ABSTRACT

A two-year combined remote sensing and soil gas geochemical study was completed in Railroad Valley, Nye County, Nevada in 1984-1985.  These studies were designed to define the relationships between light hydrocarbon seepage anomalies, known oil reservoirs, and geologic features as identified from satellite imagery and aerial photographs.

The 1984 study included a regional investigation of the Railroad Valley Basin. Thematic Mapper, airborne radar, and high altitude aerial photography were examined to identify major structural provinces and regional lineament systems. Soil gas probe samples were collected in a regional grid to define predominant seepage pathways and to identify compositional signatures. Light hydrocarbon magnitude and compositional anomalies appear to reflect preferential migration along specific fracture systems. Although direct soil gas anomalies do not occur over the reservoirs, valley bounding faults adjacent to the oil fields do exhibit a significant flux of hydrocarbons updip from all the presently known accumulations. Compositional results mapped distinct seepage anomalies, which correlated with the production of the known oil fields.

The second year study conducted in 1985 examined the Grant Canyon field and Currant seepage anomaly areas with detailed soil gas sampling and high-resolution aerial photography. The Grant Canyon study area was confirmed to have a high magnitude seep updip and adjacent to production. This seepage appeared to be focused primarily near the valley boundary fault and to a lesser extent by other prominent fault and fracture systems.  Biogenic gas production was observed within the playa sediments directly over the field suggesting the presence of a macroseep.

Soil gas compositional subdivisions, defined in the first year's regional study, repeated on resampling and subdivide the Currant area into two compositionally different anomalies, which are separated by the Currant Creek regional lineament.  These seepage anomalies correlate with specific local lineaments mapped from aerial photographs, which appear to be controlled by subsurface structures.  Repeatability of magnitude and compositional results are compared for the 1984-1985 seasons.  Contour maps drawn from regional and detailed sampling of this anomaly demonstrate the importance of adequate sampling density on geochemical interpretations.

INTRODUCTION

This research program was initiated and supported by Japex Geoscience Institute and Earth Resources Satellite Data Center, Tokyo, Japan, as one of a series of worldwide studies to evaluate the application of remote sensing in oil and gas exploration.

The objective of this two year study was to determine whether or not geochemical anomalies occur at locations that can be predicted by any of the proposed structural models of Railroad Valley reservoirs, and to establish the relationship of structures to interpretations from remote sensing data in Railroad Valley.

GENERAL GEOLOGY

Although the Basin and Range Province has been sparsely explored for oil and gas, surface macroseeps are known and shows of oil and gas have been encountered in many exploratory wells.  Railroad Valley has been the most actively explored valley within the Great Basin since the discovery of the Eagle Springs field in 1954.  Currently there are seven producing fields in Railroad Valley, including Grant Canyon, Bacon Flat, Trap Spring, Eagle Springs, Kate Springs, Sans Spring and Duckwater Creek fields.  These fields have produced over 36 million barrels of oil as of 1993, of which over 34 million barrels have been produced from the Grant Canyon, Trap Spring and Eagle Spring fields.  The Trap Spring and Eagle Springs fields are a combination of stratigraphic, truncation subcrop fault traps, which may also be true of the subeconomic Currant discovery.  Production occurs from matrix and fracture porosity in reservoirs in the lacustrine Sheep Pass Formation (Cretaceous and Eocene) and the Garrett Ranch volcanic group (Oligocene), (Dolly 1979).  The most unique feature of these fields is that production occurs from the highest position on the lowermost fault block at the basin margin.  On the adjacent higher fault block the reservoir beds were removed or altered by erosion during Basin and Range deformation (Foster 1979).  The Grant Canyon and Bacon Flat fields are also located in structurally low, intermediate fault blocks, with production reported to be from the Middle Devonian Simonson dolomite (McCaslin 1984).  The Kate Spring field produces from both Devonian dolomite and Tertiary breccias (Herring, 1994).  The Sans Spring field, discovered in 1993, produces from Garrett Ranch volcanics, which are overlain by Tertiary valley fill (Grabb, 1994). More detailed discussions of Railroad Valley oil field geology can be found in Oil Fields of the Great Basin, R.A. Schalla and E.H. Johnson editors, (1994), and Hulen et al. (1994).   The sources of the oils in Railroad Valley are probably a combination of the Late Cretaceous-Eocene Sheep Pass Formation and Mississippian Chainman shale, (Poole and Claypool, 1984).

SUMMARY OF THE FIRST YEAR RESULTS

The first year's study consisted of an evaluation of Thematic Mapper (TM) and airborne radar (SLAR) images used to map lineaments and other geologic and geomorphic features, concurrently with conducting a light hydrocarbon regional surface geochemical survey.  Soil gas samples were collected in 125 cc septum cap bottles from 4-foot deep probe holes on a one-mile grid spacing as road access allowed.  The sampling hole was made by manually pounding a solid 1/2 inch steel bar probe into the ground to a depth of 4 feet.  After the bar was removed, a soil gas probe was inserted into the hole.  The septum bottles were evacuated to approximately 100 microns vacuum before each use, and were filled with soil gas to a positive pressure of about 7 psi using a hand pump attached to the top of the soil gas probe.  Hydrocarbons were measured on a flame ionization gas chromatograph as described by Jones and Drozd (1983).

Contour maps of the regional methane and propane soil gas data gathered in 1984 are shown in Figures 1 and 2, respectively.  Both components exhibit large magnitude geochemical anomalies which originate at the basin bounding fault and extend updip onto the adjacent pediment block.  A very simplified diagram, shown in Figure 3, helps to explain how this updip migration might occur through fractures and or draped sand lenses contained within the Tertiary fills.  Although this drawing does not attempt to explain the complex relationships between source and reservoir rocks, it does offer possible migration pathways from source rocks at depth to potential reservoirs and to the surface.  These geochemical anomalies clearly confirm the presence of active light hydrocarbon seepage over and adjacent to known oil fields and other untested areas in Railroad Valley.

The excellent quality of the TM and radar imagery allowed the mapping of regional and local lineaments over the entire valley, including the large playa in the center of the basin.  Combining this data with published information suggested a regional structural framework for the basin and established the geometric relationship of the known producing fields to the lineaments.  The regional geochemical survey provided data of sufficient resolution to establish both magnitude and compositional correlations of significant geochemical anomalies with lineaments, lineament intersections, and producing fields.  The results of this study were presented at the Fourth Thematic Mapper Conference (Jones et al., 1985).

Close examination of the regional hydrocarbon soil gas data from 1984, revealed that compositional ratio plots (Pixler, 1969)[1] of anomalous light hydrocarbons are able to clearly differentiate, at least two distinct hydrocarbon combustions which are suggestive of subsurface hydrocarbon type.  Compositional ratios over and updip from the producing Trap Springs, Eagle Springs, and Grant Canyon Fields all fall within the same compositional range, despite the different possible sources of the oil and depths of production.  Samples collected in the vicinity of the non-economic heavy Currant #1 well, show much oilier compositional ratios, which are distinctly different from those of the producing fields.

This is best defined by a color compositional dot map, as shown in Figure 4, in which the size of each dot is proportional to the ethane magnitude and the color is selected from the Pixler ratio plot (see inset to Figure 4).  Choosing the standard empirical cuts from Table 1 on this data set, shows that the producing oil fields fall within the yellow, rather than within the green areas, as would be expected for the heavy oils produced in Railroad Valley.  This color compositional dot map suggests that it is possible to differentiate between hydrocarbon type from each site's relative position on these Pixler ratio plots.  Eagle Springs, Trap Springs, and Grant Canyon fields have well-controlled intermediate compositions (yellow dots) while the Currant well area exhibits much lower, oilier ratios (green dots).  Thus hydrocarbon seep compositions observed in Railroad Valley appear to differentiate productive or potentially productive reservoirs from noncommercial, heavy oil accumulations at depth.  These compositional changes are closely related areally, suggesting that the mapped compositional changes may reflect subsurface geologic boundaries which control petroleum migration and accumulation.

These observed differences in the compositional ratios over the known production can be used to help segregate untested light hydrocarbon anomalies in Railroad Valley into potentially prospective versus nonprospective categories.  Many of the large magnitude anomalies are actually split compositionally, with the oilier signature occurring updip.  This relationship may reflect the loss of volatiles or water washing of the hydrocarbons which have migrated into shallower fault blocks without adequate trapping seals.  In general, productive type signatures occur along and basinward of the basinal fault, suggesting direct migration from accumulations at depth.  It should be noted that areas defined as having a productive signature reflect a geochemical similarity with productive fields and do not necessarily suggest an economic accumulation.  However, it is reasonable to suggest that these areas do appear to have a much higher potential for accumulations of producible hydrocarbons than those areas not exhibiting such signatures, greatly reducing the geographic area to be explored by more expensive geophysical exploration tools.

SUMMARY OF SECOND YEAR PROGRAM

Two areas from the first year's study were selected for additional remote sensing and surface geochemical surveys. The results of this second year study were reported at the proceedings of the Fifth Thematic Mapper Conference, by Burtell et al. (1986).

The areas were selected to provide more detailed information on the relationship of lower-order linear features identified on aerial photographs (photolinears) to Landsat (TM) lineaments and to their significance with respect to the leakage pathways, as suggested by surface geochemical signatures.  The non-economic Currant #1 well and the Grant Canyon field areas were chosen for this detailed study because of prolific seepage levels, contrasting surface environments and the differing patterns of the lineaments associated with each field.  The Grant Canyon area was also selected in order to investigate deeper geochemical sampling methods in the playa sediments.

Standard format, color infrared aerial photography was used to map photolinears and pertinent geomorphic features.  NHAP-CIR photographs (scale 1:80,000) were used to map photolinears on a regional basis, covering not only the study areas themselves, but also to provide continuity between the two areas for purposes of correlation.  BLM-CIR photographs (scale 1:24,000) were used to map photolinears in each project area.  The length of the photolinear mapped is related in part to the scale of the photography, and in part, to the resolution and quality of the photograph as well as to the spectral characteristics of the film.  The synoptic aspect of the NHAP photographs allowed the integrating of ground features over the longer distance than on the larger scale BLM photographs.  As a result, the regional photolinear study yielded photolinears ranging in length from 300 to 1750 meters, while those mapped from the BLM photographs ranged from 100 to 1500 meters in length.  From the standpoint of the objectives of the survey, the more detailed photolinear maps derived from the BLM photographs produced the best product. One of the most useful of these fracture maps is shown by Figure 5.

Discussion of the Currant Study Area

The Currant study area was selected for this detailed evaluation study because the original regional geochemical study showed an obvious hydrocarbon compositional change across the Currant lineament (Figure 4), which trends NE-SW through the center of the study area.  Although well-expressed to the SW,  (where it marks the Trap Springs oil field) and to the NE where it corresponds to the valley of Currant Creek, (here it is marked by intense fracturing in the rocks), it is only subtly expressed at the surface in the vicinity of the Currant well.  The lineament was defined by TM, Landsat, and radar imagery across the study area, and is reflected only by drainage lines and minor topographic features on the aerial photography.  The lack of good tonal definition is due to the general cover of recent out-wash sediments which lack variety in spectral reflectance.  The vein sorted and unconsolidated nature of these sediments also tends to produce diffuse photolinear signatures.  In addition, it was hoped that the geochemical survey would determine the reason for the lack of productivity in the Currant #1 well, which appeared to be favorably sited for fracture production because of its proximity to the Currant lineament.

Geochemistry, Currant Study Area

A total of 406 four foot probe, soil-gas samples were collected in 1985 on a 0.3 km grid over the Currant study area to further identify the geochemical trends and compositional variations which appeared to be correlated with the Currant lineament.  All soil-gas samples were analyzed by FID gas chromatography for methane through butane light hydrocarbons.  Contoured hydrocarbon magnitudes as shown by Figures 6 and 7 for methane and propane, form discrete linear anomalies, and groups of anomalies, which appear to be closely correlated with selected lineament zones identified from the aerial photographs (Figures 5, 6, and 7).  Although some of the contoured anomalies in the vicinity of the Currant lineament do reflect its orientation, hydrocarbon magnitudes do not appear to be solely controlled by the lineament.

Anomalous hydrocarbon zones extend fair distances from the lineament, indicating that this regional lineament does not simply act as a fault pathway to control the local magnitudes of hydrocarbon seepage.  Hydrocarbon magnitudes appear to be controlled to a greater degree by NS and EW small-scale linear features (Figure 5), which probably reflect the location of subsurface structural faults, and fault related fracture systems.

Regional geochemical data collected in 1984 showed a distinct compositional change across the Currant lineament, as defined by compositional ratio plots (Figure 4).  A scatter plot of methane to ethane data from the more detailed 1985 grid study clearly shows two distinct populations which can be divided on a C1/C2 ratio of 8.3 (Figure 8).  A further examination of compositional ratio plots generated from the detail grid data clearly shows there are actually two independent populations; non-productive oil and an intermediate population which lies in the productive signature range previously defined by the regional geochemical data.  Color dot maps of the 1984 and 1985 data demonstrate the obvious separation and repeatability of the compositional information (Figure 9).

This compositional change suggests that the Currant lineament does have a significant influence on subsurface geologic processes and/or subsurface fluid flow.  The lineament appears as a gravity embayment as it passes through the study (Guion and Pearson, 1978), but it is not clear what form the lineament takes at depth in this area.  The lineament could be a major fracture zone, normal fault, or other structural discontinuity.  Insufficient well control does not allow a comprehensive structural interpretation.  It is, however, apparent that the Currant #1 well was drilled right in the transition zone between favorable and unfavorable surface signatures.  This relationship is very important because faults identified by lineament zones are generally not the only controlling factor for light hydrocarbon seepage, but simply provide enhanced pathways of migration for gases and fluids.  As is demonstrated by this data, the local geologic framework and source potential are also important factors for interpreting the significance of both the hydrocarbon seeps and the lineaments.

In summary, geochemical data over the Currant lineament system is able to confirm the regional structural significance of the lineament even though it is not easily definable on a local scale from aerial photographic data alone.  The lineament appears to have a profound effect on the quality of hydrocarbon reservoirs expected in the subsurface, and suggests that this and similar features are important, if not essential in the mapping of geologic features related to the formation of economic hydrocarbon accumulations.

Discussion of Grant Canyon Detail

As with the Currant Area, the Grant Canyon area was selected for study because of the economic significance of this small but prolific oil field, and of its potential for developing a useful remote sensing-geochemical exploration model to aid in the discovery of hydrocarbon traps in similar playa environments.  The study area is located primarily within the Railroad Valley playa.  The area was sampled by augering 12 to 23 foot deep holes, which were sampled for soil gases or waters by the methods of Jones and Drozd (1983).  As expected, the presence of photolinears was more difficult to detect than in the Currant area because of the minimal tonal changes in the playa.  Sufficient data were acquired to demonstrate significant areal correlations between the anomalies of the detailed grid geochemical survey and with Landsat, TM and photogeologic features mapped.

Geochemistry, Grant Canyon Detail

Geochemical sampling included the collection of 186 samples from 12 to 23 foot augered holes in both the playa lake and alluvial deposits on the eastern side of the basin, centered over the Bacon Flat and Grant Canyon fields.  Water samples from the playa lake were collected and analyzed for dissolved light hydrocarbon gases where saturated sediments precluded the collection of soil gases.  Contoured geochemical data, shown in Figures 10 and 11 for methane and propane, exhibit a series of well defined high magnitude anomalies which reflect the valley boundary fault to the east, in addition to a large area northeast of the field where a major east-west linear feature intersects the valley bounding fault and related fracture systems.  This anomaly appears to reflect the updip migration of gases associated with the Grant Canyon Field.  A large north-south trending methane anomaly is also located west of the Bacon Flat wells.  Samples in this area had up to 10,000 ppm methane and were associated with black, organic rich, near surface sediments.  A stable carbon isotope measurement from this area has a value of -78.5 parts per mil.  A similar methane anomaly to the north has an isotope value of -65.6 parts per mil, suggesting at least partial petrogenic sourcing in this area.  Contoured propane values reflect a similar pattern to the methane anomalies, although the large methane anomaly west of the Bacon Flat wells is present only as a slight increase over background values on the propane map.  Propane anomalies over the field range from 100 to 300 ppb, which is 2 to 5 times background.

As shown by Figure 4, compositional data along the valley boundary fault and updip from the Grant Canyon Field area has a signature similar to the Trap Spring Field.  Sites within the playa lake have a more gas prone signature, which is a result of influence from biogenically generated methane.  The observation that biogenic gases are associated with macro seeps has been noted many times by extensive sampling in the Gulf of Mexico (Anderson, et al., 1983, Kornacki et al., 1995, and Reilly et al., 1995).  Although geochemical data directly over Grant Canyon is influenced compositionally by biogenic methane, the large magnitude methane to butane values observed updip do have a compositional signature similar to Trap Springs and Eagle Springs Fields.  These data highlight the area as having a high potential for producible hydrocarbons to exist directly downdip from these anomalies.

IMPORTANCE OF SAMPLE DENSITY

A regional geochemical survey on one-mile grids represent a low resolution approximation to the actual size or shape of any actual geochemical anomaly.   The methane and propane detail on 1000 foot centers is very different from the 1984 regional contour maps, as shown by Figures 6 and 7.  The sharp geochemical boundaries observed in the 1985 detail study cannot be mapped from the regional 1984 geochemical data.  This is shown very clearly in Figure 12, which expands the 1984 propane contour for direct comparison with the 1985 detailed grid data. Comparison of the magnitudes and compositions of these two data sets on the color compositional maps in Figure 9 proves that the 1984 data are valid and of good quality; however, using the 1984 data to draw contours is a serious mistake which results in an erroneous interpretation in regard to the location and shape of this complex anomaly.

Fracture orientations from the aerial photography in Figure 5 define and control the sharp boundaries of the geochemical anomalies, as shown by Figures 6 and 7.  The Currant lineament cuts through the center of this major seep anomaly and clearly does influence the geochemical signature, which is interpreted to be controlled by fluid flow at depth, thus, indirectly mapping the nature of the potential subsurface reservoirs.  The shape of the geochemical anomaly shows strong control by the bounding fractures, which have no obvious relationship to the regional lineament.

A comparison of the Currant lineament with the close detailed composite interpretation from aerial photography shows that the azimuth of the Currant lineament is expressed only in the short photolineaments.  However, the regional lineament is not obvious from only the short photolineaments within the valley.  Based on just the aerial photography, we might suggest that this lineament is not important; the geochemical data however, clearly shows otherwise.

The 1984-85 surveys also showed that a large number of high magnitude seeps occur near, or on lineaments and lineament intersections in Railroad Valley (Jones, et al. 1985).  This classic relationship reflects one of the most valuable usages of remote sensing lineament studies in frontier basins.  Preferential location of geochemical samples in the vicinity of active structural zones and their intersections will usually locate a large number of the hydrocarbon seeps in any basin.  In addition, regions of intense fracturing which do not exhibit hydrocarbon seepage, strongly suggest a lack of source potential at depth in such areas.  However, as shown by these surveys, magnitudes provide only part of the information conveyed by surface geochemical data; composition may be of even more importance.

The distinct compositional change associated with this regional Currant lineament suggests subsurface hydrodynamic processes related to the lineament may form a barrier to subsurface water flow and divert fluid to the east of the lineament.  Oil accumulations east of the lineament could, therefore, be water-washed, resulting in the non-commercial heavy oil observed in the Currant #1 well.  Potential petroleum reservoirs west of the lineament may be protected from water washing, retaining their volatile constituents, and providing a gassier soil gas signature west of this regional lineament system, even though the feature is not immediately obvious from small-scale remote sensing data alone.  It is also important to note that the regional geochemical study conducted in 1984 would not have been sufficient to support this interpretation, and that close detailed data gathered in 1985 were required to properly confirm the relationships between lineaments and hydrocarbon seepage in this case.

CONCLUSIONS

Remote sensing imagery is able to identify major fault and fracture systems which are related to the structural development of Railroad Valley.  These systems define individual structural blocks which can be further evaluated by geochemical data.  Geochemical magnitude anomalies are located along fractures and faults updip from the oil fields and other prospective areas of the valley.  Anomalies over pediment surfaces may reflect fracture migrated gas which probably originates in the graben structures.  A combination of remote sensing and soil gas sampling can identify major near surface gas migration pathways in this complexly faulted valley, suggesting a model for interpretation.

Combined remote sensing and geochemical data acquisition programs also provide a low cost regional and prospect evaluation, and are able to highlight regions which have similar structural and geochemical signatures to productive areas.  The Currant study proves the importance of large, regional lineament systems on the paleo environments which determine type and quality of subsurface hydrocarbons.  The lineament does not necessarily control the type or maturity of hydrocarbon sources, but is interpreted to have significant influence on fluid circulation and other subsurface geologic processes which control petroleum and reservoir formation.  High magnitude geochemical anomalies of a productive type signature located at the intersection of one or more regional lineament systems, are present at each producing field in Railroad Valley, providing a key to finding additional hydrocarbon accumulations.  These anomalies have been validated by repeat sampling on three separate occasions; 1977, 1984, and 1985 (Richers, 1985, Jones et al., 1985 and Burtell et al., 1986).  Regional geochemical studies encompassing the Great Basin have demonstrated the usefulness of this technology throughout the basin.  The careful interpretation of detailed geochemical data, coupled with remote sensing mapping, can greatly decrease the area to be searched within frontier areas and significantly increase the probability of successful exploration tests.