EXPLORATION

May 26, 1997
The Rumsey Leduc reef (Late Devonian) in Central Alberta is an after-the-fact case history that illustrates the congruence of horizontal gradient intensity and seismic anomalies. Fig. 9 [37849 bytes] shows the stratigraphic relationship of the relevant formations that comprise the Late Devonian Woodbend Group. The Rumsey reef is situated on the Fenn-Big Valley Shoal near Stettler, Alta. The geology of that shoal was described by Andrichuk.18 Fig. 10 [39446 bytes] shows the shoal area and
Leonard A. LeSchack
Topaz Energy Exploration Ltd.
Calgary
The Rumsey Leduc reef (Late Devonian) in Central Alberta is an after-the-fact case history that illustrates the congruence of horizontal gradient intensity and seismic anomalies. Fig. 9 [37849 bytes] shows the stratigraphic relationship of the relevant formations that comprise the Late Devonian Woodbend Group.

The Rumsey reef is situated on the Fenn-Big Valley Shoal near Stettler, Alta. The geology of that shoal was described by Andrichuk.18 Fig. 10 [39446 bytes] shows the shoal area and specifically the dolomitization pattern of a 30 m slice directly above the Cooking Lake platform.

Andrichuk observed that at least 30 m of secondary dolomites indicative of subsequent Leduc reef formation can be seen underlying the Erskine, Stettler, Fenn, and Big Valley Leduc reef fields on the shoal. He further suggested that because this dolomitic trend extends about 7 miles southwest of Big Valley field, that area to the southwest may well contain productive reef buildups as yet undiscovered (in 1958).

The Rumsey reef, discovered in that area in 1982, is the most significant new productive Leduc build-up discovered on the shoal since 1958. The Rumsey reef is a "drowned reef," the drowning having resulted from subsidence of the Cooking Lake platform that halts growth of reef-building organisms before a full-grown Leduc reef is achieved. These Leduc drowned reefs are typically of small areal extent but are filled to the spill point with light gravity oil and are prolific producers. Such small reefs are particularly difficult to find by 2D seismic exploration unless the line passes directly over the apex of the reef.

The Rumsey reef was discovered by Gulf Canada as a result of a single seismic line that encouraged Gulf to lease the land, and two more lines used to define an anomaly (30 ms isochron thinning between the Viking (Cretaceous) and the Ireton (Late Devonian] formations). No clear seismic anomaly was seen at the Leduc/Cooking Lake target level, just a defocusing of the seismic energy. A well flowing up to 4,000 b/d of oil for 3 years resulted from a pinnacle reservoir covering only 16 ha (40 acres). To date, 3.8 million bbl have been recovered.19 Lemon20 acknowledged the discovery was based on serendipity as well as science.

Subsequent to the Rumsey discovery, Gulf conducted a 3D seismic survey to determine the full extent of the reef. In 1994 Gulf participated in a joint project with the author to share and make public the 3D survey in exchange for the author's HGI and DRAD survey of the same area. Gulf provided the 3D survey only after the author presented his survey results. Both surveys are shown in Fig. 11 [55597 bytes]. It is clear from Fig. 11 that the HGI survey targets the reef as does the 3D seismic survey.

The author measured at Rumsey field the magnetic effects due to well casings, pipe, infrastructure, and pipelines and has removed these effects before preparing the HGI map. Magnetic data recorded every 10 m from the 9-36-34-21w4 well showed that going either N-S or E-W, the effect of the well was no longer observed 80 m away from the well. Pipelines caused effects no farther than 20 m away.

Accordingly, total field magnetic data values within 80 m of wells and 20 m of pipelines were replaced by regional averages derived from values outside the affected zone, prior to computing the HGI. This modeling of cultural artifacts, by empirical means, especially effects of well bores and pipelines, is routinely conducted by the author in all surveys. Thus, individual cultural magnetic signatures can be accurately ascertained and man-made magnetic contamination can be eliminated prior to mapping.

The DRAD survey shown in Fig. 12 [46881 bytes] was conducted as described by Saunders et al.1-7 using the GIS-4 spectrometer. Thorium-normalized uranium and thorium-normalized potassium were computed, and the DRAD value, the difference between the two radiometric values at each station, are expressed in terms of standard deviations about the mean. The DRAD values in Fig. 12 show an apparent halo anomaly around the reef field. The author has observed similar halos, measuring in the total count mode, around the Morinville, Golden Spike, and Redwater Leduc reef fields. Saunders8 suggested that reef fields possibly are more likely to have halo rather than apical anomalies owing to probable differential compaction-related faulting around reefs.

Observations

In the two pre-drilling case histories at North and South Pierson, Man., seven wells were drilled on HGI and/or radiometric anomalies after completion of the surveys. All wells found oil, although not all in economic quantities.

At Waskada field empirical evidence suggests the DRAD survey specifically outlines the channels or "sweet spots" in the Lower Amaranth sand. When viewed in light of the results of the South Pierson survey shown in Fig. 13 [25620 bytes], it suggests that radiometric surveys are effective exploration tools for these channel sands. Similarly, empirical evidence also suggests the Waskada HGI survey outlines exclusively the Mississippian reservoirs underlying the Lower Amaranth sand.

Experience gained from the North Pierson survey that pinpointed reservoirs of Mississippian age that have subsequently successfully produced, as shown in Fig. 14 [56781 bytes], adds strength to this interpretation of the Waskada HGI anomaly map, for there is great similarity of the HGI anomaly shapes between the two.

In previous studies (LeSchack15 and proprietary surveys conducted by the author) it has been assumed that anomalies could not be attributed to any specific horizon in a multihorizon environment. Only qualitative indicators such as anomaly amplitude and anomaly shape were believed usable to make predictions.

For example, Devonian Leduc pinnacle reefs caused the highest intensity HGI anomalies, followed by Mississippian limestone reservoirs, and Jurassic and Cretaceous oil and gas sands cause the least intense anomalies.

Where encountered together, as is often the case in Alberta, the Cretaceous anomalies are superimposed on the Leduc anomalies but, owing to differing shapes and intensities, could qualitatively be separately identified by morphology. For example, in Fig. 11, anomaly patterns suggest a NW-SE Cretaceous sand channel overlays the ellipsoidal Rumsey reef.

There is much Cretaceous gas in the Rumsey area, and the gas well shown in the figure produces from the Glauconitic sandstone (Early Cretaceous). The radiometric anomalies show up as spatially associated halo or apical anomalies, thereby adding independent verification to the more location-specific HGI anomalies.

In the Waskada study, however, there is little discernible overlap between the HGI and DRAD maps, both appearing to successfully measure separate horizons. Barchyn12 observes that pressure maintenance through water and gas injection has been required in Waskada field owing to its being a low-energy solution gas drive reservoir.

Upward migrating hydrocarbon gas leaking from reservoirs is assumed to cause the diagenetic changes that HGI and radiometric surveys measure. If there is little gas pressure, there may not be enough leakage to cause discernible magnetic susceptibility change in the "Foote Horizon" the HGI measures, but perhaps it is sufficient to alter the potassium and uranium balance.

If this is true, one of the unexpected results of the Waskada study may be the suggestion of a correlation between HGI amplitude and initial gas pressure of the reservoir causing the anomaly. Leduc reservoirs, such as the Rumsey pinnacle, with the highest HGI measured, had sufficient gas pressure to flow most of its life, rather than be pumped as is the case with most Mesozoic reservoirs surveyed.

Subsequent proprietary surveys by the author appear to support this hypothesis. In the 39 section survey, 92 wells had been drilled. Of these, 55 were classified as Cretaceous producers, 15 produce from the Devonian Nisku formation, and 22 were abandoned.

HGI values associated with the location of each well were tabulated. The HGI values associated with each of the three classes of wells appeared to have a normal distribution distinct from each other. The Nisku distribution had a mean of 2.02 with a standard deviation of 1.08. The Cretaceous distribution had a mean of 0.82 with a standard deviation of 0.32, and the abandoned well distribution mean was 0.55 with a standard deviation of 0.28.

Based on these observations, theoretical probability distributions were computed for each well class. Fig. 13 shows plots of these distributions.

Orientation of anomalies

It now appears from examination of the residuals of total magnetic field intensity (the data from which the second horizontal derivative is computed) that in the studies so far conducted, there is a preferred orientation of the anomalies corresponding to their formations of origin, a potentially valuable diagnostic.

A typical residual magnetic anomaly has both a positive and negative portion, corresponding, respectively, to the north-seeking and south-seeking poles of a magnet. Zeitz and Andreasen21 provided examples for evaluating remanent magnetism anomalies observed from aeromagnetic data.

Using their technique, the azimuth of total magnetization of each residual anomaly associated with a producing well (or well with a significant show) was tabulated. The data included all wells observed in the author's case history and proprietary survey areas. The angle formed by present magnetic north and the line connecting the anomaly maximum (north) and minimum (south) was measured to approximate the true angle of azimuth of the measured total magnetization.

As an example, Fig. 14 illustrates how the azimuth of the total magnetization of the residual anomaly associated with the Rumsey reef was computed. It is the angle measured clockwise from magnetic north to the positive portions of the line connecting the maximum of the positive anomaly to the minimum of the negative anomaly.

The azimuth of the Rumsey reef anomaly is clearly at variance with the current azimuth of magnetic north. Four other Leduc reefs miles away had essentially the same residual anomaly azimuths.

Anomalies associated with Nisku reservoirs produced a family of azimuths that, although similar to each other, were statistically different from the Leduc anomalies. Similarly, Mississippian (Alida) anomalies were associated with azimuths much the same as each other but different from the Leduc and Nisku anomalies.

Finally, anomalies associated with the many different reservoirs of the Cretaceous were associated with a wide variation of azimuths, many suggesting reverse polarities, which were largely different from those associated with the Paleozoic reservoirs.

Making use of the association of reservoirs with horizontal gradient intensities as illustrated in Fig. 13, a plot of the total magnetization azimuths versus associated HGI values is presented in Fig. 15 [50907 bytes] in polar coordinate form. Each of these is identified with the producing horizon in a nearby well. Although only 58 anomalies are plotted, a trend with useful economic ramifications is suggested. The combination of azimuth of total magnetization and associated magnetic horizontal gradient intensity could be indicative of the reservoir which caused the anomaly.

The data presented in Fig. 15 are empirical and subjective, inasmuch as the residual anomalies are not always clearly defined when estimating their azimuths and intensities. It is also unclear why anomalies associated with reservoirs of different ages should exhibit different magnetic azimuths.

The diagenetic anomalies of enhanced magnetization in the "Foote horizons" appear to be at depths of 100 m to 250 m in the areas discussed meaning that they probably occur in Cretaceous or Cenozoic formations. Assuming that the enhanced magnetizations in the "Foote horizons" are less than about 125 million years old and that they average the geomagnetic field direction over thousands of years, we might expect the residual anomalies to exhibit azimuths (relative to present-day magnetic north) between 312-341° (normal polarity) or between 132-161° (reversed polarity), based on paleomagnetic reference directions provided by D.R. Van Alstine.22

The large spread in observed azimuths of the residual anomalies, especially the NE and SW azimuths associated with the Cretaceous and Nisku reservoirs probably reflects a superposition (vector sum) of normal-polarity (present field) and reversed-polarity (probably Early Tertiary) magnetizations. If so, the different azimuthal clusters of residual anomalies might reflect differences in composition and grain-size distribution of the authigenic magnetic minerals (magnetite, maghemite, pyrrho- tite, and/or greigite) in the "Foote horizons."

Coarser magnetic grains ( 10 microns) would record the present-field normal-polarity direction, and finer grains (< 10 microns) would record the ancient reversed-polarity direction. verification of this interpretation would require analysis of drill cuttings as described by foote in the papers cited above, followed by detailed paleomagnetic rock-magnetic analysis of these cuttings to determine the magnetic mineralogy and grain-size distributions.

Although the statistics used to create Figs. 13 and 15 were not drawn from a large sample, the trend of the separate probability distributions supports the hypothesis that the greater the original solution gas pressure in the reservoir, the larger the resulting HGI anomaly. Also, the azimuth of the total magnetization vector of the anomalies appears to be an indicator of the age of the reservoir that caused the anomaly.

If these hypotheses can be proved correct, there are important economic ramifications for HGI anomaly interpretation. In this geographical area, Devonian producers make 500-1,000 b/d of oil, while Cretaceous and Mississippian producers make 25-50 b/d of oil.

Survey costs

Although HGI/DRAD survey costs vary for many of the same logistic reasons that seismic costs do, an attempt to make a comparison of typical costs per section including permitting, data acquisition, data processing, and interpretation is shown in Table 1 [10458 bytes].

Costs are given in Canadian dollars and are based on multisection surveys searching for targets at a depth of 1,000-1,300 m in Central Alberta. It is also based on the author's current mode of collecting radiometric data which reduces overall data collection time from 2.5 min per point (as needed in the above-discussed case histories) to 1.5 min per point.

This is achieved using a Scintrex GSA-61 (1.8 l sodium iodide crystal) sensor, mounted on an eight wheel drive ATV, that requires only one-third the dwell time previously needed to obtain the same radiation counting statistics. As can be seen from Table 1, the total cost per section for the combined ground-based HGI and DRAD survey is approximately 15% the total cost for a 3-D seismic survey.

Conclusions

Seven wells were drilled to test anomalies found in the pre-drilling HGI and radiometric surveys at North and South Pierson and all found oil. Three new fields were discovered. The probability of this level of success occurring by chance is minimal.

The DRAD survey of the test site at Waskada field showed good qualitative correlation between anomalies and production. When examined in conjunction with the total count radiometric survey at South Pierson, it can be concluded that this technique is an effective way for mapping potential drilling locations in the Lower Amaranth, and presumably, similar channel sands.

The concomitant HGI survey at Waskada appears to quite clearly map the Tilston limestone underlying the Lower Amaranth sand. Most past surveys have shown that HGI and radiometric data mapped more or less the same anomalies with the HGI data being more location specific. As a result, the radiometric survey is used to help independently confirm the HGI anomalies.

In the Waskada study, however, two separate horizons are mapped with little apparent "crossover" effect. Because of the original low solution gas pressure in the Lower Amaranth sand, it is postulated that there is not enough gas leakage to observably affect magnetic susceptibility in the "Foote Horizon," but enough to alter the balance among the potassium and uranium ions at the surface.

As a result, the DRAD survey measured essentially the Lower Amaranth sand and the HGI survey measured primarily the Tilston limestone. Therefore, it is suggested that both surveys may be valuable in exploring this geological environment in the Williston basin where both reservoirs can be productive and seismic is not effective.

At Rumsey, the HGI anomaly is essentially congruent with the 3D seismic survey of the reef. The DRAD anomaly forms a halo around the reef as has been observed at several other Leduc reef fields. This suggests that HGI and DRAD surveys could be effective in locating small but prolific Leduc reefs for further definition by more expensive 3D seismic surveys.

Additionally, empirical evidence suggests that a combination of HGI values and true magnetic azimuth of the residual magnetic anomalies may be indicative of the reservoir that is responsible for the anomaly. Considerably more data are needed prior to suggesting a more substantive relationship.

Since the author began using HGI and/or radiometric surveys for predicting hydrocarbons in wildcat wells, eight wells were predicted to encounter oil, and seven of them (85%) did find oil. The one failure (also in the Williston basin) was drilled largely on the basis of a congruent seismic anomaly, which also proved erroneous.

The author's experience is that HGI and radiometric surveys substantially reduce exploration costs. The total cost, including permitting, surveying, data processing and interpretation, for the combined HGI and radiometric survey costs approximately 15% to total cost for a 3D seismic survey.

The HGI/DRAD combination is particularly effective when integrated with geological and seismic studies and assures a substantially higher probability of success than with geology and seismic work alone. These techniques clearly indicate probability of encountering hydrocarbons; however, they do not necessarily indicate whether any given well will be economic.

Acknowledgment

The author thanks Kevin J. Christie, chief geoscientist, and Dennis Dorval, geophyicist, of Gulf Canada Resources Ltd., for facilitating our surveys over the Rumsey Reef, providing the 3D seismic survey for correlation, and for permission to publish the results. The author also thanks Dr. Donald F. Saunders for providing valuable guidance on survey strategy during the conduct of the author's field work, and along with Dr. John M. Andrichuk for reviewing this article and providing many helpful suggestions. An abbreviated version of this article was published in CSEG Recorder.

References

1. Saunders, D.F., Burson, K.R., Brown, J.J., and Thompson, C.K., Combined geological and surface geochemical methods discovered Agaritta and Brady Creek fields, Concho County, Tex., AAPG Bull., Vol. 77, No. 7, 1993, pp. 1,219-40.

2. Thompson, C.K., Saunders, D.F., and Burson, K.R., Model advanced for hydrocarbon microseepage, related alterations, OGJ, Nov. 14, 1994, pp. 95-99.

3. Saunders, D.F., Burson, K.R., and Thompson, C.K., Observed relation of soil magnetic susceptibility and soil gas hydrocarbon analysis to subsurface petroleum accumulations, AAPG Bull., Vol. 75, No. 3, 1991, pp. 389-408.

4. Machel, H.G., and Burton, E.A., Causes and spatial distribution of anomalous magnetization in hydrocarbon seepage environments, AAPG Bull., Vol. 75, No. 12, 1991, pp. 1,864-76.

5. Foote, R.S., Use of magnetic field aids oil search, OGJ, May 4, 1992, pp. 137-142.

6. Foote, R.S., Integrating airborne and subsurface magnetic data: proceedings Fifth Thematic Conference, Remote Sensing for Exploration Geology, Reno, Nev., Sept. 29-Oct. 2, 1986, pp. 233-247.

7. Saunders, D.F., Burson, K.R., Branch, J.F., and Thompson, C.K., Relation of thorium-normalized surface and aerial radiometric data to subsurface petroleum accumulations, Geophysics, Vol. 58, 1993, pp. 1,417-27.

8. Saunders, D.F., personal communication, 1995.

9. Donovan, T.J., Hendricks, J.D., Roberts, A.A., and Eliason, P.T., Low-altitude aeromagnetic reconnaissance for petroleum in the Arctic National Wildlife Refuge, Alaska, Geophysics, Vol. 49, 1984, pp. 1,338-53.

10. Olea, R.A., Kriging: understanding allays intimidation, AAPG Geobyte, Vol. 7, No. 5, 1992, pp. 12-17.

11. Nettleton, L.L., Elementary gravity and magnetics for geologists and seismologists, SEG, 1971, 121 p.

12. Barchyn, D., The Waskada Lower Amaranth (Spearfish) oil pool, Southwestern Manitoba: a mode for Spearfish exploration in Saskatchewan, in Lorsong, J.A., and Wilson, M.A., eds., Oil and Gas in Saskatchewan, Spec. Pub. No. 7, Sask. Geol. Soc., Regina, 1984, pp. 103-111.

13. Kent, D.M., Depositional setting of Mississippian strata in Southeastern Saskatchewan: a conceptual model for hydrocarbon accumulation, in Lorsong, J.A., and Wilson, Eds., Oil and Gas in Saskatchewan, Spec. Pub. No. 7, Sask. Geol. Soc., Regina, 1984, pp. 19-30.

14. Martin, R., Paleogeomorphology and its application to exploration for oil and gas (with examples from Western Canada), AAPG Bull., Vol. 50, No. 10, 1966, pp. 2,277-2,311.

15. LeSchack, L.A., Ground-based magnetic horizontal gradient intensity and radiometric surveys, a cost-effective hydrocarbon exploration tool: three case histories in Western Canada, CSEG Recorder, Vol. 19, No. 8, 1994, pp. 7-13.

16. Foote, R.S., Horizontal gradient interpretation of low-altitude cesium vapor magnetometer data compared with measurement of rock magnetic susceptibility of drill cuttings, in Davidson, M.J., Unconventional methods in exploration for petroleum and natural gas IV, Dallas, Southern Methodist University Press, 1986, pp. 107-112.

17. Tompkins, R., Direct location technologies: A unified theory, OGJ, Sept. 24, 1990, pp. 126-134.

18. Andrichuk, J.M., Stratigraphy and facies analysis of Upper Devonian reefs in Leduc, Stettler, and Redwater areas, Alberta, AAPG Bull., Vol. 42, No. 1, 1958, pp. 1-93.

19. Lemon, T., and Taylor, B., The Rumsey Leduc pinnacle reef: where are the rest?, CSEG national convention, May 4-6, 1993.

20. Lemon, T., Gulf Canada Resources Ltd., Calgary, personal communication, 1993.

21. Zeitz, I., and Andreasen, G.F., Remanent magnetism and aeromagnetic interpretation, in Hansen et al., eds., SEG Mining Geophysics, Vol. II, Theory, SEG, 1967.

22. Van Alstine, D.R., Applied Paleomagnetics Inc., Redmond, Wash. 98053, personal communication, 1996.

23. Wendte, J.C., Cooking Lake platform evolution and its control on Late Devonian Leduc reef inception and localization: Redwater, Alberta, Bull. Canadian Pet. Geol., Vol. 42, No. 4, 1994, pp. 499-528.

The Author

Leonard A. LeSchack began experimenting with radiometric and magnetic HGI technologies in 1987 to aid prospect generation for his Calgary oil exploration company, Aquamarine Energy Exploration Ltd. He established Topaz Energy Exploration Ltd. in 1996 as a survey company to apply the technologies. He was a geophysical trainee with Shell Oil Co. in Houston and an assistant seismologist on Antarctic over-ice traverses during the International Geophysical Year 1957-58. During 1967-86 he operated LeSchack Associates Ltd. in the U.S. to conduct geophysical research for the U.S. government and numerous oil companies. He holds a BS degree in petroleum geology from Rensselaer Polytechnic Institute, Troy, N.Y., and studied geophysics at the University of Wisconsin graduate school.
Copyright 1997 Oil & Gas Journal. All Rights Reserved.