WILLISTON WAULSORTIAN MOUNDS—2: Geochemical innovations point to vast Lodgepole oil expanse

Nov. 16, 2009
New technologies reveal that Lodgepole reef oil discoveries near Dickinson, ND, may be part of a supergiant oil field that covers a much larger area of the Williston basin.

New technologies reveal that Lodgepole reef oil discoveries near Dickinson, ND, may be part of a supergiant oil field that covers a much larger area of the Williston basin.

Geochemistry innovation

Since collapse chimneys were not created equal, the ability to rank them and obtain at the same time a nonseismic confirmation of their location would obviously be helpful.

Fig. 12, an explorationist's desideratum (unequivocal geochemical measurements of whatever with high signal-to-noise ratio, repeatability, and significance) led the author to delve the morass of geochemical literature in the hope of finding a jewel lurking in its murky depths.

The author subdivided the subject into: 1) indirect geochemistry, 2) alkane geochemistry, and 3) aromatic geochemistry.

Indirect geochemistry

Even a casual acquaintance with this subject in my opinion indicates this superstition is no more useful for finding oil than is astrology and that further investigation is a waste of time.

Alkane/alkene geochemistry

This technology posits that since petroleum does occasionally rise from underground reservoirs to create visible macroseepages, microseepages that only instruments can detect should also exist.

The logic is impeccable, but countless failures and two pertinent studies show that alkane microseepage geochemistry cannot work—even though it has been carried out since 1929.

Smith and Ellis6 demonstrated in 1963 that the decomposition of roots and grasses in soils creates the same alkanes and alkenes assumed to come from microseepages and warned that "these findings seriously challenge the usefulness of this approach to prospecting for petroleum," a tactful way to state that alkane geochemistry doesn't work.

Persistent doubts about the value of alkane geochemistry led the Geological Survey of Canada and industry in 1970 to run a huge survey in Alberta that covered 540 sq miles and used 4,561 core holes to retrieve samples later dissolved in acid and analyzed for desorbed alkanes "a la Horwitz."7

Limestones are said to adsorb and desorb 10 times more alkanes than clays and clays 10 times more alkanes than sandstones. This huge sampling variation, which changes from sample to sample, has not been and cannot be adequately corrected for. The GSC's conclusion was that the recovered alkanes correlated mostly, if not wholly, with soil lithology, were most likely generated in soils, and were not microseepages rising from the subsurface.

The report's senior author volunteered that the method is useless.

Direct aromatic geochemistry

It would be nice for aromatic hydrocarbons to only pop out of microseepages, but studies indicate that they, too, may originate in near surficial sediments.

One likely reason for the recent appearance of aromatics in shallow sediments (and so-called biogenic gas) is the emergence of instruments that allow detection in parts per billion. To circumvent this problem we first determine, in the field, the background level of aromatics, and then know that the soil-air aromatics we sample mostly derive from microseepages.

Two types of commercial aromatic hydrocarbons geochemistry were available. One requires burying traps in soil for 3 weeks before analyzing their catch; the other detects benzene, toluene, ethylbenzene, and xylene directly in soil-air.

Our tests were conducted on the axis of the collapse chimney of our first discovery in Dickinson (Fig. 13). The chimney, offset from the contamination of surface production facilities, offers a perfect testing ground.

Conoco used seismic to site its vertical No. 1 dry hole. No. 2, a directional reentry from No. 1 and entirely based on our A II, hit the reef and had an 1,800 b/d IP. No. 3 was drilled in defiance of A II vehement prediction of a dry hole. This well missed the reef core and hit the tight flank, as did No. 1. An attempt to convert it into an injection well was a complete failure, following an immediate communication between wells 2 and 3.

The "buried traps" aromatic geochemistry found that chimney and background gave identical measurements. The other method using soil-air aromatic geochemistry proved hugely nonrepeatable.

Clearly, if we wanted the profiles of Fig. 12 (they are actual A III geochemistry, published here for the first time), somebody had to construct ad hoc instruments and invent algorithms to interpret their measurements.

Fortunately the author's partner is an engineer, an essential requirement for although all the elements of our prototype are off-the-shelf items, the machine assembled according to instructions gave measurements charitably described as amusing. It took months of redesign and testing before analyzer, carrier gas, calibration gas, digitized display, sampling hardware, and handling protocols that constitute A III aromatics analyzer worked to our specification of 1 ppb (which is like measuring 1 km to an accuracy of 0.001 mm).

Rudely, the field tests revealed three unexpected problems which, unless solved, would make our own aromatic hydrocarbon geochemistry useless.

Mobile equilibrium

The three aromatic gases benzene, toluene, and ethylbenzene, sampled in soil-air, are found between their melting and boiling temperatures.

This means that Le Chatelier mobile equilibrium partitions them between liquid and gaseous phases (alone analyzed), which in turn means that depending on the soil-air temperature samples at the same location may register as anomaly or background.

Obviously, in most cases, in order to be meaningful the raw measurements need be normalized, which has been accomplished.

Venting

The rising aromatic microseepages emerge from the phreatic zone to mix with the air of the vadose zone; there changes in pressure and temperature cause the commixture to leak out to atmosphere at rates that depend on the vertical transmissibility of near-surface rocks—a hard shale being the best and gravel the worst cap.

Thirteen raw benzene measurements taken at 20 ft intervals over the heart of the Gruman 18-1 anomaly (Fig. 13) varied between 18 and 134 ppb, a huge variation due to venting which makes the raw measurements meaningless. If uncorrected, our soil-air geochemistry would be invalid as many others had been.

The effective corrections for venting to atmosphere (too bulky for this summarization) have been worked out.

A V finger- printing

The various oil and gas sources are reflected in their respective aromatic microseepages.

Thus, it is possible to distinguish the Lodgepole petroleum microseepages from that of shallow gas channels, for instance. It is also possible to map (then drill) odd-shaped shallow gas accumulations by running a gridded survey as was done in Montana (Fig. 14). This obviously beats random drilling.

We have called this innovation A V technology. I know no other method that can do this.

Results to date

The account of the first applications of A I and A II technologies gave the results of 21 exploration wells in the Fort Worth basin of North Texas. All these wells hit the target reefs.

In Dickinson the first three were sited solely on AI and AII technologies and found that not all Lodgepole reefs had equal oil potential.

The need to highgrade the identified Lodgepole reefs led to A III technology. No well has been drilled yet using the A V innovation.

To date eight wells have been drilled in the Williston basin using these new technologies, and six of the eight found high reef and collapse chimney. One well missed (surveyor's error), and one well hit forereef, which area we will avoid in the future.

Completions and observations

The recoveries of the Dickinson wells demonstrate the importance of drilling close to the (X) of the A II target (Fig. 15), seeing that 400 ft away from the X of an A II has invariably yielded dry holes. The author's regression of this plot indicates that the expected recovery of a well drilled on an (X) to be 4 million bbl (Fig. 2).

Our drilling shows that A I finds reefs and A II finds reef-generated collapse chimneys, but that the Madison is the fractured reservoir and leaving acid in the formation longer than it takes to change connections and reverse it out results in the precipitation of the products of attack of carbonates by acid as the acidity of the solution decreases.

Also, the acid extracts sludge from the Lodgepole oil that effectively plugs the fractures. In the last three wells, the operators used acid, and although hydrocarbons were present, the fracture system was plugged with the products of the acid precipitates. This experience will avoid similar plugging in the future.

The problems of producing the postulated Dickinson supergiant oil field should be solved, at least for those operators willing to apply sound science and the lessons learned.

Conjectural reserves

There is no shortage of drilling locations seeing that A II has found thousands of outcropping collapse chimneys in the basin such as that shown in Fig. 16.

Ground shot of surface-visible A II feature with a diameter of 700 ft (Fig. 16). Photo by Robert J. Angerer Jr.

Only a tiny area of the potential ca. 25,000 sq miles of the Dickinson oil field has so far yielded production; this restricts reserve estimates to the category of wild guesses.

The simplest of these anchors on Eland oil field, said to cover 51⁄4 sq miles and have an estimated ultimate recovery of 32 million bbl, which gives a recovery of about 6 million bbl/sq mile.

Prorating this figure (or 0.02% of the prospective area) to the area of the Dickinson supergiant oil field yields a most satisfying—if highly aleatory—150 billion bbl potential and puts it in the category of Alberta's Athabasca tar sands. Dividing this figure by 10 still yields a giant field.

References

1. Price, L.C., and LeFever, J., "Dysfunctionalism in the Williston Basin: The Bakken/mid-Madison petroleum system," Bull. Can. Petro. Geology, Vol. 42, No. 2, June 1994, pp. 187-218.

2. Downey, Joe S., "Geohydrology of the Madison and associated aquifers and parts of Montana, North Dakota, South Dakota and Wyoming," USGS Prof. Paper 1273-G, 1984.

3. Downey, Joe S., and Dinwiddie, George A., "The Regional Aquifer System—The Northern Great Plains and parts of Montana, North Dakota, South Dakota and Wyoming: summary," USGS Prof. Paper 1402-A, 1988.

4. Azad, J., "Remote sensing imagery analysis method lifts veil on pinnacle reefs in US basins," OGJ, May 29, 2000, p. 33.

5. Story, Chip, "3-D Images Active Gas Changes," AAPG Explorer, June 2002.

6. Hunt., D., et al., "Paleokarst recognition and 3-D distribution," P032, EAGE 65th Conference and Exhibition, Stavanger, Norway, June 2-5, 2003.

7. Smith, G.H., and Ellis, M.M., "Chromatographic analyses of gases from soil and vegetation, related to geochemical prospecting for petroleum," AAPG Bull., Vol. 47, No. 11, November 1963.

8. McCrossan, R.G., "An evaluation of surface geochemical prospecting for petroleum, Caroline area, Alberta," conference Third International Geochemical Exploration Symposium, Toronto, Apr. 16-18, 1970.

Montana

Bill Barrett Corp., Denver, said it expensed four vertical wells in the quarter ended Sept. 30, 2009, at its Circus prospect in the Montana overthrust southeast of Helena.

The company focused on the Upper Cretaceous Cody shale, 900-2,000 ft thick at 3,000-7,000 ft, "to identify a large, repeatable natural gas resource play, but test results instead indicate more complex geology than anticipated that is not aligned with the company's strategy and timeline for development."

Barrett tested the Cody shale in three vertical wells drilled in 2008. Results varied but were noncommercial and included gas flows up to 1.1 MMcfd, oil flows up to 117 b/d, and large quantities of water.

Utah

Clayton Williams Energy Inc., Midland, Tex., plans to drill a third exploratory well in the last quarter of 2009 in the Central Utah thrust belt.

Target at the Maple Canyon prospect, in Sanpete County north of Providence oil field, is the Jurassic Navajo sandstone formation.

Williams, which holds 22,755 net acres and has a one-third interest in the play in Sanpete County, has previously drilled two dry holes in the northwestern part of the county.

Correction

Here is the correct Fig. 7 that was intended to run with Part 1 of the foregoing Williston Lodgepole article by Jamil Azad (OGJ, Nov. 9, 2009, p. 32).