EFFECTS OF DIFFERENTIAL COMPACTION FRACTURING SHOWN IN FOUR RESERVOIRS

Gilbert E. Thomas
Thomas & Associates
Denver

With the advent of horizontal drilling in the U.S., fractured reservoirs have become a major target in the ongoing search for hydrocarbons.

Early optimism resulting from the high initial production rates for horizontal wells, however, is now being replaced by a more realistic view as many such wells are exhibiting rapid declines in production. This is especially true for the Mississippian Bakken shale play in the Williston basin and the Cretaceous Niobrara "chalk" play in the Denver basin.

This rapid production decline and the difficulty in finding other "sweet spots" to drill in fractured reservoirs are the two main problems plaguing the exploration side of horizontal drilling in the U.S.

Underlying these difficulties is the fact that fractured reservoirs are a mystery in themselves. As Petroleum Information Corp. says in a new publication about the Niobrara: "Little is yet known about the precise reasons for its sporadic productivity; the exact nature of the fracture system(s) within it; ... or the deeper structures and structural trends with which productivity appears associated."

Regarding these unknowns, the author's 30 years of experience in evaluating fracture systems and their causes both for tectonically-generated fractures and gravity-generated differential compaction fractures have shown that the problems of sporadic productivity, fracture system nature, and associated deep "structures" are mainly the result of believing that fractured reservoirs are primarily caused by tectonic deformation.

This paper will examine four fractured-reservoir fields in the U.S.: Silo (Niobrara), Wyoming; Elkhorn Ranch (Bakken), North Dakota; Pearsall (Austin chalk), Texas; and the Syndicated Options Ltd. 9372 Ferguson Brothers well (Ordovician carbonates), Kentucky. The paper will show that differential compaction fracturing is more of a major factor in long-term, sustainable production in a fractured reservoir than is tectonic fracturing.

First, a general discussion of the two types of fracturing and how they affect reservoir production.

TECTONIC FRACTURING

Recent articles about the tectonic generation of fractures in a reservoir have centered on the assumption that basement reactivated faults or lineaments are the cause.

For example, Fischer and Rygh,1 interpreting the cause of large, vertical fractures in the Bakken shales, believe the fractures "are related to mechanical stress and form as basement block adjustments are made along linear fault trends."

Significantly, they believe that these fractures are not a prime factor in production: "Such fractures contribute to the overall productivity of a well but are probably not essential for production."

Hamilton-Smith et al.2 believe the Ordovician fractured carbonates in Clinton County, Ky., "are probably due to tectonic deformation ... [which] Regionally ... is reactivation of basement faults." Similar assumptions about tectonic lateral adjustments on basement faults have also been used recently to explain the fractures in Silo field Niobrara reservoir.

If such tectonic reactivation of basement, subtle faults, or lineaments is truly the primary factor in fracture generation, then one should expect such megashears as the Lewis & Clark or such basement major faults as the Kentucky River fault zone to display fractured reservoir production along their extent where stratigraphic conditions are favorable.

But this is clearly not the case. So if the major faults or lineaments do not show the fractured reservoir effect, there is no reason to expect subtle, buried features to do so.

The Lake Basin fault zone in Montana, an en echelon series of small, normal faults, is an example of fractures formed by tectonic lateral adjustment on an underlying basement megashear near the Lewis & Clark.

The zone illustrates the problem with tectonically-generated fractures in a fractured reservoir. To be effective, a fractured reservoir not only has to have many fractures, but they must also be interconnected to provide the necessary permeability to profitably drain the reservoir.

The faults and fractures of the Lake Basin fault zone are clearly not interconnected (poor drilling results along the zone attest to this).

We have experienced similar results for tectonic fractures in many places-Lake Basin is the rule rather than the exception-and the reason is probably related to the episodic nature of tectonic adjustments on basement faults or lineaments. If adjustments were continuous throughout geologic time, interconnected fractures and profitable reservoirs would probably be commonly associated with major faults or lineaments.

In sum while tectonically-generated fractures, whether of regional or localized variety, are the rule in rock deformation, the episodic nature of tectonism is not conducive to producing interconnected fractures, which are absolutely necessary for sustained production from a fractured reservoir.

COMPACTION FRACTURING

Differential compaction fracturing, contrary to tectonic fracturing, is not episodic in nature.

The ever-present force of gravity, which generates compaction fracturing wherever differential strain occurs in the sedimentary section, is always active. So wherever adequately thick sedimentary units have been deposited over basement topographic irregularities, differential compaction fractures occur over and around the basement topographic features.

While it is true that mere compaction of sediments doesn't produce fractures and results in a progressive reduction of porosity with depth, differential compaction because of localized differential strain does produce fractures and affects permeability, the necessary ingredient for a fractured reservoir.

Furthermore, this always active process produces many microfractures which act as the vital connecting medium in an effective fracture system.

Do not think, however, that because gravity is always active in localized differential compaction zones throughout the sedimentary section that fractures extend continuously from the basement to the surface or that every stratigraphic unit is equally fractured. The effect of the ministrain in the differential compaction zones not only decreases upwards through the sedimentary section, but also is affected by the thickness and lithology of each unit.

A thick, massive limestone at depth might fracture more than a similar unit higher in the section but neither would fracture as intensely as a thin limestone separated by shales at either depth. It is precisely in this latter environment of thin limey chalk units or thin sandy units separated by shale or marl where differential compaction fracturing is most effective.

It is also in this same environment that differential compaction fractures in conjunction with tectonic fractures, produce the best fractured reservoirs of all.

The tectonic fractures in this case are commonly the most widespread (although nonconnected), while the differential compaction fractures where intersecting the tectonic fractures, supply the interconnectiveness necessary for localized sustained production (see Thomas3 for a fuller discussion of differential compaction fracturing).

SILO FIELD, WYO.

This field is a good example for the importance of recognizing the presence of differential compaction fracturing as a prime factor in the permeability of a fractured reservoir.

In an excellent article on the resistivity conditions at Silo field, Johnson and Bartshe4 found that, "The cause of the anomalously high resistivity is confirmed to be directly related to the presence of oil bearing fractures."

They further found the Silo fracture system has two fracture directions striking nearly at right angles-a widespread west-northwest/east-southeast trend and a more narrowly confined northeast-southwest trend.

Because low oil production is associated with the northwest-southeast system, they interpret this system to have "a network of small, poorly interconnected fractures." Conversely, the high oil production associated with the northeast-southwest system led them to interpret this latter system as having "more open, widely communicated fractures ..."

It is important to note that Johnson and Bartshe's northeast-southwest direction of fractures, filled with oil giving them their highest resistivity anomalies, is the same direction of fractures that Thomas had already inferred as differential compaction fracture zones, along with the highest annual production rates for vertical wells in Silo field occurred.

One might think that after the appearance of the Thomas article, the horizontal well site selectors would consider the northeast-southwest fracture direction for maximum potential.

But as Johnson and Bartshe point out a year later: "To date most horizontal wells drilled in Silo field have been oriented to intersect the west-northwest/east-southeast fracture systems [the tectonic direction]. This may account for the encouraging initial results of many of the horizontal completions and the subsequent rapid production declines" [this paper's emphasis].

Fig. 1 shows the paleotopographic features mapped in the Silo area plus the inferred zones of differential compaction fractures expected from the configuration and width of the paleotopographic features.

Also shown are the vertical-well locations and their annual average production rates as of December 1989, horizontal well locations as of June 1991, and the northeast-southwest fracture trends of Johnson and Bartshe based on resistivity data.

The coincidence between the differential compaction zones and the resistivity-determined fracture zones is readily apparent as are the locations of the best-producing wells within the zones. Note that the best producers almost always occur on the flanks of the paleotopographic highs where differential compaction strain is at a maximum.

The poorest producing vertical wells and dry holes are found either on the top of the broad paleotopographic high (where differential compaction strain is minimal) or between highs.

Note, too, the dramatic decrease of vertical-well production north of the Antelope Draw alignment where large paleotopographic highs are not present (two horizontal wells have been drilled in this poor producing area).

Also of note are localities A, B, C, and D. Seven of the best 10 vertical-well producers, including the top three, occur in these "pockets" or "elbows" formed where paleotopographic highs converge or intersect.

This arrangement suggests the paleotopographic pockets or elbows act as accumulation sites for hydrocarbons migrating from the west-southwest out of the basin. Once these sites are filled to the brim in localities that also contain differential compaction fractures, high production rates, as along the Goertz trend, are a possibility.

Undoubtedly, the widespread northwest-southeast tectonic fractures in Silo field have also contributed to vertical-well production, but the best productivity appears to occur only where the northwest-southeast tectonic fractures are intersected by the northeast-southwest differential compaction fractures.

As Johnson and Bartshe conclude, "Wellbores that communicate only with the west-northwest/east-southeast fracture system ... initially produce at high rates but, due to the limited volume of the interconnected fracture system, will decline precipitously and yield little cumulative oil."

Because only one of the nine horizontal wells shown in Fig. 1 appears to have tested a northeast-southwest differential compaction fracture zone, the poor, sustained production for these wells found to date, is likely due to the failure to recognize the presence of differential compaction fractures in the Silo reservoir and their critical function of supplying necessary permeability locally.

ELKHORN RANCH AREA, N.D.

The importance of recognizing differential compaction fractures in a fractured reservoir play can also be seen in the Mississippian Bakken shale play in the Williston basin.

Fig. 2 illustrates the paleotopographic features mapped in the Elkhorn Ranch field area and the inferred differential compaction fracture zones expected to develop over and around the basement features (over the narrow north-northeast and northwest features and around the larger north-northeast features).

For comparison, the inferred zones of increased fracturing in the Bakken as determined from well log analysis' are also shown.

It is apparent from the coincidence of the well log fractures with the differential compaction fractures that here, too, the compaction fractures are playing a major role in the fractured reservoir.

Note how the two sets of fractures coincide along the northwest paleotopographic trends at localities A and B. Also note how the well-log fracture zones change trend where the north-northeast paleotopographic highs are encountered.

It is evident here from Fig. 2 that a horizonal drilling program based on paleotopography and differential compaction fractures could more than hold its own versus one based solely on tectonic fracturing.

In fact, the premise of basement paleotopography and differential compaction fracturing would do much to solve the current problem plaguing not only the Bakken play but much of the horizontal drilling throughout the country, i.e., the difficulty in defining the best productive localities in a fractured reservoir.

Hansen and Long5 describe the problem in the Williston basin: "This fracture trend mapping reveals that the Bakken fairway is not uniformly productive. Rather, it consists of some heavily-fractured 'sweet' spots separated by relatively unproductive 'dead' zones."

PEARSALL FIELD, TEX.

Fig. 3 shows the gentle south flank of the Pearsall anticline near the town of Dilley, Tex.

According to Galloway, et al.,6 the fractures in the Austin chalk reservoir "resulted from regional extensional stress acting on brittle chalk units sandwiched between plastic shale and marl."

If the initial production rates of the horizontal wells shown in Fig. 3 are any indication of the intensity of fracturing in the Austin chalk and hence the interconnectiveness of the fracture system, then it is apparent from Fig. 3 that fracturing and permeability of the reservoir is highly variable. Production rates vary from more than 1,000 b/d to less than 250 b/d in a few thousand feet.

This variable production can perhaps best be understood by comparing the location of each well with the subsurface features present beneath the Pearsall anticlinal area.

These features include paleotopographic highs that trend north-northeast and northwest, the same trends as in Wyoming and North Dakota (the basement paleotopographic grain through the U.S. is north-northeast and northwest, paralleling the major Proterozoic arches and troughs) and a northeasterly, subtle structural trend, parallel to the main axis of the Pearsall anticline.

These subtle anticlines are deep seated and although not apparent in the structure contours on the top of the Austin chalk are indicated in the topographic patterns at the surface.

Note that many of the highest initial production rates in Fig. 3 are associated with these northeast anticlines (localities A, D, and E) while the other highest rates occur on the flanks of northwest paleotopographic highs (localities B and C). Perhaps more important, the lowest producers are found away from the paleotopographic and structural features.

These associations show that the deep-seated features beneath the Pearsall anticline are influencing the development of interconnected fractures in the Austin chalk reservoir by the process of differential compaction. Just by knowing where the underlying features are would be a great advantage in selective siting of wells even in such a profitable fractured reservoir as the Austin chalk.

So while tectonic regional extension fracturing at Pearsall might contribute a more or less equally spaced system of fractures, it is where differential compaction fractures locally intersect the regional set that the best interconnected system occurs thereby causing the highest initial production rates and presumably, the best sustained cumulative production.

CLINTON COUNTY, KY.

On Sept. 25, 1990, Syndicated Options Ltd. of Austria No. 9372 Ferguson Brothers vertical well encountered oil in the Middle Ordovician High Bridge group at a depth of 1,008 ft.

Initial production rates were reported to be 130-400 bbl/hr with cumulative production for the first 8 weeks said to be nearly 150,000 bbl.

According to a Kentucky Geological Survey report, the reservoir "is apparently a fractured carbonate rock and the fracturing is probably associated with reactivation of a basement fault..."6

Again, a fractured reservoir is "probably associated" with a reactivated basement fault.

But the gravity and magnetic maps in the area only display linear trends suggestive of basement faulting some 10 miles to the east of the discovery well.

If fracturing from a basement fault 10 miles to the east actually accounts for the fracturing near the well, then other wells drilled in the area should encounter similar "regional" fracture conditions.

But this is not the case as Hamilton-Smith et al. report, "Of the 32 permits issued in Carter Coordinate B-53 (the area of Fig. 4) since Oct. 1, 1990, only one has been reported as a producer to date ... located approximately 1,500 ft northwest of the Ferguson Brothers well, is reported to be flowing at the rate of 50 bbl/hr ..."

Both the discovery well (triangle) and the offset producer are shown in Fig. 4 plus the paleotopographic features found in the area. Based on this basement perspective, the interpretation offered here for the prolific production in the Ferguson Brothers well is a combination of flank differential compaction and hydrocarbon accumulation in an "elbow" as shown at locality A of Fig. 4.

Within the north-northeast paleotopographic trough, hydrocarbons would migrate up the plunge of the Cincinnati arch complex (the area lies to the east of the main arch axis) to where the northwest paleotopographic high (locality A) forms an "elbow" with the north-northeast paleotopographic high.

Further hydrocarbon migration would be hindered by the "elbow" obstruction leading to accumulation presumably to the "brim point." Because this "elbow" site would also be the focus of flank differential compaction fracturing over the paleotopographic highs, locality A would be a prime prospect for a fractured-reservoir extraordinary producer.

The fact that the only offset producer in the area occurs on the edge of the "elbow" lends credence to this interpretation and suggests that the Ferguson Brothers well encountered a unique situation of maximum fracturing and accumulation that does not appear to be very large (less than a mile wide) from a paleotopographic perspective.

A similar situation might exist at locality B against a larger and better expressed northwest paleotopographic high.

Hydrocarbon accumulation, however, might not be as great as at A because B lies in the migration "shadow" of the northwest paleotopographic high at A.

CONCLUSIONS

This article has illustrated the presence of differential compaction fractures in four fractured reservoirs-Silo field, Wyo., Elkhorn field, N.D., Pearsall field, Tex. and Clinton County, Ky-and pointed out that it is the compaction fractures, not the tectonic fractures that provide the necessary interconnected fractures necessary for sustained production.

The recognition of the role that differential compaction fractures play in a fractured reservoir will do much to help solve the plaguing problems of sporadic productivity and the defining of other "sweet spots" in a fractured reservoir.

The industry widespread belief that only tectonically-generated fractures are present in a fractured reservoir can only lead to the needless drilling of many less than profitable wells.

REFERENCES

Fischer, D.W. and Rygh, M.E., Overview of Bakken formation in Billings, Golden Valley, McKenzie counties, N.D.: OGJ, Nov. 20, 1989, pp. 71-73.
Hamilton-Smith, T., et al., High-Volume Oil Discovery in Clinton County, Kentucky Geological Survey Informational Circular 33, 1990.
Thomas, G.E., Silo area, Wyoming: Differential compaction fracturing, OGJ, Oct. 13, 1990, pp. 89-91.
Johnson, R.A. and Bartshe, R.T., Using resistivity to assess Niobrara fracture patterns in horizontal wells: OGJ, Sept. 2 and 9, 1991.
Hansen, W.B. and Long, G.I.W., Criteria for Horizontal and Vertical Prospects in the Bakken Formation, Williston Basin: Montana Geological Society 1991 Guidebook-Bakken Formation, 1991, pp. 151-163.