during deposition.
The Carlile Formation in the Great Plains of Kansas and Colorado consists of four widespread members which, in ascending order, are the Fairport Chalk, the Blue Hill Shale, the Codell Sandstone, and the Juana Lopez Member. Based on regional correlations (Weimer, 1983, p. 14), the Blue Hill is not present in the Wattenberg area and the Juana Lopez is too thin (< 1 ft; 0.3 m) and lenticular to map. Therefore, the isopach map of the Carlile (Figure 9; see animation of Figures 2, 5-12, 17-20, and 25) reflects primarily the thickness of the lower Carlile (Fairport) and the upper Carlile (Codell Sandstone) (Weimer and Sonnenberg, 1983). The Carlile thins across the area from approximately 80 ft (24 m) in the southeast to < 25 ft (7.6 m) to the west. Two unconformities influence the Carlile thickness: one is at the base of the Codell Sandstone; the other is at the contact of the Codell with the overlying Fort Hays Member of the Niobrara. These unconformities will be discussed later.
The lower Carlile Shale in the Wattenberg area is a black, noncalcareous shale which, by log correlations, is equivalent to the lower Fairport. The Codell Sandstone is a gray, very-fine-grained bioturbated sandstone. It thickens westward across the area from a wedge edge to 28 ft (8.5 m). The lower Carlile shows a reverse pattern of westward thinning because of the unconformity at the base of the Codell. Both units are believed to be marine in origin.
The Niobrara Formation varies in thickness from 240 to 450 ft (73.2 to 137.2 m) (Figure 10; see animation of Figures 2, 5-12, 17-20, and 25). Isopach trends for the Niobrara are east-west in contrast to north-south for part of the Benton Group. Regionally, four limestone (chalk) intervals and three intervening shale intervals within the Niobrara were mapped in the subsurface (Figure 4). All units contain marine fossils and are thought to be relatively deep, neritic deposits. Of these seven mapable units, only the lower member has been given a formal name, the Fort Hays Member. The remaining six intervals are grouped together as the Smoky Hill Member. The four limestone units are easily recognized on mechanical logs from wells because of high resistivity and a poorly developed spontaneous-potential (SP) curve deflection. The limestone units (with minor shale beds) vary from a wedge edge to 80 ft (24.2 m) thick; the intervening shale units (with minor limestone beds) vary from 20 to 150 ft (6.1 to 45.7 m) thick.
An unconformity at the top of the Niborara is responsible for the east-west isopach trends, and the interpretation of its origin helps define the paleostructure of the Wattenberg area (Weimer, 1980).
STRUCTURAL AND PALEOSTRUCTURAL ANALYSIS
A structure contour map on top of the Codell Sandstone in the west-central portion of the Denver basin shows that J Sandstone and Codell production straddles the axis of the basin (Figure 2; see animation of Figures 2, 5-12, 17-20, and 25). The basin is asymmetric, with a gently dipping east flank and a steeply dipping west flank. The basin axis lies close to and parallels the Front Range. Laramide uplift of the Front Range along basement faults has dominated the overall structural style. Small drape folds over basement faults form sharp anticlines in the Loveland, Berthoud, and Boulder fields (Figure 2; see animation of Figures 2, 5-12, 17-20, and 25). Minor changes in regional strike or dip within the Denver basin result principally from small-scale subtle warping of strata over basement fault blocks. Fold trends are northwest and east-northeast.
Although presently in a structural low, the Wattenberg area was a structural high at the start of deposition of the Pierre Shale. A key to understanding the paleostructure and configuration of basement fault blocks is to analyze the origin of the thinning in the Niobrara Formation (Figure 10; see animation of Figures 2, 5-12, 17-20, and 25).
A key unit with which to interpret paleostructure is the upper limestone of the Niobrara Formation (Figure 11; see animation of Figures 2, 5-12, 17-20, and 25). The isopach map (Weimer, 1980) shows the unit varies from a wedge edge to 80 ft thick (24.2 m) and is absent over an east-west-trending area that is 50 miles (80.5 km) long and 10 miles (16.1 km) wide. Outcrop sections north of Boulder show the upper limestone is present, although perhaps thinner than normal. Based on micropaleontological work, LeRoy and Schieltz (1958) reported an unconformity at the Niobrara-Pierre contact in the Dagg Mesa section (4 mi (6.4 km) north of Boulder), but the amount of missing Niobrara was not determined.
The diagnostic mechanical log patterns from closely spaced wells clearly show the upper limestone and the upper part of the underlying shale of the Niobrara to be truncated because of erosion in the central part of the Wattenberg field (Figure 4 and Figure 11; see animation of Figures 2, 5-12, 17-20, and 25).
An isopach map from the top of the J Sandstone to the base of the Pierre Shale ( Figure 12; see animation of Figures 2, 5-12, 17-20, and 25) indicates a thin area of less than 750 ft (228 m), thickening southward to more than 900 ft (274 m). The greatest erosion is over the top of what is interpreted as an east-west-trending anticline (horst block). After the Niobrara was deposited, structural movement occurred during the time of early Pierre deposition, resulting in an uplifted shoal area on the sea floor. This area was subjected to beveling by marine processes or possibly by subaerial processes, although no evidence for subaerial exposure was found. The erosion may have occurred during a regional drop in sea level, with the top of the paleostructure subjected to more intense wave or current action than the flank areas. After the period of erosion, marine shales of the lower Pierre were deposited.
If the lower Pierre Shale was deposited in a nearly horizontal attitude ( Figure 4), then the isopach map can be used as a paleostructure contour map (Figure 12; see animation of Figures 2, 5-12, 17-20, and 25). The isopach map shows greatest thinning in an east-west-trending area 10 miles (16.1 km) wide and 50 miles (80.5 km) long. This paleostructure probably persisted during the deposition of several thousand feet of the lower half of the Pierre Shale. Natural gas that was free to migrate in the J Sandstone during this sequence of events accumulated in a combined structural and stratigraphic trap. This trap was subsequently warped into the present low structural position at the bottom of the Denver basin.
Uncertainty exists as to whether the Wattenberg gas is thermogenic or a mixture of both thermogenic and biogenic. Because of above-normal heat flow and isotopic composition in the Wattenberg area, Momper (1981) and Rice and Threlkeld (1982, 1983) believe the gas is thermogenic. Clayton and Swetland (1980, p. 1631) state that "methane gases from the Wattenberg field near Denver are isotopically too light to have formed by late-mature thermal cracking." If significant quantities of the gas are biogenic and, therefore, of early origin, then the early paleostructure played an important role in the accumulations.
The relationship of the Wattenberg paleohigh to other paleostructural elements in the Denver basin was described by Sonnenberg and Weimer (1981). Comparison of regional isopachs of the J Sandstone in the Denver basin show that in general the J Sandstone is thin in areas where the Niobrara is thin. In addition, valley-fill deposits of the J Sandstone are best developed in thick areas between paleostructural high areas. Also, extensional fracturing associated with paleodrape folds along the margin of the basement blocks probably developed a natural open fracture system in the J Sandstone. The trends would be east-west and northwest and northeast.
LITHOLOGIES OF RESERVOIRS
Both the Fort Collins and Horsetooth members of the J Sandstone produce gas in the Wattenberg field, but most of the production is from the Fort Collins Member. The two members are similar texturally and compositionally but appear to have slightly different diagenetic histories.
The Fort Collins Member is a very-fine-to fine-grained sandstone composed of approximately 80% quartz, 10% argillaceous matrix, 5% rock fragments, and 5% feldspar. The sandstones are subrounded to rounded and poorly to well sorted. The main matrix material of clay is interpreted to result from bioturbation which, in a complete section, decreases vertically in abundance. The decrease in matrix is due to higher energy conditions, which tend to winnow out the clay fraction, and also to a change from deposit-feeding to suspension-feeding organisms. The lithologies of outcrop sections described by Clark (1978) and Suryanto (1979) west of the Wattenberg field are similar to cores from producing wells.
The porosity is mainly intergranular with minor amounts of microporosity found in the matrix material and secondary interparticle porosity associated with leached feldspars and lithic rock fragments. The order of diagenetic processes in the sandstone is as follows: compaction, growth of clay rims (chlorite), formation of quartz overgrowths, calcite cementation and contemporaneous feldspar and lithic rock fragment solution, subsequent calcite solution, growth of illite-smectite, and late-stage fracturing. Many of these processes were ongoing throughout the burial history of the sandstone. The main diagenetic event in the Fort Collins Member is the formation of quartz overgrowths (Figure 14 and Figure 15). The euhedral overgrowths are separated from the detrital quartz grains by a clay rim and show crystal faces directed into pores which generally results in triangular pore shapes. The percentage of overgrowths appears to diminish as matrix content increases.
The Horsetooth Member is a fine- to medium-grained well-sorted, well-rounded sandstone composed of 75 to 80% quartz, 5 to 10% rock fragments, and 5% feldspar. The porosity is mainly intergranular with minor microporosity and intraparticle porosity. The diagenetic processes affecting the Horsetooth Member are similar to the Fort Collins Member, with the exceptions of more intense silica cementation and the presence of a late-stage kaolinite clay cement. The combination of silica overgrowths and kaolinite cement can totally occlude primary porosity. These cements in the Horsetooth Member may help trap the gas in the Horsetooth Member and, therefore, appear to be the seal on the eastern side of the Wattenberg field.
The Codell Sandstone in the Wattenberg area is a bioturbated, silty, shaly, very-fine-grained sandstone. The sandstone is poorly sorted and compositionally immature. The overall lithology of outcrop sections described by Lowman (1977) is similar to lithology observed in cores from producing wells. The average composition (bulk X-ray analysis) of the sandstone in the Hamilton Brothers Pratt 1-30 (sec. 30, T4N, R68W) is as follows: quartz, 45%; feldspar, 6.8%; matrix, 40.2% calcite, 4.0%; pyrite, 3.4%; and siderite, 0.6%. The matrix composition consists of illite, 49%; chlorite, 21%; quartz, 15%; feldspar, 5%; pyrite, 7%; and calcite, 3% (analyses provided by Cities Service Co.). Most of the clays are interpreted to be detrital in origin, with clay mixed with sand by the process of bioturbation. SEM photomicrographs show authigenic illite lines and bridges pores. Calcite, pyrite, and siderite are also authigenic.
Because of the high clay matrix within the Codell, the sandstones probably had low initial porosity and permeability, which has been further reduced by compaction and diagenesis. Tensional fracturing associated with structural movement undoubtedly enhances reservoir performance.
The J Sandstone reservoir in the Wattenberg field is productive over an area of 600,000 acres at depths of 7,600 to 8,400 ft (2,310 to 2,560 m). Reservoir parameters were summarized by Matuszczak (1973) as follows: net pay thickness varies from 10 to 50 ft (3 to 15 m); porosity ranges from 8 to 12%; permeability ranges from 0.05 to 0.005 md; reservoir pressure and temperature average 2,900 psig and 260°F, respectively; average water saturation is 44%.
Over the Wattenberg field, the J Sandstone is generally thin and ranges in thickness from approximately 75 to 150 ft (22.8 to 45.7 m) (Figure 5; see animation of Figures 2, 5-12, 17-20, and 25). Most of the gas production comes from the Fort Collins Member (delta-front sandstone, Figure 13 and Figure 16), which ranges in thickness from a wedge edge to greater than 80 ft (24.4 m). However, the main pay section in the upper Fort Collins averages between 10 and 20 ft (3 and 6 m). Thin areas, or areas where the Fort Collins Member is absent, are caused by erosion during an incisement of drainage prior to deposition of valley-fill deposits of the Horsetooth Member (Figure 16, Figure 17, Figure 18; see animation of Figures 2, 5-12, 17-20, and 25). The Horsetooth ranges in thickness from less than 20 ft (6.1 m) to greater than 140 ft (42.6 m). Thick valley-fill deposits with channel sandstone complexes occur on the northeast and southwest portions of the mapped area.
Because of the widespread distribution of the marine Fort Collins Member, the log character of the J Sandstone appears to be fairly uniform across the Wattenberg field. A normal log suite consists of resistivity, SP, gamma-ray, neutron, and density logs; typical log patterns are illustrated from the Amoco Rocky Mtn Fuel 1 (Figure 13). Through the Fort Collins Member, the gamma-ray and SP curves have a funnel shape, indicating an overall textural coarsening-upward sequence (more sand and less shale). The log response is typical for delta-front sandstones in the Wattenberg field. The resistivity curve is almost a mirror image of the gamma-ray and SP curves. The highest resistivities occur in the upper portion of the delta-front sandstones and are caused by gas being present in the reservoir. Porosity values from the neutron and density logs through the upper delta front are 3% and 6 to 10%, respectively. The neutron-density crossover is caused by a gas effect.
In the Rocky Mtn Fuel well (Figure 13 and Figure 16), the Horsetooth Member consists of alternating bioturbated shaly sandstone and laminated sandstone and shale. The bioturbated interval is interpreted as marine-bay deposits that formed during the Mowry transgression related to sea-level rise. The low SP, high gamma-ray, and low resistivity readings reflect the abundance of clay in the sandstone. In other areas of the field, cleaner sandstones are present in the Horsetooth Member, and the log response changes accordingly ( Figure 16).
Gas produced from the J Sandstone is rated as 1,150 Btu/ft3 (Momper, 1981). A typical gas analysis from a well in the northern portion of the Wattenberg field consists of the following (Amoco Esther Gaumer 1, sec. 21, T3N, R66W): nitrogen, 0.43%; carbon dioxide, 3.42%; methane, 77.64%; ethane, 10.66%; propane, 3.31%; isobutane, 00.57%; n-butane, 0.95%; isopentane, 0.43%; n-pentane, 0.26%; hexanes, 0.56%; heptanes plus, 1.75%.
Gas-oil ratio is 38,481 mcf/bbl separator oil (D. Perez, Amoco, personal communication, 1983). The gas is considered to be thermogenic in origin (Rice and Threlkeld, 1982). Higher-than-normal temperature gradients have been mapped by Meyer and McGee (1985), and areas of best production are related by them to a large "hot spot" anomaly.
Cumulative gas production from wells in the Wattenberg field is generally less than 500 mmcf per well (Figure 19; see animation of Figures 2, 5-12, 17-20, and 25). Although production is erratic, the major production is from the west-central portion of the field from the Fort Collins Member of the J Sandstone. In only a few small areas have wells produced more than 1 billion cubic feet (bcf). If the field development remains on 320-acre spacing, the cumulative production map suggests that ultimate production will not reach 1 trillion cubic feet (tcf) because only a small percentage of the 1,000 wells will produce 1 bcf. Published reserve figures of 1.3 tcf (Matuszczak, 1973, 1976) may have to be revised. Cumulative production of oil associated with gas also shows an erratic distribution ( Figure 20; see animation of Figures 2, 5-12, 17-20, and 25). Main areas of oil production are in the central and southeast portions of the field. Cumulative production per well is generally less than 5,000 bbl, although several small areas show production in excess of 10,000 bbl per well.
The Codell Sandstone reservoir, approximately 400 ft (122 m) stratigraphically above the J Sandstone, has been developed as a new petroleum-producing horizon in the Wattenberg field area (Weimer and Sonnenberg, 1983). Until recently, the Codell was overlooked as a potential reservoir for oil and gas and was bypassed during drilling and well completion in the J Sandstone. The Codell may contain significant reserves of hydrocarbons, but the present economics of the play is highly debatable. It has generally been ignored because the sandstone is difficult to identify on electric logs and is almost impermeable (Figure 21 and Figure 22). Nevertheless, new well completions in a broad area (possibly 1 million acres) allow petroleum geologists to do frontier exploration in a mature basin. The technology frontier is how to extract commercial quantities of petroleum from a widespread tight reservoir.
A structure contour map of the top of the Codell Sandstone in the west-central portion of the Denver basin shows that Codell production straddles the axis of the basin (Figure 2; see animation of Figures 2, 5-12, 17-20, and 25). Currently, stratigraphically entrapped hydrocarbons in the Codell have been found in more than 100 discovery wells. Production depths range from 4,000 to 8,000 ft (1,219 to 2,438 m). Porosities indicated by density logs range from 8 to 24%. Porosities indicated by core analyses range from 8 to 10%. Permeabilities from core analyses are generally less than 0.5 md. The average thickness of reservoirs in the area is between 14 and 16 ft (4.3 and 4.9 m). At about 8,000 ft (2,438 m) the reservoir pressure and temperature range from 3,000 to 3,500 psig and 200 to 240°F, respectively (pressure information courtesy Energy Oil). Initial potential production volumes from the new discoveries range from 200,000 to 1.1 mmcf per day and 11 to 300 bbl of oil or condensate per day.
Oil discoveries are most abundant on the west flank of the basin, whereas gas condensate discoveries are dominant on the east flank of the basin. Some operators believe the gas condensate discoveries may be in retrograde condensate reservoirs. A retrograde condensate reservoir contains hydrocarbons in a single phase (gas) until the reservoir pressure declines to the dew point, at which time liquid condenses from the reservoir fluid. This can seriously affect production by reducing relative permeability.
The Codell is designated as a tight gas sand by the Federal Energy Regulatory Commission (Figure 2; see animation of Figures 2, 5-12, 17-20, and 25). Prices for the gas from tight gas sands were a major incentive for exploring the Codell.
The earliest production from the Codell was from a fractured reservoir in the Boulder field. In the mid-1970s, Byron Oil developed the southern part of the Spindle field and commingled production from the Terry, Niobrara, and Codell. Recent exploration of the Codell began in mid-1979 with discoveries in the Boulder Valley field by Martin Oil. The next major discovery was by Energy Oil in mid-1981 in the Hambert field area. Since that time, the area of exploration has been greatly enlarged.
No general spacing rules exist for Codell fields; however, operators generally offset oil discoveries on 80-acre spacings and gas discoveries on 160-acre spacings.
Large portions of the exploration area for the Codell are within the Wattenberg and Spindle fields, which produce hydrocarbons from the Terry, Hygiene, and J sandstones. Several wells which produced from the J have been recompleted in the Codell Sandstone.
Average drilling costs during 1981 and 1982 in the west-central portion of the Denver basin were $100,000 for a dry hole, $300,000 for a completed well, and $80,000 for recompleting a J Sandstone well to produce from the Codell. Costs for acreage in the area of Codell production have risen sharply. Acreage formerly leased for $30 to $50 per acre bonus now leases for more than $100 per acre bonus.
Across the Wattenberg area, the Codell unconformably overlies the Greenhorn or Carlile shale and is unconformably overlain by the Niobrara Formation. The Codell ranges in thickness from a wedge edge to approximately 30 ft (9.1 m) thick (Figure 21 and Figure 23; see animation of Figures 23 and 24). The thickest Codell occurs in the west portion of the Wattenberg area. The absence of Codell Sandstone in the southeast portion of the mapped area is probably because of erosional truncation, although absence by facies change is possible. Because of the unconformity at the base of the Codell, the lower Carlile thickens eastward from less than 20 ft (6 m) on the west flank of the Denver basin to a maximum of 250 ft (76 m) (Figure 21 and Figure 24; see animation of Figures 23 and 24).
The Fort Hays Member of the Niobrara Formation ranges in thickness from less than 20 ft (6.1 m) in the northwest portion of the Wattenberg area to greater than 30 ft (9.1 m) in the southwest portion of the mapped areas (Figure 25; see animation of Figures 2, 5-12, 17-20, and 25). Thinning to the northwest may be because of onlap onto the erosional surface at the base of the Fort Hays in this area. Fracturing within the Niobrara Formation probably enhances production in the Codell Sandstone. Where fracturing is important to reservoir quality, the Codell and overlying Fort Hays Limestone may be a commingled reservoir.
The electric log character of the Codell Sandstone appears to be fairly uniform in bore holes in the exploration area. A common log suite consists of resistivity, SP, gamma-ray, neutron, and density logs; and typical log patterns are shown for the Dome Frank 1-13 well (Figure 22). The gamma-ray and SP curves change little through the Codell and underlying Carlile Shale. The deep-induction resistivity reading for the Codell is 6 to 8 ohm-m, which is only slightly higher than the reading for the underlying shale. Porosity values from neutron and density logs of the Codell are 18 to 20% and 12 to 18%, respectively. Core analyses of the Codell from the Dome well indicates the porosity is only 10%. The core consists of silty, clayey, bioturbated, very-fine-grained sandstone. The low SP, high gamma-ray, and low resistivity readings reflect the abundance of clay in the sandstone. The neutron log is affected by the clay and slightly affected by the gas in the sandstone. The higher density log values for porosity are caused by the gas.
Gas produced from the Codell Sandstone is rated at 1,185 Btu/ft3 and is thermogenic in origin (Rice, 1983). A typical gas analysis from the Hambert field area (T4N, R65W) consists of the following: helium, 0.01%; hydrogen, trace; carbon dioxide, 2.63%; nitrogen, 0.85%; methane, 77.11%; ethane, 12.92%; propane, 4.06%; isobutane, 0.57%; n-butane, 0.96%; isopentane, 0.31%; n-pentane, 0.26%; and hexane, 0.32% (composition courtesy Energy Oil).
GEOLOGIC MODELS FOR ORIGIN AND DISTRIBUTION RESERVOIRS
Unconformities related to sea-level changes, and either regional or local tectonics, have significantly influenced the origin and distribution of reservoir rocks. Over large areas, the upper, more porous portions of the J and Codell reservoirs were removed by erosion prior to deposition of overlying strata. Better petroleum production is found where more complete stratigraphic sections remain.
Origin of J Sandstone Reservoir
At the end of Skull Creek deposition (T1, Figure 26; see animation of Figures 26, 27 and 29), a regressive event began which deposited shoreline and shallow marine sandstones with a transitional contact with underlying Skull Creek shales. Depositional patterns over basement fault blocks, where slight fault block movement influences sedimentation, depends on the depositional environment (Figure 26; see animation of Figures 26, 27 and 29). Rivers and associated deltas positioned themselves in structural and topographically low areas (i.e., grabens), whereas delta-margin or interdeltaic sedimentation occurred along an embayed coast over structural horst blocks. Delta-front and shoreface sands extended seaward from the shoreline a distance controlled by effective wave base. The shoreline prograded seaward to position T2, and a sheet-like sand body was deposited over a large area (Fort Collins Member--pay sandstone of Wattenberg). These depositional patterns developed during a high sea-level stand.
A drop in sea level occurred (T3), during which all or a large portion of the depositional basin (Skull Creek seaway) was drained. River drainages were incised into older strata, especially in topographic lows that correspond with the graben fault block areas (Figure 27; see animation of Figures 26, 27 and 29). Over much of the Denver basin, the base of the incisement is on the T1 or T2 sand complex. Locally the erosional surface cut into the Skull Creek Shale. This drop in sea level is related to the worldwide low sea level reported by Vail et al. (1977) as occurring approximately 97 million years before the present. The geographic distribution of the major incised valleys during the low stand is shown on an isopach map of the J Sandstone (Figure 28, modified after Haun, 1963; Matuszczak, 1976).
A rise in sea level occurred (T4, Figure 29; see animation of Figures 26, 27 and 29), during which the incised valleys were probably modified and filled with fluvial and estuarine sandstone, siltstone, and shales. With a continued rise in sea level and minor renewed fault-block movement, strata were eroded from the top of the horst blocks and a widespread but lenticular thin transgressive-lag deposit of conglomeratic or coarse-grained sandstone formed over the horst blocks on a surface of unconformity. Following T4 the entire region received marine siltstone and shale deposition (Mowry or Graneros shales).
An important unconformity separates T1 and T2 deposits from T4 deposits. The basin-wide surface of erosion (T3, Figure 27; see animation of Figures 26, 27 and 29) may be within sandstone deposits in the basin (i.e., valley-fill sandstones rest on older regressive sandstones) or between sandstone and marine shale deposits (i.e., valley-fill sandstones rest on Skull Creek Shale). Another case is where the unconformity is at the base of the Mowry Shale or top of the regressive sandstones (e.g., portions of the Wattenberg field).
Previous correlations that show the J Sandstone to be deposited throughout the basin during one major regressive event are in error. Sandstones above the unconformity (Horsetooth Member) are younger than the regressive sandstone at the top of the Skull Creek (Fort Collins Member), although, because of tectonic movement, the older sandstones are now at a stratigraphic high position (Figure 4, Figure 29; see animation of Figures 26, 27 and 29). Therefore, the two types of sandstones cannot have a facies relationship. Recognizing the basin-wide unconformity associated with the J Sandstone, especially in relation to paleostructure, has significance in future petroleum exploration in the Denver basin.
The main gas reservoir at Wattenberg is the Fort Collins Sandstone beneath the unconformity. However, in the northeast and east portions of the field, gas is produced from sandstones within the younger valley-fill deposits of the Horsetooth Member. When productive, these lenticular sandstones are thought to have fluid continuity with the Fort Collins Member across the wall of the ancient valley.
Origin of Codell Sandstone Reservoir
Thickness and facies patterns suggest that a model for marine shelf, slope, and basin sedimentation can be reconstructed for the Greenhorn and Carlile (Figure 30). The origin and distribution of the Codell Sandstone is related to a combination of tectonics, shelf sedimentation and sea-level changes. A broad structural doming in central Wyoming (Weimer, 1983) during the sea-level drop at about 90 million years (m.y.) resulted in erosion of marine lower Carlile strata (Figure 21 and Figure 30). Chert and phosphate pebbles in thin coarse-grained sand occur as a lenticular lag on the erosional surface. Codell sand was deposited as marine or shoreline sand in the southeastern Denver basin during the low stand of sea level (C, Figure 30). With the subsequent rise in sea level, the Codell sand was deposited in brackish or marine environments in scour depressions in the northern Denver basin and as marine-shelf shaly sand and marine bars in the Wattenberg area (D, Figure 27; see animation of Figures 26, 27 and 29). As sea level continued to rise, the Juana Lopez was deposited as a thin widespread shelf palimpsest deposit of sand and shell fragments, mainly in the southern basin area, and the Sage Breaks Shale was deposited over the Codell Sandstone in the northern Denver basin (Weimer and Sonnenberg, 1983). The original geographic extent of the Sage Breaks is unknown.
A second drop in sea level between 89 and 89.5 m.y. in conjunction with regional doming in northern Colorado resulted in erosion from the area of the Wattenberg field area of the Sage Breaks Shale, Juana Lopez Limestone, and the upper part of the Codell Sandstone. The Niobrara Formation was deposited on the erosional surface during the subsequent sea-level rise.
The unconformity at the top of the Carlile Formation places the upper portion of the Fort Hays Member of the Niobrara Formation on the Codell Sandstone, or the lower Carlile interval (D, Figure 30). The unconformity has been interpreted by Weimer (1978) to result from erosion followed by marine onlap on a broad northeast-trending structural element called the Transcontinental arch. Sparse faunal evidence suggests that the hiatus corresponds to the time span represented by 5 or 6 faunal zones.
This unconformity has an important control on the distribution of sandstone in the Codell. Regional studies suggest that, in the Wattenberg field area, post-Codell erosion removed strata estimated to range from 30 to 100 ft (9 to 30 m) in thickness. Because of thin lags of coarse sand and chert pebbles, the erosion is interpreted to have removed the marine central bar facies, the higher-energy deposits that may have contained porous and permeable sand. Remnants of this sand facies represent exploration targets for better production from the Codell.
Both the J and Codell sandstones are tight sands productive from a large area that lies across the axis of the Denver basin. No water is produced from either reservoir in the structural low areas of the field. The present traps appear to be largely stratigraphic-diagenetic, although subtle structure probably played a role that was dependent on the time of entrapment and the nature of migration.
The Wattenberg accumulations belong to the class of deep basin traps described by Masters (1979) and Gies (1981). These traps have no downdip water and may or may not have updip sandstone equivalents that produce water. Trapping is because of subtle permeability variations caused by diagenesis, by a lack of buoyancy forces so that the gas stays in static equilibrium, or by subtle permeability changes caused by facies or unconformities. All of these factors may contribute to entrapment in the Wattenberg field.
Published figures for estimated ultimate recovery for petroleum in the J Sandstone reservoir at the Wattenberg field are 1.3 tcf gas and 30 million bbl of condensate (Matuszczak, 1976; Momper, 1981). Cumulative production to January 1982 for 895 wells was 252 bcf gas and 2.6 million bbl of condensate (Oil and Gas Statistics of Colorado, 1981) (Figure 31, Figure 32, Figure 33). The most prolific areas of the Wattenberg field are T1-2N, R66-67W and T2S, R64-65W (Figure 19 and Figure 20; see animation of Figures 2, 5-12, 17-20, and 25). The gas wells produce at low rates and have a life of 20 years or more. A typical production decline curve from the west portion of the field is illustrated by the Amoco Rocky Mtn Fuel 1 (sec. 8, T1N, R67W) (Figure 33). Production declines at a rate of 50%/yr for the first 1 or 2 years and then declines at a rate of 10 to 20%/yr through the life of the well. Based on production history and projection into the future, figures for ultimate cumulative production cited above appear to be too high by a factor of 3 for gas and 5 for oil (Figure 19 and Figure 20; see animation of Figures 2, 5-12, 17-20, and 25).
Peak production for the J Sandstone reservoir occurred in 1976 when 38.7 mmcf of gas and 400,000 bbl of condensate were produced (Figure 31, Figure 32). Annual production totals have declined slightly since that time in spite of the continued development drilling. Development well drilling was greatest in 1975 when 235 new gas wells went on production. Amoco Production is the major operator in the field.
Exploration and development of the Codell has occurred since about 1981; consequently, experimentation in completion techniques continues. Several wells have been producing for over 1 year, and production data are available. Typical production decline curves are illustrated by the Energy Oil Grant Arens 1 well (Figure 34) on the gas-prone eastern flank of the basin and the Machii-Ross Barclay Crisman 1 well (Figure 35) on the oil-prone western flank of the basin.
The Energy Oil Grant Arens 1 well (sec. 22, T4N, R65W) was perforated at depths of 7,094 to 7,108 ft (2,163 to 2,167 m) and had an initial potential flow of 750 MCFGD and 80 BOPD. A maximum monthly production of 597 bbl of oil and 9.4 mmcf of gas was recorded in November 1981 (Figure 34). After one year, production has declined 80% for oil and 68% for gas. Cumulative production through June 1983 was 4,755 bbl of oil and 83,718 mcf of gas. If the production decline rate remains constant, the estimated ultimate recovery from the well will be 6,000 bbl of oil and 86,000 mcf of gas. However, the rate of decline has decreased, which suggests the ultimate recovery will be larger.
The Machii-Ross Barclay Crisman 1 well (sec. 20, T3N, R66W) in the Wattenberg field was recompleted from the J Sandstone (Figure 35). The Codell was perforated between depths of 7,390 and 7,404 ft (2,253 to 2,257.3 m) and flowed 64 bbl of oil and 270 mcf of gas per day. A maximum production of 1,269 bbl of oil and 6.7 mcf of gas was achieved in November 1981. In one year, both oil and gas production declined 71%. Cumulative production through June 1983 was 10,039 bbl of oil and 56,841 mcf of gas. If the production decline rate remains constant, the estimated ultimate recovery is 11,000 bbl of oil and 67,000 mcf of gas. However, there has been a decrease in the rate of decline, which suggests the ultimate recovery will be larger.
These two examples indicate the ultimate reserves for each Codell well may be very small; but because of the large potentially productive area, total reserves are significant. Even with high gas prices, many of the producing Codell wells appear to be presently noncommercial or marginally commercial. In terms of economics, the recompleted J Sandstone wells are the most promising because the cost of recompleting a well is much less than the cost of drilling a new one.
Larger reserves per well should be found in either fractured reservoirs associated with structural anomalies or in the central-bar facies of marine-shelf sandstones. Wells currently in the exploration area apparently have not penetrated a marine central-bar facies. This facies would reflect a higher-energy depositional environment, and sandstones should be better sorted with better reservoir characteristics.
The best exploration procedure would be to establish production in the J Sandstone and eventually plug the well back to the Codell, or to devise a program for dual completion. In the Wattenberg field, the Codell Sandstone contains considerable reserves of hydrocarbons, and many of the wells now producing from the J Sandstone probably will be recompleted in the Codell. If 1,000 wells in the Wattenberg field were also productive from the Codell, and if the cumulative production from each well averages 10,000 bbl of oil and 50,000 mcf of gas, then ultimate reserves could be calculated as 10 million bbl of oil and 50 bcf of gas.
DRILLING AND COMPLETION TECHNIQUES
A typical drilling program for a J well in the Wattenberg field consists of the following: (1) drill a 12 1/4-in. (31.1-cm) surface hole and set 8 5/8-in. (21.9-cm) casing through the Fox Hills aquifer; (2) drill a 7 7/8-in. (20-cm) hole to total depth; (3) log the well and set 4 1/2-in. (11.4-cm) casing to total depth; (4) perforate casing through the J interval and stimulate (Smith et al., 1976). The wells are drilled with water-based mud. Operators increase mud weight (i.e., mud up) a few hundred feet (30.5 m) above the J Sandstone to condition the hole for logging. While drilling the J Sandstone, the water loss is maintained at 8 cm3 or less to minimize damage to the reservoir.
The Wattenberg field has commercial production because massive hydraulic fracturing is used to increase permeability around each well bore. Amoco Production Co., the main operator in the field, used laboratory experiments to determine fluid characteristics and proppants applicable to the J Sandstone reservoir (Fast et al., 1977). Low permeability and high reservoir temperatures presented problems for conventional fracture stimulation technology. To determine the optimum fracturing treatment, areas of the field were designated as experimental. The results of the field experiments showed that in certain areas of the Wattenberg field, larger treatments were more economically desirable. Other areas of the filed appeared economically unattractive regardless of treatment size (Fast et al., 1977). The standard Wattenberg fracture stimulation job currently consists of 180,000 gal (681,374 l) of cross-linked gelled water with 832,000 lb (377,388 kg) of 20/40-mesh sand used as a proppant (D. Perez, Amoco, personal communication). Artificially generated fractures are estimated to penetrate 3,000 ft (914 m) into the J Sandstone reservoir (Fast et al., 1977).
Codell wells are generally drilled with water-based mud that contains bentonite for additional density. Prior to reaching the Niobrara Formation, operators increase the mud weight to ±9.0 lb/gal (1.07 g/cm3) and try to maintain a water loss rate of 8 cm3. While drilling the Codell, mud weights are kept between 9.1 and 9.5 lb/gal (1.08 and 1.13 g/cm), depending on pressures encountered, and the water loss rate is maintained at 6 cm3 or less. The low rate of water loss minimizes damage of the Codell reservoir since the Codell Sandstone contains clays (e.g., montmorillonite) that swell and migrate in the presence of fresh water.
After drilling is completed, logs are run and casing is usually set at total depth. The casing is then perforated, tubing is set, and the Codell is stimulated. Typical stimulation consists of a small acid treatment (1,000 to 2,000 gal; 3,785 to 7,570 l) followed by hydraulic fracturing. An average fracture stimulation, which costs approximately $55,000, consists of 110,000 gal (416,395 l) of gel water (KCl base); 15,000 lb (6,803 kg) of 100-mesh sand; and 150,000 lb (68,038 kg) of 20/40-mesh sand. The lower part of the Codell is perforated in an attempt to keep the fracturing fluids from entering the Niobrara Formation. Hydraulic fracturing is done through tubing to avoid killing the well after the fracturing and because tubing strength is much greater than casing strength.