Steven C. KrehbielGeophysicist Morgantown, W.Va.
The U.S. Department of Energy has conducted a series of studies investigating the hydrocarbon potential of the Pacific Northwest, intended to supply useful background information to companies interested in exploring this under-drilled area.
A 140 mile (230 km) seismic reflection survey was a key component of the project, designed to describe the regional structural geology and to determine the practicality of using reflection techniques in this region. The results of the project support previous indications that significant thicknesses of untested sedimentary rocks with hydrocarbon potential may exist in the Pacific Northwest.
The boundaries of the study area (Fig. 1) were based on a magnetotelluric survey conducted by the U.S. Geological Survey that delineated an anomalously conductive rock layer in an area roughly bounded by Mt. St. Helens, Mt. Adams, and Mt. Rainier. This feature, named the Southwestern Cascades Conductive Anomaly (SWCC), was interpreted to be caused by a sedimentary sequence underlying the surface volcanic layer.1 2
Previous exploration activity had not tested these inferred sediments, but other drilling in the Pacific Northwest has found numerous shows of both oil and gas. This culminated in the discovery of Mist gas field (more than 42 bcf cumulative production) of northern Oregon, the region's only commercial field to date.
Much of the older exploration was conducted before the development of modern plate tectonic theory. With the improved understanding of the regional framework, explorationists have a starting point for using the older data to develop new plays.
REGIONAL FRAMEWORK
The regional structure and resulting depositional systems have been controlled by the convergence of the Juan de Fuca and North American plates.
Continental/oceanic convergent margins of this type often show extensional features on the continental plate. However, simple extension does not adequately describe the more complex structural style in this part of the Pacific Northwest.3 These include the oblique angle between the Juan de Fuca spreading center and the continental margin (Fig. 2), and the northerly motion of the Pacific and Juan de Fuca plates with respect to the North American plate.
This combination of relative movement between the plates may affect the regional structural style, since the shallow dip of 3 near the coast to about 25 under the Cascades4 puts the downgoing oceanic slab and the bottom of the North American in proximity for a considerable distance behind the edge of the margin. This allows coupling between the plates and the transfer of some of the lateral motion from the oceanic plate to the continent.3 Paleostructure is further complicated by widespread rotation of a segment of the North American plate beginning in Eocene, as suggested by paleomagnetic data.3
The combination of the right-lateral movement, tectonic extension, and long-term regional rotation makes categorizing the resulting structural style difficult. Available data indicate a pattern of high angle, normal faults, often bounding extensional features such as pull-apart basins, together with localized compressional structures that may result from adjustments to the regional movement. These characteristics are similar to a wrench/divergent style.56 However, the geological structure is complex and difficult to analyze,7 particularly when smaller areas are examined without recognizing large scale mechanisms.
The regional geology has also been influenced by the westward growth of the continental margin since late Mesozoic time. Details of the growth process are not well understood. However, allochthonous terranes have been identified throughout the western margin of the North American continent8 and often appear to have been rafted to the continental margin as part of the subduction process.
One of the largest of these allochthonous terranes is the Coast Range of Washington, sometimes referred to as Siletzia. It is largely composed of the Crescent formation, a mid-Eocene unit of pillow basalts and tholeiitic basalt flows interbedded with volcaniclastic and sedimentary material. A variety of interpretations for the origin of the Crescent have been proposed.9
One suggestion is that the basalts originated as a sea-mount complex related to the passage of the Yellowstone hot spot. Another interpretation hypothesizes that the basalts originated from a volcanic ridge or a short lived spreading center10 close to the mid-Eocene coastline. In both scenarios, the material was extruded and quickly accreted to the continental margin, forming the present coastline.
The westward growth of the continental margin caused a matching migration of the magmatic arc, leaving behind abandoned remnants of older activity. Examples include old magmatic arcs, forearc basins, accretionary prisms, and trench fill. These artifacts have been distorted over time by subsequent geological processes, but might be found from the original continental boundary to the coast.
STRATIGRAPHY
The margin growth and late Eocene addition of the Siletzia complex to the continental margin had profound effects on the subsequent depositional systems.
While earlier sediments (Fig. 3) are essentially marine, as characterized by the sandstones of the Eocene McIntosh and Megler formations, younger rock types are increasingly continental in origin.
This can be demonstrated by the downward progression from the continental rocks of the Oligocene Lincoln Creek and Ohanopacosh formations, to the late-Eocene, fluvial/deltaic Skookumchuck formation, and finally grading to the mid-to-late Eocene marine McIntosh formation.
Volcanic influences first became evident in the late Eocene, with the intermittent flows and volcaniclastics of the Northcraft and equivalent Cowlitz formations. The level of volcanic activity gradually increased until the current environment was developed.
HYDROCARBON POTENTIAL
The sediments in the study area could be expected to have increasingly favorable hydrocarbon potential with increasing age and depth, matching the change from the younger, volcanic depositional environment to marine.
Representative borehole data showing this progression come from the deepest well in the region, the 10,800 ft Shell 1 Thompson drilled in 1959.
The well was drilled in the Chehalis basin, a forearc basin west of the study area, and tested much of the late Eocene nearshore/continental sequence (Fig. 4). It was spudded in the volcanic Oligocene Lincoln Creek formation before completely penetrating the late Eocene Skookumchuck formation, including a thick volcanic and volcaniclastic layer of Northcraft formation interbedded within the Skookumchuck. A marine sandstone was reached at TD, tentatively identified as the McIntosh formation.
Although several gas shows were found in the Skookumchuck formation, the well was neither economic nor particularly encouraging. Reservoir quality of the Skookumchuck sands was poor, and thermally immature rocks were found throughout most of the total depth of the well. Rocks within the oil generation window were only found at the marine sandstone reached at TD.
SEISMIC DATA
The DOE geophysical acquisition and processing methods for this study used available techniques: a Vibroseis source and a sign-bit recording system (see table), followed by conventional data processing methods.
The data set consists of six DOE seismic lines, recorded between 1989 and 1991, and two commercial lines that were recorded in the early 1980s. This line geometry (Fig. 5a) gave two east/west profiles connected by a north/south tie line. Note that DOE lines 1 and 2 were combined in the data processing stage and are referred to as line 1 & 2.
A depth-converted line drawing of the longest east/west profile, about 83 miles in total length (Fig. 5a) and derived from the southern-most commercial seismic line with DOE lines 1 & 2 and 3, is shown in Fig. 3b. It begins on the Crescent formation outcrop on the eastern side of the Coast Range that forms the western edge of the Chehalis basin.
From there, the profile passes directly over the location of the 1 Thompson well and continues eastward through the SWCC to the western flanks of Mt. Rainier.
The Chehalis basin is clearly visible as a shallow depression, with stratified reflectors extending for approximately 2 sec (12-15,000 ft) at the depocenter. High angle normal faults are numerous and appear to have been formed during the late-Eocene to early Oligocene.4
Data quality is generally good over the western part of the line but deteriorates at roughly the western limits of the volcaniclastic material and flows extending from the Cascades.
The DOE data, benefiting from their higher fold, show interpretable reflectors along the rest of the profile. Discernible details include: a northwest-trending surface feature called the Morton anticline at CDP 700, effects from a Miocene-age intrusive just north of the profile at CDP 7800 (schematically represented on the line drawing), and a wide depression centered at CDP 9700, immediately west of Mt. Rainier.
Reflector quality is fair to poor throughout much of the DOE profile, apparently related to a combination of complex structure and a troublesome surface noise.11
The noise appears to be a trapped wave or reverberation within the surface layer that proved difficult to completely remove by traditions stacking and noise reduction techniques. Resulting reflectors were somewhat discontinuous but could be seen throughout most of the profile. Bands of reflectors, interpreted as representing major lithological boundaries, could usually be followed more easily than individual reflecting events.
The shallow reflectors can be correlated to the Skookumchuck near the Morton anticline, based on surface exposures and a 2,000 ft stratigraphic test drilled in 1983.
Deeper reflectors persist below this band for as much as 3 sec of total travel time. These reflectors are interpreted as representing the marine McIntosh formation and older, unknown sediments.
Most of the faults at Morton and elsewhere along the profile appear to be high-angle and normal, consistent with the extensional style seen more clearly in the Chehalis basin. They are shown, somewhat schematically, as bounding faults to a graben forming the syncline near Morton. In reality, though, the fault pattern is likely to be more complex than shown on the line drawing. An area of active dextral movement, the St. Helens Seismic Zone, extends from Mt. St. Helens through the western limb of the Morton syncline.12
To the east of the Morton anticline, the shallower data can be correlated from surface exposures to the volcanic Oligocene Ohanopocosh. A distinctive reflector pattern at the broad down warp near Mt. Rainier correlates to a Miocene aged interbedded basalt/lacustrine sequence, filling a broad syncline interpreted as the upper part of a large graben or crustal downwarp associated with Mt. Rainier.
The deeper reflective bands can be followed throughout this section and are generally parallel to the shallower events, indicating that a relatively, simple structural style probably existed before being distorted by significant Miocene movement. However, no readily identifiable basement reflector is seen throughout the SWCC.
SOURCE ROCK MATURITY
Source rocks within the study area have a complicated pattern of maturity characteristics.
The late-Eocene sediments drilled to date have usually been gas prone, thermally immature rocks with low maturity/depth values, as found in the Shell 1 Thompson. However, the total organic carbon content of the Skookumchuck and Lincoln Creek formations in the Thompson well is good, above 1%, and suggests they could act as source rocks in a better thermal environment.
The kerogen type is exclusively the gas-prone, herbaceous Type III within the Skookumchuck formation.
The 1 Thompson well is similar to wells drilled in other forearc basins in its maturity characteristics. However, the Humble 1 Perry well, located 5 miles to the northeast, discovered what appear to be two distinct thermal trends over its 10,685 ft depth.
Mature rocks within the oil window are found at less than 2,000 ft before shifting to a maturity/depth trend similar to that of the 1 Thompson below an 1,100 ft thick layer of volcanically derived rocks, the Goble formation.
The 1 Everett Trustee and 1 Rosa Meyer wells also found early mature rocks at less than 4,000 ft and measured much steeper maturity/depth gradients than found in the Thompson well and the lower parts of the Perry well.
Additionally, a 2,000 ft coal test hole near the crest of the Morton anticline reportedly found very mature rocks in the main gas generation phase. These characteristics are seen graphically (Fig. 6) on the cross section A-A'.
This anomalous thermal behavior appears related to groundwater movement above 2,500 ft, which carried heat westward from the Cascades and from more localized sources, that produced a separate, shallow geothermal regime.
The dual nature of the geothermal gradients makes maturity studies difficult, requiring additional knowledge of both the regional thermal gradient of the deeper sediments and the effects of the interaction between the shallow and deep mechanisms for valid conclusions.
However, any affect on the maturity of the SWCC sediments would more likely be favorable than detrimental to their hydrocarbon potential. While as yet unproven, the deep marine sediments of the McIntosh and older formations appear to be within the oil generative window and would be likely to have organic material with more attractive kerogen types than have been drilled to date.
DISCUSSION
Although pessimistic in some ways, the data from the Thompson and other nearby wells, when combined with the surface information, maturity studies, and the DOE seismic data, hint at additional potential.
For example, considerable organic material is found in the fluvial/deltaic Skookumchuck formation, and suitable organic content should also be expected in the marine McIntosh formation and older sediments. In the right environment, either could act as hydrocarbon source rock.
Clean sandstones that could be suitable reservoir rocks are found at the McIntosh outcrops and from a high porosity (30%) sandstone within the Skookumchuck formation, utilized as a reservoir for Jackson Prairie gas storage field in the Chehalis basin.
Isolated pockets of higher porosity sands are found in the productive Cowlitz formation, equivalent to the Skookumchuck, at Mist gas field. These may be related to more energetic current flows in restricted seaways as the Siletzia terrane approached the continental margin and potentially could have analogies in the Skookumchuck formation.
Also, maturity trends from bottomhole temperatures and vitrinite reflectance values indicate a thermal gradient trend that increases eastward from the Thompson well. Reliability of this trend is questionable, though, because of the few borehole tests and the shallow geothermal trend.
The shallower, stratified reflectors seen throughout the seismic profile can be correlated to the late-Eocene found in the Chehalis basin, using the outcrops near Morton as reference points. Potential structural traps seem common in these sediments, and stratigraphic traps might be equally abundant.
The regional dip of these reflectors suggests that they may continue under the Cascades to the Eocene sediments tested on the Columbia River Plateau at 15,000-17,000 ft in depth. Since the correlation is interrupted primarily by the Miocene-aged Cascades, this correlation seems plausible.
If these Eocene sediments extend under the Cascades, they would form a large, essentially untested sedimentary province that could have significant hydrocarbon potential. If the province is continuous, the nearly commercial gas shows from the Eocene sediments at the Columbia River plateau wells may indicate that similar source and reservoir rocks could be found at shallower depths west of the Cascades.
Realistically, though, well control is too sparse to project the basic parameters of source, reservoir, and thermal characteristics from the Thompson well to the SWCC. This makes the present investigation of the SWCC heavily dependent on geophysical data and surface information. Furthermore, the effects of the later Miocene intrusives and volcanism on any hydrocarbon accumulations are unknown.
CONCLUSION
Both the seismic reflection interpretation and maturity studies indicate that the study area contains sediments potentially favorable for hydrocarbon accumulations.
Mist gas field and the shows found throughout the region indicate that hydrocarbon generation has occurred in the late Eocene sediments, but this study suggests that the best potential of the region may well be in the older, untested rocks.
The geologic model suggested by this study indicates an extremely dynamic system that has evolved significantly over time, with complex interactions between structural formation and deposition.
The exploration uncertainties involved in so complicated a region make this a high risk area. However, it is one of the few untested areas in the lower 48 with good potential. For a company willing to accept the risks, the Pacific Northwest represents a frontier area in the truest sense.
ACKNOWLEDGMENTS
I expressly thank my main partner in this effort, Mary Guide, and the later efforts of Mark Thomas. They would both be co-authors of this paper in more fair times. I appreciate the encouragement and advice of Gary Latham with this project.
I also thank Hugo Pulju of Lauren Geophysical, Denver, for permission to use the line drawing of A2-3 for this paper. Many other first rate people contributed to the project, specifically Hank Shasse and William Lingley, Washington State Department of Natural Resources; Dan Hollis, Geophysical Systems; Rick Steinick, Golden Geophysical; John Contino, Geotrace Technologies; Peter Hales, Weyerhauser Minerals; and Tom Wilson, West Virginia University. I wish we had more time to have spent with both Steve Pappajohn and Tom Ise, independent geologists in Washington, but appreciated their superior expertise in this area and the work they did for this project. Their contributions were essential.
REFERENCES
1. Stanley, W.D., Tectonic Study of the Cascade Range and Columbia Plateau Based Upon Magnetotelluric Soundings, Journal of Geophysical Research, Vol. 89, p. 4,447.
2. Stanley, W.D., Gwilliam, W.D., Latham, Gary, The southern Washington Cascades, a previously unrecognized thick sedimentary sequence?, AAPG Bull., Vol. 76, No. 10, 1992, pp. 1,569-85.
3. Wells, Ray E., and Paul L. Heller, The relative contribution of accretion, shear, and extension to Cenozoic tectonic rotation in the Pacific Northwest, GSA Bull., Vol. 100, 1988, p. 325.
4. Mooney, Walter D., and Weaver, Craig S., Regional crustal structure and tectonics of the Pacific States: California, Oregon and Washington, GSA Memoir 172, 1989, p. 129.
5. Harding, T.P., Identification of Wrench Faults Using Subsurface Structural Data: Criteria and Pitfalls, AAPG Bull., Vol. 74, 1990, P. 1,590.
6. Harding, T.P., Identification of Wrench Faults Using Subsurface Structural Data: Criteria and Pitfalls: reply, AAPG Bull., Vol. 75, 1991, P. 1,786.
7. Park, R.G., Geological Structures and Moving Plates, Chapman & Hall, New York, 1988.
8. Ben-Avraham, Z.A. Nur, D. Jones, A. Cox, Continental Accretion: From Oceanic Plateaus to Allochthonous Terranes, Science, Vol. 213, p. 47.
9. Duncan, R.A., A Captured Island Chain in the Coast Range of Oregon and Washington, Journal of Geophysical Research, Vol. 87, p. 827.
10. Ise, F.T., Washington and Oregon-Are There Other Rocks to Explore, OGJ, Aug. 12, 1985, p. 112.
11. Jarchow, Craig M., Catchings, Rufus D., and Lutter, William L., How Washington crew got good, thrifty seismic in bad data area, OGJ, June 17, 1991, p. 54.
12. Weaver, Craig S., Grant, Wendy C., and Shemeta, Julia E., Local Crustal Extension at Mt. St. Helens, Wash., Journal of Geophysical Research, Vol. 92, p. 10,170.
Copyright 1993 Oil & Gas Journal. All Rights Reserved.
