EXPLORATION North Dakota's Dickinson Lodgepole discovery: a preliminary exploration model

Julie A. LeFever
North Dakota Geological Survey
Grand Forks, N.D.

S.P. Halabura
North Rim Exploration Ltd.
Saskatoon, Sask.

David W. Fischer
Independent geologist
Grand Forks, N.D.

Carol D. Martiniuk
Manitoba Energy and Mines
Winnipeg

Interest in the Mississippian Lodgepole formation of North Dakota has intensified since the successful completion of the Duncan Oil Inc. 1-11 Knopik flowing 2,707 b/d of oil and 1.55 MMcfd of gas (430 cu m of oil and 43,891 cu m of gas. This is the first successful wildcat since Conoco Inc. initiated the play in February 1993.

The play began when Conoco drilled an in-field wildcat in an attempt to establish deeper production in Dickinson oil field (Fig. 1)(19604 bytes). The discovery well, 74 Dickinson State, was completed in a "clean" lower Lodgepole limestone section that is thought to represent a "Waulsortian mound" (Fig. 2,(27793 bytes)Tables 1(7643 bytes), 2(19646 bytes). Until second quarter 1995, the activity associated with the Lodgepole "mound" play has been primarily that of a land acquisition and seismic play. The success of the 1-11 Knopik wildcat has increased the interest and the number of companies involved in the play.

The most important questions asked concerning the Lodgepole play are whether or not it will step out of the Dickinson area, what are the factors that control the development of these "mounds," what controlled the development of the reservoir and trap, and how it was charged with oil. Other than the reservoir section, the most significant feature observed from wireline logs of the area is the anomalously thick Bakken formation (Mississippian-Devonian). This observation is important to understanding the Lodgepole play and can be used to help explore for similar features elsewhere in the basin.

Production from the Lodgepole formation in the North Dakota portion of the Williston basin has been restricted to a few scattered wells prior to Conoco's discovery in 1993.1 Before 1993, operators in the state considered the Lodgepole to be essentially nonproductive. Examination of the Lodgepole in Canada shows that this is not the case, where the formation produces not only from shelf carbonates but also from reservoirs that may be stratigraphically equivalent to those in Dickinson field (Fig. 3)(31866 bytes).

The first commercial oil production from the Williston basin came from the Lodgepole formation in Daly field, Manitoba. The California Standard 15-18 Daly, in 15-18-10-27wPm, was completed in early 1951 as a marginal producer.2 Currently 319 wells are producing from the Lodgepole in Daly field. Cumulative production is 30,942,818 bbl (4,917,100 cu m) through December 1994, with remaining established oil reserves of 6,005,314 bbl (954,300 cu m).

The reservoir is highly undersaturated, with the primary reservoir drive provided by oil expansion above the bubble point, solution gas, and limited bottom water drive along the downdip edge of each pool.3 In addition to Daly, Kirkella field, lying further north along this trend, produces from the same facies (Fig. 3).(31866 bytes)

Several observations based on examination of the reservoir at Daly can be applied to the Lodgepole play in the Dickinson area. One of the mechanisms at work in Daly field is Devonian salt dissolution and collapse, which is responsible in part for the subtle facies changes, irregular isopach patterns, and the local development of the "mound-like" portion of the reservoir. Structural features at Daly are influenced by basement tectonism related to the Birdtail-Wasada axis (Precambrian province boundary).4

Regional setting

The Williston basin is an intracratonic basin that encompasses most of North Dakota, parts of South Dakota and Montana in the U.S., and parts of Saskatchewan and Manitoba in Canada (Fig. 1)(19604 bytes). A complete stratigraphic succession is present in the central portion of the basin and represents all of the ages from Precambrian through Recent. The basin reaches a maximum thickness of 16,200 ft (4,939 m) in McKenzie County, N.D., which is in the central part of the basin.

Since the discovery of oil in the Williston basin in 1951, production has been primarily from the Madison group. This sequence reaches a maximum thickness of 2,000 ft (610 m) and consists of three formations, in ascending order, the Lodgepole, Mission Canyon, and Charles. In Canada the Bakken formation is included in the group. The Lodgepole conformably overlies the Bakken formation in the central part of the basin in North Dakota and Montana and in Saskatch- ewan and Manitoba (Fig. 4)(24648 bytes). It unconformably overlies the Three Forks, Birdbear, and Duperow towards the margins of the basin in North Dakota, South Dakota, and Montana. In turn, the Lodgepole is conformably overlain by the Mission Canyon formation in the central basin and is cut by the post-Mississippian erosional event along the basin margin. The Spearfish/Amaranth formation (Triassic-Jurassic ?) unconformably overlies the Lodgepole in this area.

Lodgepole stratigraphy

The majority of the production from the Lodgepole formation in the Williston basin is located in Manitoba. Several fields produce from a complex succession of shallow marine carbonates within the Lodgepole in the province (Fig. 3).(31866 bytes) The stratigraphy associated with the main producing fields has been formalized since their discovery in the early 1950s. There are two nomenclature schemes that have been used for the lower portion of the Lodgepole formation in Manitoba. One scheme applies to Daly and Kirkella fields; the other, based on Virden field, applies everywhere else in the lower Lodgepole in Manitoba. One terminology describes the shelf sequence (Virden) while the other describes the shelf-slope sequence (Daly-Kirkella).5 6

Young and Rosenthal6 related the north-south orientation of the Lodgepole shelf/slope break in Manitoba to basement faults. These faults focused the dissolution of the Devonian salts and influenced subsidence rates during the deposition of Lodgepole strata. In north-central North Dakota, it appears that the terminology used at Daly field for the lower Lodgepole applies to areas of known salt dissolution. Shoreward of the Daly trend, the stratigraphy follows Virden terminology (Fig. 5) (29480 bytes).

The stratigraphy of the Lodgepole formation in Virden field was formally defined by Stanton in 19567 and 19588. In those papers, Stanton stated that the nomenclature was applicable only to a narrow belt that parallels the eastern erosional edge in Manitoba and North Dakota. In the Virden area, the Lodgepole is divided into the Scallion, Virden, Whitewater Lake, and Flossie Lake members (Fig. 5,(29480 bytes)Table 3)(47646 bytes).

The terminology for the Daly field area was proposed by Organ and Russin.2 The original nomenclature they proposed was not intended for use outside Daly field; however, certain correlations using this terminology can be made throughout the region and can also be carried southward along the original extent of the Prairie salt in North Dakota into the Dickinson area.

Six different facies of the lower Lodgepole formation have been identified and named in the Daly area. They are, in ascending order: the Basal limestone facies, Cromer shale facies, Cruickshank crinoidal facies (secondary reservoir), Cruickshank shale facies, Daly member (primary reservoir), and an unnamed member (Fig. 5,(29480 bytes)Table 3)(47646 bytes). The facies changes observed in these members across the collapse area can be used to determine the timing of dissolution.

Correlations of the upper Lodgepole units, the Whitewater Lake and Flossie Lake members, are possible across the southwestern portion of Manitoba and throughout North Dakota and can be identified throughout capping the Lodgepole "shelf." While all beds comprising the Lower Lodgepole are deposited in basin-filling clinothems, all beds comprising the Upper Lodgepole lack such bedding style. The prominent facies changes observed in Manitoba and North Dakota affect rocks of the lower Lodgepole, below the Whitewater Lake member (Table 3).(47646 bytes)

Deposition

The complex stratigraphic relationship between the Daly and Virden areas in Manitoba, and their associated depositional environments, has been documented by Young and Rosenthal.6

Overall, the Lodgepole formation was deposited in subtidal, open marine conditions across a shallow dipping ramp.9 Biota present suggest the environment was well oxygenated and of normal salinity, unlike the anaerobic conditions during Bakken deposition. In general, the mud-supported rocks are suggestive of a lower energy environment. The grain-supported rocks represent higher energy, shoaling environments.

Regional equivalents

The facies described for the Lodgepole formation throughout Manitoba can be carried south across the border into north-central North Dakota and may continue along the original extent of the Prairie salt towards the Dickinson area. The facies in the lower Lodgepole of the Dickinson area is correlative with the facies that have been identified at Daly field. Correlations of strata across the Dickinson Lodgepole pool are shown in Fig. 6.(24610 bytes) The stratigraphic position for the productive facies may correlate with those at Daly field.

The basal limestone facies in the Dickinson area lies immediately above the Bakken and below what is referred to as the "false Bakken" (Fig. 6)(24610 bytes). This facies thickens toward the margin of the basin. The overlying argillaceous limestone, "false Bakken" is correlative to the Cromer shale facies, while the productive "mound" facies observed in the Conoco and Duncan wells is correlative with the Cruickshank crinoidal facies. It is recognized that there are noted lithologic variations between the facies at Dickinson and Daly. These variations represent depositional differences related to bathymetric position and the post-Mississippian unconformity surface that cuts the stratigraphic section at Daly. Correlations also show that the overlying Cruickshank shale facies, the Daly member, and the unnamed limestone facies are also present in the Dickinson area. The prominent gamma-ray deflection that appears on the logs above the Daly section in the Dickinson area is correlative with the lower Whitewater Lake member. The Cruickshank crinoidal facies and the Daly member in Daly field display neutron-density cross-over on the Compensated Neutron Formation Density logs and are readily recognized on this basis. These log characteristics are also present on the logs from Dickinson field.

Understanding the relationship of the Lodgepole formation and its facies as well as the influence of Devonian Prairie salt dissolution is important to understanding the "mound" play.

Salt dissolution

Devonian Prairie

The Prairie formation has the thickest single salt present in the Williston basin.10 It extends southward from Canada and is present over a significant portion of the basin (Fig. 1)(19604 bytes). It attains maximum thicknesses of 638 ft (195 m) in North Dakota, 300 ft (92 m) in Montana, 715 ft (218 m) in Saskatchewan, and 425 ft (130 m) in Manitoba.11 12 13

The Prairie thins to zero towards the basin margins. In addition there are recognizable salt dissolution areas. Capping the Prairie salt are red to green nonfossiliferous dolomites and calcareous shales of the "Second Red Bed." These dolomites and shales limited the dissolution of the underlying salts during deposition of the overlying Dawson Bay formation (Devonian).

Dissolution of the Prairie salt and its relationship to hydrocarbon trapping has been discussed by numerous authors for various parts of the Williston basin, and the mechanisms for salt collapse are readily accepted for specific areas.10 11 13 14 15 The Prairie acts as a "lever" connecting deeper basement movement to sedimentological anomalies recognized higher in the section. Explorationists en- countering anomalous features in these areas assume they may be totally or partly due to the removal of underlying salt.

Salt collapse features in these areas can be recognized from geophysical surveys, isopach and structure maps, and wireline logs. Information is further enhanced by examination of available cores that display collapse-fill facies that include insoluble residues and collapse breccia.

In areas where dissolution is complete, indirect indicators of salt dissolution and collapse must be used. There are several features that are indicative of dissolution. For example, in north-central North Dakota, these features include thickness anomalies in the "Second Red Bed," isopach thickness variations in overlying formations, fracturing as indicated by drillstem test recoveries, and the presence of thin, localized beds of salt within the Mission Canyon formation. One of the key indicators in this area is the presence of overly thickened Bakken or the presence of all three members of the Bakken. Where this occurs salt collapse is assumed. These anomalies can be used to map an erratic but steady progression of the Devonian salt edge across north-central North Dakota.

Traditionally, the southern edge of the Prairie salt has been considered to be the result of deposition, not dissolution. Detailed stratigraphic mapping along the southern limit suggests that this is not the case. Areas in Billings, Golden Valley, Dunn, and Stark counties show an anomalously thickened "Second Red Bed" due to dissolution rather than deposition10 (Fig. 7)(24144 bytes). The Bakken, again one of the key regional indicators of collapse, also has anomalous isopach variations. Similar anomalies can also be mapped in overlying formations.

Theory

Multiple stage collapse related to the dissolution of the Prairie salt is well documented throughout portions of the Williston basin. Examination of some of the processes in areas of salt collapse can be used to explain the features observed at Dickinson.

The process by which a collapse structure is formed is important to understanding the unique reservoir character and structural setting formed within rocks comprising such structures. In dealing with multiple stage salt collapse it is understood that salt removal is episodic. There is evidence that faulting induced by basement-related tectonic movements may control sites of dissolution and collapse.

The removal of the salts by subsurface waters leads to the creation of fluid-filled caverns within the salt mass. When the cavern reaches a critical size, the overlying rocks cannot support the overburden weight, resulting in failure of the roof and collapse of overlying beds into the cavern, thus forming a collapse breccia. Displacement of the fluid creates a series of injection fractures above and below the salt cavern. The brine that contains the insoluble residues, originally present in the salt, is forcefully displaced upward or downward as a result of the collapse. These insoluble residues often are incorporated within the collapse breccia or fracture network as either a clay matrix or as a fracture fill. The episodic forceful movement of the fluid throughout this fracture system results in the formation of distinctive "string of pearls" carbonate reservoirs composed of irregular vugs and open voids connected by a fracture network. Multiple stages of both carbonate dissolution and cementation may also be present and are related to the episodic movement of fluid through the system.

The process of salt dissolution, brine displacement, and overburden collapse creates localized depressions that are points of increased sedimentation. The resulting abnormally thick deposits of an overlying unit or units indicate the timing of collapse. This added sediment forms a "compensation zone" forming a "single stage" collapse feature.

If salt dissolution continues on a more regional scale, the remaining salt mass is removed creating multistage collapse features as the broader area of overburden collapses. Any single-stage structures present within the broader regional dissolution area now become "inverted" in that the thicker sediments forming the compensation zone cause doming of the overlying rocks, and "multistage" structure is formed.

Another type of structure results from a regional dissolution of a limited amount of Prairie salt. The partial removal results in vertically restricted collapse of the overlying beds. The geological result of this limited regional dissolution is the creation of porous fracture and breccia zones in normally dense rocks.

Collapse structures found within potash mines in southern Saskatchewan contain a variety of collapse-induced breccias, ranging from oligomictic to polymictic to "crackle" breccia.16 The breccias, if not plugged with insoluble residue, are extremely porous and permeable. Typically, the structures display complex reservoir distribution, internal structural architecture, and anomalous stratigraphy. The mechanical properties of the rock strata determine the type of breccia and/or fractures that formed. Rocks with similar properties respond in a similar manner.

Fig. 8(15278 bytes) shows the morphology of a typical single to early multistage collapse structure. The cavern fill illustrated on the figure typically consists of a mixture of the previously mentioned breccias. Locally, pockets and layers of very permeable and porous rock occur with rock of poorer reservoir quality. The entire cavern fill is cut by fractures that provide hydrodynamic continuity. The cavern fill may be reservoir quality depending upon the amount of post-collapse fracture fill and cementation. In many structures there are carbonate strata that have not been reduced to rubble during the collapse process but rather sagged into the cavern with a minimum of disruption. These intervals are typically extensively fracture, forming a very permeable reservoir mass. If the collapse structure occurs within an area where the undisturbed Paleozoic section is relatively tight, the breccia and fracture zones associated with the collapse structure can provide a reservoir capable of trapping migrating hydrocarbons. This is especially true of collapse structures that are isolated from areas of regional salt dissolution and that intersect regional carrier beds.

Collapse structures may be either vertically and laterally sealed or may be connected to regional aquifers. Often collapse structures are part of larger complexes that in turn connect to form laterally extensive dissolution systems. Collapse structures that are laterally sealed but vertically open may act as "chimneys" that allow higher reservoir beds to be charged with oils generated by stratigraphically lower source rocks. Collapse structures that are connected to linear basement features and which are laterally open may act as conduits for the regional migration of oil. Evidence for these fractures may be present on Compensated Neutron Formation Density logs (Fig. 9)(24617 bytes). Additionally, these fractures may act as the reservoir "pipeline," resulting in changes of oil column thickness across the reservoir. Oil moving through the system later may be stored in the fracture or within the vuggy porosity. Vuggy porosity that is present but not connected to the open fracture network at the time of migration will show no staining. Returning to Fig. 8,(15278 bytes) it becomes apparent that the oil trapping potential of a collapse structure is a function of the contrast between the porosity and permeability of the cavern fill and the lack of porosity and permeability of the surrounding, undeformed rock. If the surrounding undeformed rock is porous and permeable, then fluids will simply pass through the collapse structure into the surrounding rock. If the surrounding rock is generally dense, then oil may be trapped within the collapse structure. Furthermore, the distribution of fluids within the fractured rock of a collapse structure may be complicated by the presence of major displacement faults created during the process of collapse (Fig. 9).(24617 bytes) These displacement faults may act as "floor seals" limiting the lateral migration of hydrocarbons within the structure.

Dickinson applications

A comparison of features observed in the Dickinson area with those in areas of salt dissolution and collapse provide the evidence that this was the probable mechanism for initiation of "mound" growth.

Isopach maps constructed for the Dickinson area display thickness anomalies. Locally, thicker sections occur on isopachs of the "Second Red Beds," combined Birdbear (Devonian) to Bakken formations, Bakken formation, and of the Tyler formation. A notable thinning occurs between the base of the last Charles salt and the State "A" marker.

Cross-sections through the Lodgepole pool in the Dickinson area indicate through compensating thickening (Fig. 10)(29791 bytes) the timing of dissolution and collapse. Underlying the main portion of the "mound" is an anomalously thickened upper Bakken shale section 40 ft (12 m) instead of the regional 8-10 ft (2-3 m). The thickened Bakken section was deposited in a structural low as a result of the initial phase of Prairie salt dissolution and collapse. Following the multiple stage dissolution model, collapse of adjacent areas inverted the thickened Bakken section. The Bakken then became the localized high acting as the pedestal for "mound" growth (Fig. 11)(39081 bytes). Subsidence continued in the area adjacent to the "mound" at varying rates. In the areas flanking the "mound," lower energy argillaceous limestones were deposited. As subsidence continued and/or sea level rose, the growth of this assemblage or "mound" continued vertically as argillaceous limestones continued to be deposited along the flank. The difference in shape and distribution between the Daly and Dickinson areas is probably related to water depth and subsidence rate. Daly is higher on the shelf/slope allowing the "mounds" to spread out laterally with a significant component of talus or apron style mechanical deposition on the flanks. The Dickinson area is at the break in slope and growth is more vertical. By the middle of Lodgepole time, a deepening event flooded the area and growth of the biotic assemblage ceased. Dissolution through the area continued sporadically, as evidenced by small scale adjustments in the stratigraphic section.

The lower Lodgepole producing interval has been described as a crinoid and bryozoan dominated boundstone to bafflestone with associated grainstones and wackestones.17 The reservoir rocks are generally light colored with varying amounts of micrite. The rock properties of this clean carbonate "mound" section would allow fractures to propagate easier than the adjacent argillaceous limestones surrounding the mound. The reservoir is characterized as highly fractured with numerous large vugs enabling it to produce high volumes of fluid. Where not intersected by fractures isolated vugs lack oil staining. Multiple phases of cementation have also be described from the Dickinson cores.17 The presence of fractures, vuggy porosity, and cementation can be explained by the multistage collapse model.

Oil source

Two different oil sources have been proposed for the Madison group in the Williston basin. It has been concluded that the source for mid-Madison oil was not the Bakken.18 19 Osadetz et al.18 suggest a Lodgepole source for the mid-Madison. Price and LeFever20 state that the beds adjacent to the reservoir rocks, not the Lodgepole formation, are the probable source for the mid-Madison oils.

The source of the oil in the Lodgepole in Dickinson field is from the underlying "Bakken source system."21 The "Bakken source system" includes the lowermost Lodgepole formation, the Bakken formation, and the uppermost Three Forks formation.19 Three distinguishing characteristics type the oil from the Conoco 74 Dickinson State to the Bakken system based on whole gas chromatogram (Fig. 12)(14131 bytes). Firstly, Bakken oils and shale extracts have an odd carbon preference index (CPI) significantly greater than 1.00. Values determined range from a low of 1.06 CPI (n-C25 to n-C35) to a high of 1.13 CPI (n-C28 to n-C30). Oils derived from a carbonate source (e.g.: "Lodgepole") have CPI values that are always less than or equal to 1.00. Therefore, the oil from the Conoco well cannot be from a carbonate source. Secondly, the pristane-phytane ratio for this oil is 1.96. Bakken oils have values of 1.64 to 2.08, whereas Madison oils ("Lodgepole") have values ranging from 0.76 to 1.44.19 Thirdly, the production of a high gravity oil with significant quantities of gas and no water is characteristic of Bakken production (Table 2).(19646 bytes)

The Bakken formation underlying the Dickinson field area is immature to marginally mature. The Rock-Eval hydrogen index (HI) for Bakken in this area is about 500. This is substantiated by measured values for the pristane/n-C17 and phytane/n-C18 of 1.27 and 0.80, respectively. These values are very high because of the oil's immaturity. The oil also has slightly or moderately great- er contents of n-C20 to n-C30 n-paraffins because of the oil's immaturity; immature oils have more of these compounds and Bakken shale extracts near the depositional edge shales in this area of the basin are more paraffinic than the norm as documented in Price et al.22

Dickinson field does not differ from other potential Lodgepole play areas be- cause it overlies immature to marginally mature Bakken shale. This does not appear to be the important factor in determining the success of a potential prospect. Production of 40.2 to 41 gravity oil from the middle Bakken section at Daly field can be considered as unexpected inasmuch as the Bakken shale in the area is immature, well out of the oil generation window. Additionally, significant quantities of 33-35 gravity oil have been produced from the Lodgepole at Daly, Virden, and all of the other fields along those trends in Manitoba. All of those fields are out of the oil generation window and require either long distance migration23 or migration of oil fractionated from bitumen from immature source rocks.24 This oil is able to migrate because of physical disruption of immature shales25 from fracturing due to salt collapse.

Conclusions

The salt collapse model can be used to readily explain the features observed at Dickinson field. Using this preliminary model, the play expands regionally throughout North Dakota, Montana, Manitoba, and into Saskatch- ewan along the projected original extent of the Prairie salt. The use of Daly terminology throughout this area may help delineate the timing of collapse. It is also important to recognize that in a play of this nature there are going to be numerous variations in style and type of "mounds" and in fact Daly and Dickinson may only represent end-points.

The principal conclusion of this article is that it is not important whether or not the Dickinson pool is located within a Lodgepole "Waul- sortian mound;" rather what is important is that the pool is a structurally-controlled oil accumulation. The process of progressive multistage dissolution of the Prairie salt and the resultant collapse of overlying beds created a highly fractured and therefore porous and permeable section and a viable vertical migration pathway from a marginally mature source rock.

A lesson from the Dickinson pool that can be applied elsewhere in the Williston basin is that any manifestation of Prairie (and other) salt dissolution and collapse should be carefully examined by explorationists. Salt dissolution can be used not only as a possible mechanism to initiate "mound" growth but also as a means of delineating sites of regional tectonic activity. These reservoirs may not be restricted only to the lower Lodgepole formation. In fact, the identification of properly timed salt collapse structures may offer an alternative exploration strategy to companies wishing to duplicate the economic success of the Dickinson pool.

Acknowledgments

The authors thank Sidney B. Anderson, Robert F. Biek, Thomas J. Heck, Richard D. LeFever, Leigh Price, and Richard Sereda for their critical reviews of this article and Yockim Resources for use of maps and data.

References

1. LeFever, J.A., and Anderson, S.B., A little known carbonate buildup in the Lower Lodgepole formation, northwestern North Dakota, in Lorsong, J.A., and Wilson, M.A., eds., Oil and gas in Saskatchewan, Saskatchewan Geological Society Spec. Pub. No. 7, 1984, pp. 31-45.

2. Organ, D.W., and Russin, G.M., Mississippian stratigraphy of the Daly oilfield, Canadian Mining and Metallurgy Bull., 1956, pp. 193-198.

3. Manitoba Energy and Mines.

4. McCabe, H.R., Stratigraphy of Manitoba, an introduction and review, in Turnock, A.C., ed., Geological Association of Canada Spec. Pub. No. 9, pp. 167-187.

5. Sereda, R.D., Regional evaluation of the Lower Mississippian Souris Valley beds of southeast Saskatchewan, University of Saskatchewan, unpublished master's thesis, Saskatoon, 1990.

6. Young, H.R., and Rosenthal, L.R.P., Stratigraphic framework of the Mississippian Lodgepole formation in the Virden and Daly oilfields of southwestern Manitoba, in Christopher, J.E., and Haidl, F., eds., 6th International Williston Basin Symposium, Saskatchewan Geological Society Spec. Pub. No. 11, 1991, pp. 113-122.

7. Stanton, M.S., Stratigraphy of the Lodgepole formation, Virden-Whitewater area, Manitoba, 1st International Williston Basin Symposium, North Dakota Geological Society and Saskatchewan Geological Society, Bismarck, N.D., 1956, pp. 79-83.

8. Stanton, M.S., Stratigraphy of the Lodgepole formation, Virden-Whitewater area, Manitoba, in Goodman, A.J., ed., Jurassic and Carboniferous of western Canada, AAPG John Andrew Allan Memorial Volume, Tulsa, Okla., 1958, pp. 372-390.

9. Heck, T.J., Depositional environments and diagenesis of the Mississippian Bottineau interval (Lodgepole) in North Dakota, University of North Dakota, unpublished master's thesis, Grand Forks, N.D., 1979, 227 p.

10. LeFever, J.A., and LeFever, R.D., Relationship of salt patterns to hydrocarbon accumulations, North Dakota Williston Basin, in Hunter, L.D.V., and Schalla, R.A., eds., 7th International Williston Basin Symposium, Montana Geological Society, Billings, Mont., pp. 69-88.

11. Holter, M.E., The Middle Devonian Prairie Evaporite of Saskatchewan, Saskatchewan Department of Mineral Resources, Geological Sciences Branch, Report No. 123, 1969, 133 p.

12. Manitoba Energy and Mines, Mineral Resources Division, Structure contour and isopach mapDevonian Prairie evaporite formation, Stratigraphic Map Series DPE-1, 1980, 1 sheet.

13. Oglesby, C.A., Deposition and dissolution of the middle Devonian Prairie formation, Williston basin, North Dakota and Montana, unpublished master's thesis, Colorado School of Mines, Golden, Colo., 1988, 79 p.

14. DeMille, G., Shouldice, J.R., and Nelson, H.W., Collapse structures related to evaporites of the Prairie formation, Saskatchewan, GSA Bull., Vol. 75, 1964, pp. 307-316.

15. Martiniuk, C.D., and Barchyn, D., Petroleum potential of the pre-Mississippian, southwestern Manitoba: an introduction and review, Manitoba Energy and Mines Petroleum Branch Open File Report 14-93, 1993, 35 p.

16. Halabura, S.P., Danyluk, T., and Gerhardt, S.A., Reservoir development, structure, and hydrodynamics of salt dissolution and collapse structures: examples from the Saskatchewan potash mining belt, in Hunter, L.D.V., and Schalla, R.A., eds., 7th International Williston Basin Symposium, Montana Geological Society, Billings, Mont., p. 454.

17. Burke, R.B., and Diehl, P.E., Fracture and vugular porosity in the Dickinson Waulsortian-like mound: potential horizontal drilling target:, proceedings Second International Williston Basin Horizontal Drilling Workshop, North Dakota Geological Survey and Saskatchewan Energy and Mines, 1994, pp. B3-1-28.

18. Osadetz, K.G., Brooks, P.W., and Snowdon, L.R., Oil families and their sources in Canadian Williston basin (southeastern Saskatchewan and southwestern Manitoba), Bull. Canadian Petroleum Geology, Vol. 40, 1992, pp. 254-273.

19. Price, L.C., and LeFever, J.A., Does Bakken horizontal drilling imply a huge oil-resource base in fractured shales?, in Schmoker, J.W., Coalson, E.B., and Brown, C.A., eds., Geological studies relevant to horizontal drilling: examples from western North America, RMAG, Denver, 1992, pp. 199-214.

20. Price, L.C., and LeFever, J.A., Dysfunctionalism in the Williston basin: the Bakken/mid-Madison petroleum system, Bull. Canadian Petroleum Geology, Vol. 42, 1994, pp. 187-218.

21. Price, Leigh, U.S. Geological Survey, personal communication.

22. Price, L.C., Ging, T., Daws, T., Alonza, L., Pawlewicz, M., and Anders, D., Organic metamorphism in the Mississippian-Devonian Bakken shale North Dakota portion of the Williston basin, in Woodward, J., Meissner, F.F., and Clayton, J.L., eds., Hydrocarbon source rocks of the Greater Rocky Mountain region, RMAG, Denver, Colo., 1984, pp. 83-134.

23. LeFever, J.A., Martinuik, C.D., Dancsok, E.F.R., and Mahnic, P.A., Petroleum potential of the middle member, Bakken formation, Williston basin, in Christopher, J.E., and Haidl, F., eds, 6th International Williston Basin Symposium, Saskatchewan Geological Society Spec. Pub. No. 11, 1991, pp. 74-94.

24. Price, L.C., and Clayton, J.L., Extraction of whole versus ground source rocks, fundamental petroleum geochemical implications including oil-source rock correlation, Geochemica et Cosmochimica Acta, Vol. 56, 1992, pp. 1,213-22.

25. Price, L.C., Basin richness and source rock disruption, a fundamental relationship, Journal of Petroleum Geology, Vol. 17, 1994, pp. 5-38.

Bibliography

Berg, C.A., Virden-Roselea and North Virden fields, Manitoba, First International Williston Basin Symposium, North Dakota Geological Society and Saskatchewan Geological Society, Bismarck, N.D., 1956, pp. 84-93.

Bluemle, J.P., Anderson, S.B., and Carlson, C.G., Williston basin stratigraphic nomenclature chart, North Dakota Geological Survey Misc. Series No. 61, 1981, 1 sheet.

Canadian Society of Petroleum Geologists, Lexicon of Canadian Stratigraphy, Vol. 4, western Canada including eastern British Columbia, Alberta, Saskatchewan, and southern Manitoba, 1990, 772 p.

LeFever, J.A., Horizontal drilling in the Williston basin, U.S. and Canada, in Schmoker, J.W., Coalson, E.B., and Brown, C.A., eds., Geological studies relevant to horizontal drilling: examples from western North America, RMAG, Denver, 1992, pp. 177-197.

McCabe, H.R., Mississippian oil fields of southwestern Manitoba, Department of Mines and Natural Resources, Mines Branch Publication No. 60-5, 1963, 50 p.

McCabe, H.R., Mississippian stratigraphy of Manitoba, Department of Mines and Natural Resources, Mines Branch Publication No. 58-1, 1959, 99 p.

The Authors