3

AN EVALUATION OF THE DEPARTMENT OF THE INTERIOR'S 1989 RESOURCE ASSESSMENT

As Chapter 1 described, the Department of the Interior's 1989 resource assessment caused concern among some petroleum industry representatives. The assessment projected significantly lower undiscovered resource values than two prior DOI assessments, issued in 1975 and 1981. Certainly, part of the decrease was due to discoveries made between assessments. Once an exploratory drill uncovers a petroleum field, the field moves from the “undiscovered” category to the category of “identified” resources awaiting extraction. But, more important, the estimates from the different years may have varied because of limitations of the assessment methodology.

This chapter presents the Committee on Undiscovered Oil and Gas Resources ' evaluation of the methodology used in the 1989 assessment. It begins with a brief overview of the assessment boundaries, important for understanding the assessment results. Then, it moves to a two-part evaluation of the assessment methodology: the first part covers the USGS; the second part covers the MMS. (Recall from Chapter 1 that the USGS inventoried resources for oil-producing regions onshore and in state waters, while the MMS assessed resources beneath the Outer Continental Shelf.)

ASSESSMENT BOUNDARIES

Understanding the assessment boundaries—which oil and gas resources the assessment includes and which it excludes—is crucial for policymakers who are formulating energy strategy and attempting to gain a comprehensive view of future oil and gas resources available for development. This section evaluates



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures 3 AN EVALUATION OF THE DEPARTMENT OF THE INTERIOR'S 1989 RESOURCE ASSESSMENT As Chapter 1 described, the Department of the Interior's 1989 resource assessment caused concern among some petroleum industry representatives. The assessment projected significantly lower undiscovered resource values than two prior DOI assessments, issued in 1975 and 1981. Certainly, part of the decrease was due to discoveries made between assessments. Once an exploratory drill uncovers a petroleum field, the field moves from the “undiscovered” category to the category of “identified” resources awaiting extraction. But, more important, the estimates from the different years may have varied because of limitations of the assessment methodology. This chapter presents the Committee on Undiscovered Oil and Gas Resources ' evaluation of the methodology used in the 1989 assessment. It begins with a brief overview of the assessment boundaries, important for understanding the assessment results. Then, it moves to a two-part evaluation of the assessment methodology: the first part covers the USGS; the second part covers the MMS. (Recall from Chapter 1 that the USGS inventoried resources for oil-producing regions onshore and in state waters, while the MMS assessed resources beneath the Outer Continental Shelf.) ASSESSMENT BOUNDARIES Understanding the assessment boundaries—which oil and gas resources the assessment includes and which it excludes—is crucial for policymakers who are formulating energy strategy and attempting to gain a comprehensive view of future oil and gas resources available for development. This section evaluates

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures three key parameters related to the assessment boundaries: the definitions of conventional and unconventional resources, the economic assumptions, and the reserve growth of existing petroleum reservoirs. Conventional/Unconventional Boundary One controversial aspect of the DOI's assessment is its limitation to “conventional” resources. As Chapter 1 explained, industry groups—most notably the Potential Gas Committee—do not share the DOI's delineation between conventional and unconventional resources and include in their own assessments certain resources that the DOI defined as outside of the assessment boundaries. The DOI assessment defined conventional resources as “crude oil, natural gas, and natural gas liquids that exist in conventional reservoirs or in a fluid state amenable to extraction techniques employed in traditional development practices” (U.S. Department of the Interior, 1989). Further, the assessment specified that these resources must exist in “discrete accumulations.” The USGS, which was responsible for the onshore assessment that covered the bulk of unconventional resources, acknowledged that the boundary between conventional and unconventional resources is hazy in many situations. The DOI's reasons for setting aside unconventional reservoirs from analysis were: (1) they are mostly discovered and their geographic and stratigraphic distributions are largely known; and (2) in-place volumes are large and recoverability is uncertain because of technical and economic factors (U.S. Department of the Interior, 1989). Further, the DOI judged that the geologic data and assessment methods available for evaluating unconventional resource occurrences were inadequate. In the committee's opinion, the DOI's decision to exclude unconventional oil resources from its assessment is justifiable. Except for thermally stimulated heavy oil recovered primarily in California, unconventional oil deposits, like those from tar sand and oil shale, have contributed little to domestic oil production. Indeed, current and future production costs of these resources will be higher than costs of conventional resources. On the other hand, the DOI's decision to exclude unconventional natural gas resources presents problems. Natural gas deposits in low-permeability sandstones, fractured shale, and coal beds—sources the DOI labelled “unconventional”—are making an important contribution to current production, variously estimated at about 1.5 to 2 trillion cubic feet (Tcf) per year. Much of

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures this volume is produced from low-permeability gas-bearing sandstones and is primarily the result of the accelerated drilling and development that occurred between 1977 and 1982, especially between 1979 and 1982 under incentives from the Natural Gas Policy Act. Production of coal-bed methane and natural gas from fractured shale have been increasing in recent years, with the former supported by increased understanding of the resource's distribution, new production technologies, and tax incentives. The recent (December, 1989) extension of coal-bed methane tax credits and the reimplementation of tax credits for tight-gas reservoirs for two years will assist further development of these resources. Production of all major forms of unconventional natural gas has been aided by extraction research funded by the Gas Research Institute and the U.S. Department of Energy. While acknowledging some of the special characteristics of unconventional natural gas accumulations, the committee concludes that it is both feasible and appropriate to include these resources in a national resource assessment.1 Of course, the estimates of unconventional resources should be provided separately from estimates of conventional resources. Industry decisions about whether to produce unconventional resources are guided by a unique set of technical characteristics and economic incentives (in the form of tax credits). Thus, estimates of unconventional resources are most useful when they are provided separately, because future production of these resources may depend upon economic and technical conditions different from the conditions that influence the production of conventional resources. In-Place Resources: Recoverable/Unrecoverable Boundary As discussed in Chapter 2, the classification of resources as “recoverable” or “unrecoverable” with current technology is at best a snapshot in time. Advances in reservoir characterization, drilling, and completion continually increase the percentage of a reservoir's total petroleum supply that is recovered. Yet, the USGS/MMS assessment (like most other assessments) included no estimates of in-place resources, only estimates of resources recoverable with 1   In this chapter, the committee's conclusions and recommendations are underlined.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures current technology. Estimates of in-place resources are important in judging the potential for new recovery technology to increase the producible petroleum supply. Because the assessment was limited to recoverable resources, on average only about one-third of discovered oil and 60 to 85 percent of discovered natural gas was incorporated in the models used to estimate undiscovered resource volumes. Excluding consideration of in-place resources in these models imposes a technological overlay too early in the assessment process. A more thorough evaluation would first estimate each play 's in-place resources and then multiply these estimates by recovery efficiencies to estimate technically recoverable resources. Using estimates of in-place resources in combination with recovery efficiencies would allow the estimates to be updated more easily to account for advances in recovery technology. The committee recommends that in future assessments, the USGS and MMS develop methods to estimate in-place resources in each play. The agencies should base their estimates of technically recoverable resources on in-place resources, by applying recovery factors to in-place resource estimates . Economic Boundaries The assessment's economic boundaries hinged on the computation of a “minimum economic field size” (MEFS): the smallest field that assessors determined can be developed profitably. Calculating the MEFS requires a myriad of economic assumptions. For the 1989 assessment, the fundamental assumptions were: The MEFS must yield a net average after-tax return of 8 percent on the development investment. The absolute price of natural gas would not exceed 75 percent of the btu-equivalent price of crude oil. The 1987 oil price was $18.00 per barrel. The 1987 natural gas price was $1.80 per million cubic feet (Mcf). Real oil prices would decline at an annual rate of 3 percent between 1987 and 1989 and then increase 4 percent (with a range of plus or minus 1 percent) per year beginning in 1990. Real gas prices would decline 2 percent annually between 1987 and 1989 and then increase 5.5 percent (with a range of plus or minus 1 percent) per year beginning in 1990.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures The future costs for reservoir development and transportation facilities would remain fixed at 1985-1986 levels. With these assumptions, DOI analysts calculated the MEFS on the basis of a 1987 field discovery and a decision to develop the field beginning in that year. The USGS and the MMS used the same assumptions in determining the MEFS. However, because onshore and offshore assessments differ in their overall approach and methodology, the USGS and MMS used different boundaries to separate the economically recoverable resources from the subeconomic resources. USGS Economic Boundaries In its analysis of the lower 48 states onshore and the offshore state waters, the USGS defined plays using field sizes down to 1 million barrels of crude oil or the energy-equivalent volume of natural gas (denoted as Mmboe; 1 Mmboe is equal to 6 billion cubic feet (Bcf) of gas). The 1 Mmboe volume was used to truncate the geologically assessed distribution of undiscovered fields and to provide a lower boundary for the play assessment methodology (Attanasi, 1988). In part, this truncation was a function of the minimum field size included in the national data base used as input for the assessment. For the lower 48 states onshore, the USGS found that virtually all fields of 1 Mmboe or larger were economic to develop, based on depth and field size. However, for the assessment of fields smaller than 1 Mmboe, the USGS calculated the MEFS by province or region and then determined what fraction of these small fields was economic to develop. The calculation of the MEFS was done by discounted cash-flow analysis using assumed future oil and gas prices, inflation rates, rates of return, development costs, and timing of development (U.S. Department of the Interior, 1989). Estimates of development costs accounted for drilling depth, especially for small fields, and water depth for state waters (Attanasi, 1988). Overall, only 1 in about 2,100 fields containing more than 1 Mmboe was found to be uneconomic to develop in the onshore lower 48 states. Fifty-two percent of 50,241 undiscovered oil fields and 47 percent of 27,014 nonassociated gas fields smaller than 1 Mmboe were considered economic to develop (Attanasi, 1988). The committee concludes that since tens of thousands of undiscovered small fields were considered economical to develop, the chosen minimum size cutoff of 1 Mmboe was too high for the overall assessment.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures It is important to note that for both Alaska and state waters in the Gulf and Pacific, where development is more expensive, the USGS used a higher MEFS. For the Gulf and Pacific, the USGS assumed a minimum economically developable crude oil field size of 2 Mmboe. For Alaska, economic constraints were even greater. For the overall assessment, in moving from technically recoverable to economically recoverable resources, applying the MEFS and the economic parameters (most notably the $18.00 per barrel crude oil price and the $1.80 per Mcf natural gas price) reduced the undiscovered resources onshore in the lower 48 states by only a small amount. Nationally, for crude oil, there was a 20 percent reduction (6.7 billion barrels) between onshore mean undiscovered recoverable resources and onshore mean undiscovered economically recoverable resources. Of that 6.7 billion barrels, however, 5.3 billion barrels were in the Alaskan region. For natural gas, there was a 26 percent reduction (65.3 Tcf) between onshore mean undiscovered recoverable resources and onshore mean undiscovered economically recoverable resources. Of that 65.3 Tcf, 56.7 Tcf were in the Alaskan region. Thus, for the onshore lower 48 states, the economic overlay produced only a small constraint to the undiscovered resource volumes. The important implication of these results—an implication policymakers must recognize—is that higher oil prices will NOT by themselves transform large volumes of undiscovered resources from technically recoverable to economically recoverable, except in Alaska. Assuming that the DOI assessment is correct, a higher oil price may speed recovery of these resources by offering a higher profit, but it will not lead to much higher total (ultimate) recovery unless the price increases stimulate development of technological advances that expand the boundaries of technical recoverability (for example, to deeper waters). We note, however, that the estimation of technically recoverable resources in the USGS assessment used an economic screen (for example, a minimum field size). The use of such a screen is contrary to the meaning of the term “technically recoverable” and implies that some high-cost but recoverable resources were left out of the assessment. MMS Economic Boundaries For its resource evaluation, the MMS divided the OCS into “planning areas.” Each planning area consisted of one or more basins. Each basin contained one or more groups of identified or postulated reservoirs that the MMS called “plays,” although these plays were actually summations of prospects

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures and not true, geologically defined plays. The MMS used a mathematical computer model, PRESTO (Probabilistic Resource Estimates—Offshore), to produce its final resource assessments. The PRESTO model produces a range of resource estimates with a corresponding estimate of the probability of occurrence. It simulates drilling of the modeled prospect and hydrocarbon discovery. MMS analysts derived probability distributions of resource volumes from multiple runs of the model. PRESTO determines the decision to develop, which would involve platform and production well installation, by comparing the prospect resources with a MEFS. For the 1989 assessment, the MEFS was determined outside the PRESTO model with a model called MONTCAR, a discounted cash-flow analysis program that calculates the volume of resources needed to balance various operating costs. In calculating operating costs, MONTCAR considers water depths, drilling depths, distance from shore, and other operating conditions for each prospect (Minerals Management Service, 1985). This results in a range of costs for different areas. The MMS's calculation of the MEFS was prospect specific. It did not incorporate costs associated with production infrastructure, such as pipelines and onshore processing facilities, that might be shared between multiple discoveries. If, on a specific trial of the PRESTO model, the computed resources exceeded the MEFS, the model stored the results for developing the final range of outcomes. If the resources for that prospect were less than the MEFS, the model set the resource volumes to zero for that trial. The model also calculated minimum basin reserves (MBR) and minimum area reserves (MAR) to determine if the aggregate estimated undiscovered resources were adequate to justify required transportation and plant facilities (U.S. Department of the Interior, 1988). The MEFS for the base-case economic scenario ($18.00 per barrel of oil and $1.80 per Mcf of gas) ranged from 3 Mmboe in the Gulf of Mexico and the Pacific, to 5 Mmboe in the Atlantic, to between 44 and 299 Mmbo in Alaska (depending on location). The maximum MEFS ranged from 190 Mmboe in the Pacific to 690 Mmboe in the Gulf of Mexico, and from 300 Mmboe in the Bering Sea of Alaska to 1,000 Mmboe in the Atlantic. MAR volumes ranged from zero in the Gulf and Pacific (that is, infrastructure is already available in these areas), to 120 Mmboe in the Atlantic, to between 77 and 810 Mmbo in Alaska (U.S. Department of the Interior, 1988). For regions with an established producing infrastructure, the MEFS and MAR had little effect on development of the undiscovered resource. In more remote or severe operating portions of producing regions, or in regions where no production has been established,

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures economic boundaries were present. Such boundaries were significant where costs of building infrastructure are high, as in remote, deep-water areas and other places where operating conditions are difficult. Economic boundaries were moderate for the Atlantic region but were much more significant for Alaskan waters because the cost of building infrastructure there is so high. Inferred Reserves and Reserve Growth Though the DOI assessment focused on petroleum from undiscovered reservoirs, it also included a separately reported estimate of inferred reserves (the postulated incremental but unknown volume estimated to be recoverable from known reservoirs). Inferred reserves are an important component of the resource base. The DOI assessment, for example, estimated that inferred oil reserves are 63 percent as large as undiscovered oil reserves, based on mean values. Therefore, the accuracy of the method used to estimate inferred reserves is an important issue. The DOI estimates of inferred reserves were based on a statistical time series of ultimate recovery by year of field discovery that ended in 1979. This time series was compiled and published by the American Petroleum Institute (API) and the American Gas Association (AGA). It captures historical and traditional reserve growth sources, such as extensions and new pools. The time series data reflect drilling experience probably no more recent than 1977. The period since 1977 has seen a substantially increased understanding of reservoir heterogeneity, a greater understanding of the consequences of sweep efficiency in waterfloods, and the potential for strategically targeting infill drilling and recompletions. As a result, the data through 1977 on reported inferred reserves do not reflect the increased knowledge of reservoirs gained in more than ten years of drilling. At the time of the assessment, there was no way to avoid this shortcoming. However, since the assessment was completed, the Energy Information Administration (EIA) has prepared an evaluation of oil and gas reserves by year of discovery that could help document the increasing efficiency in the conversion of discovered resources to reserves (Energy Information Administration, 1990). The estimated ultimate recovery (EUR) of oil and natural gas was compiled for six dates between 1977 and 1988 by year of discovery in groups of five years each. These new data can be used, for example, to show that for Railroad Commission District 8 in Texas—a leading oil-producing district—the annual average increase in EUR for 1977-1988 was 1.5

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures BOX 3.1 The USGS'S Legislative Mandate The U.S. Geological Survey has deep roots in the monitoring of the nation's natural resources. In 1879, as the nation pushed its boundaries west, Congress, at the recommendation of the National Academy of Sciences, created the USGS to map the new territories. Congress mandated in the Organic Act of 1879 that the USGS oversee “classification of public lands and examination of the geological structure, mineral resources and products of the national domain.” In response, the new agency dispatched teams of scientists on horseback to document the west's natural resources. Since 1879, the USGS's mission to research energy resources in various parts of the nation has evolved through a patchwork of federal laws, including: Public Law 29, passed in 1935, and the Appropriations Act of Fiscal Year 1959 mandated that the USGS extend its mineral resource investigations to Puerto Rico, Antarctica, and the Trust Territory of the Pacific Islands. The Wilderness Act of 1964 requested that the USGS assess mineral resources of areas proposed or established as wilderness sites. The Alaska National Interest Lands Conservation Act of 1980 required that the USGS assess the oil and gas potential for federal lands in Alaska. Although the USGS is responsible for developing and disseminating the geologic information needed to help formulate policy and ensure the wise development of the nation's energy resources, there is no provision in the federal statutes for a comprehensive national program to inventory oil and gas (National Research Council, 1988).

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures percent for fields discovered between 1920 and 1934. This was an increase from a rate of 1.0 percent in the period 1971-1977 for those same discoveries. The DOI should determine if the EIA's new data offer a method of updating previously inadequate recognition of reserve growth potential of known heterogeneous reservoirs. The DOI should also determine whether use of these new data are suitable for, and will continue to be available for, an improved inferred reserve assessment methodology. If these data are not appropriate or will not continue to be updated, other methods to define reserve growth potential should be developed. DETAILED EVALUATION OF USGS ASSESSMENT METHODS The USGS has the responsibility to assess the oil and gas resources of the onshore portions of the United States (see Box 3.1). Its jurisdiction includes the submerged lands contiguous to the coastal states to a distance of three miles from their coasts. This section evaluates the quality of the USGS assessment for these areas. Organization and Staff Oil and gas resource assessments carried out by the USGS are the responsibility of the Office of Energy and Marine Geology in the Geologic Division. Two branches, headquartered in Denver, provide the staff. For the 1989 assessment, the USGS divided the nation into 9 regions and subdivided the regions into 80 provinces (see Figure 3.1). The provinces were assigned to 42 geologists from either the Petroleum Geology Branch or the Sedimentary Processes Branch. Following procedures set forth by an assessment coordinating committee, each geologist carried out the assessment for his or her province. Each of the 42 province geologists was responsible for preparing a report on the geology and oil and gas plays in each province. (The committee found, however, that some of these reports were not prepared and completed as open-file reports until after the assessment was finished.) Work was reviewed by the assessment coordinating committee, which made the final decisions about whether to use or modify the data provided by province geologists. The USGS reported that each of the 42 province geologists spent not more than half time on the assessment over a two-year period. The project

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures FIGURE 3.1 The nine regions into which the DOI divided the nation for purposes of assessing undiscovered oil and gas. leader was the only USGS staff member who spent almost full time on the assessment. The committee found that experienced geoscience personnel were distributed unevenly among the provinces, and that their distribution did not correlate with the importance of the provinces as producers of oil and gas or their potential for containing undiscovered resources. Current research interests of participating geologists rather than the resource assessment appeared to dictate staff allocations. This resulted in a concentration of attention on the Rocky Mountains, California, and Alaska and a comparatively low level of effort in the Gulf Coast, Midcontinent, and Illinois Basin areas. In the lower 48 states,

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures FIGURE 3.4 Form B, the second form that MMS assessors used to record PRESTO input parameters in Alaska.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures judges that it would be more appropriate to relate vertical fill-up to vertical closure and area of closure, as well as to productive acreage. Further, in defining input parameters like vertical fill-up, the MMS should develop better guidelines to allow geoscientists' judgments to be flexible by play or province instead of imposing region-wide, rigid directives. The second problematic aspect of the procedure was the imposition of arbitrary productive acre limits, apparently to prevent estimates from becoming “too high.” In the Chukchi Sea, for example, the lowest enclosing contours of the prospect included 96,000 acres. The numbers used in the resource evaluation were: minimum, 9,600 acres; most likely, 19,200 acres; maximum, 28,800 acres. Thus, acreage sizes used were 10 percent, 20 percent, and 30 percent of the acreage within the lowest enclosing contour. A potential reserve generated by the “ most likely” values used for this prospect would be obtained from the following computation: 19,200 acres × 310 barrels/acre ft. × 200 ft. = 1.190 Bbo (Two-hundred feet is the thickness of the zone.) By contrast, if a “most likely” acreage were 50 percent of the total closure, or 48,000 acres, the comparable reserve would equal 2.976 Bbo (using the same recovery factor and net pay). The committee concludes that while conservative judgments about productive acreage may be appropriate for lease sale tract evaluation purposes, they are not appropriate when applied to resource assessments designed to evaluate undiscovered oil and gas in a play. Probabilistic Dependence The MMS's assessment procedure, like the USGS's procedure, did not incorporate a way to address possible probabilistic dependencies between uncertain quantities. In the Atlantic offshore region, for example, the MMS used five generic play types for the 1989 assessment: plays underlying rifts (Triassic grabens), plays along the shelf edge (Jurassic carbonates), sedimentary pinch-outs seaward of the Jurassic carbonates, lower Cretaceous drapes over basement highs seaward of the Jurassic carbonates, and faulted anticlines on the continental shelf. Seismic profiles indicate that structures that are well-defined at depth “die out” and porosity increases upward from the Triassic strata. The implication is that structure, size, and porosity are negatively correlated as functions of depth. In addition, structural relief and extent appear to be correlated with depth. These empirical findings suggest that an assessment procedure in which all

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures uncertain quantities assessed are assumed to be probabilistically independent will misrepresent an essential feature of this particular depositional environment. In particular, if structure quality is depth dependent, prospect risk must also be depth dependent. The committee recommends that the MMS modify its assessment procedure to allow the incorporation of probabilistic dependencies between uncertain geophysical quantities like structure, size, porosity, and depth. MMS assessment procedures were based on the assumption that individual prospect outcomes are mutually independent conditional on the presence of at least one field. Careful study of the statistics of well-explored plays will clarify whether or not such an assumption is reasonable. No such studies have, to our knowledge, been completed. Because of the nature of risking, assuming independence can tend to increase the estimated risk of failure to find recoverable resources. For example, the committee reviewed a case in the Pacific OCS that we believed exhibited dependency of risk. The MMS estimated the individual chance of success (1.0 minus the percent risk of failure) to be .73 based on an assumption of complete independence. According to the committee's calculations, assuming high dependency among zones would have yielded a .85 chance of success (see Appendix A). This .12 difference in the success chance when prospect outcomes are assumed dependent instead of independent can increase the estimated recoverable resources by many millions of barrels. Thus, a tendency to prefer an assumption of independence of risk will tend to result in overly conservative resource estimates. Consequently, it is critical that the MMS (and the USGS) refine its current risk assessment practice. The committee recommends that the MMS undertake a study to determine if a pattern of dependencies is present in observed sequences of prospect drilling outcomes. The MMS should redesign assessment methods and forms to incorporate dependencies where present. Assumptions As discussed previously, the MMS's resource assessment method relied on many decisions regarding the selection of input parameters entered into PRESTO. Many input parameters can be determined from the general data base, experience, and reasonable projection or extrapolation. In areas with a long history of exploration and production, the range of choices for input parameters can often be well defined and fairly narrow in span. In frontier areas, however, the range of choices can be extreme and very difficult to establish. MMS geoscientists and engineers for the most part appear to have used good

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures judgment in selecting PRESTO input parameters. Some specific instances, however, require examination and potential revision. In Alaska, a recovery factor of 40 percent for oil was applied universally in the 1989 assessment with secondary recovery initiated at the inception of production. Because of the lack of federal OCS production history in Alaska, this assumption may well be a reasonable generality. It does, however, conflict with the USGS's assumption of a 32.3 percent recovery factor on the North Slope. The committee recommends that as the data base expands, the MMS should shift to the use of play-oriented recovery factors related to identifiable lithologies, depths, and overall reservoir characteristics. In the Pacific OCS region, the recovery factors the MMS used for the Monterey reservoirs were based upon gross formation thickness, using a barrels-per-acre-foot recovery from analogs from productive Monterey fields. The committee advises that the MMS attempt to define more precise methods, based upon net pay, for determining recovery factors. In fairness, it should be realized that industry now uses similar gross thickness calculations or barrels-per-acre analogy for reserve estimates. Nevertheless, continued study and review could possibly improve methods for future assessments. In Alaska, the MMS applied an arbitrary area cut-off so that prospects less than one-half a leasing block were not included. They similarly excluded prospective reservoirs with less than 100 feet of net pay and reservoirs occurring at depths of less than 3,000 feet. Prospects judged to be subeconomic (if appropriately risked) were not modelled in the conditional case. These four exclusions are assumptions based upon economic screens that the committee considers to be inappropriate in determining the oil and gas resource endowment. The committee recommends that the MMS develop methods for separating technically recoverable resource calculations from those that determine economic resources, which is difficult with current PRESTO methodology. Economic Input The committee generally viewed the MMS's economic parameters as reasonable based on current industry data and practices. Likewise, the MMS's assumptions regarding minimum economic field sizes appeared reasonable and well defined in the various operating areas. The determination of the MEFS for prospects in different plays, water depths, operating conditions, producing or frontier areas, and drilling depths is a major component of the PRESTO

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures program. In Alaska, for example, in areas of difficult operating conditions, such as in the Beaufort and Chuchki Seas, minimum field sizes were appropriately escalated to account for expensive production and transportation infrastructure requirements, and separate calculations were made for “satellite” versus “standalone” fields. Future assessments will include producing environments in even deeper water and possibly more remote areas and hazardous operating conditions. Future technological developments and innovative technologies for drilling and producing operations will require even more economic screening models to present the available options. Continued MMS efforts in the engineering and economic modelling and data gathering programs should be encouraged to insure continued high-quality results. Uncertainties In frontier areas with little or no established commercial production, the uncertainties in input parameters, as well as in resulting estimates, are very large. For example, the 1989 assessment indicated that the mean value of the Alaska OCS recoverable oil resource is 3.4 billion barrels. This single number may be misleading if not compared to the range of possible values, as indicated by the F95 and F5 fractiles. (The F95 fractile is the petroleum volume for which there is a 95 percent probability that the area contains more than that volume; the F5 fractile is defined analogously.) In Alaska, this range is very broad: 0.6 to 9.4 billion barrels. The committee is concerned that the MMS's method of reporting resource values focused too much on mean values. In reporting future assessments, the MMS should emphasize more the role uncertainty plays in resource estimates and use more graphic displays to demonstrate visually the ranges of uncertainty. This type of reporting would assist the non-technical user in understanding the process and results, and possibly offset the tendency to focus only on the single number, the mean, which is now most often reported and used. Statistical Methods Number of Undiscovered Fields MMS assessors obtained distributions showing the number of undiscovered fields by probabilistic sampling and economic screening of identified prospects.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures The identified prospects had specific geographic locations, risk factors, physical information relating to field-size distribution, and economic screening information. In addition to identified prospects, the MMS allowed for the inclusion of unidentified prospects in each basin. The number and field-size parameters of unidentified prospects were extrapolated from the statistics of identified prospects. (As mentioned previously, it appears that unidentified prospects were included infrequently.) The MMS assessment procedure converted a prospect to the inventory of undiscovered fields by a coin toss, with conversion probability equal to one minus the risk factor for that prospect. Prospects were risked independently of one another, given that at least one prospect in the basin would be converted. The probability that no prospect in a basin would be converted was taken to be the basin risk. This basin risk implicitly introduced a degree of dependence among prospects, discussed later. If the prospect survived the coin toss, then a field size was assigned. The field was retained in the inventory only if its assigned size exceeded a minimum economic size screen specific to that prospect. The economic screen was set to 1 Mmboe for all prospects for “undiscovered recoverable resources” estimation (as opposed to “undiscovered economically recoverable resources ” estimation). The total number of undiscovered fields in a basin thus became a random fraction of the total number of identified and unidentified prospects. The MMS made no systematic attempt to calculate the distribution of the number of prospects that the seismic grid missed, which could be considerable even in basins with apparently extensive exploration. Likewise, the MMS did not estimate the number of prospects that were crossed by the seismic grid, but were not identified. Whenever unidentified prospects were included, their size distribution was such that they could never survive the economic screening. This is one more piece of evidence pointing to the need for the increased consideration of conceptual plays, discussed earlier in this chapter. In contrast to the MMS, the USGS did not use concepts of specifically located and specifically described putative fields. For each defined play, the USGS generated multiple undiscovered fields by sampling repeatedly from a fixed, play-specific, field-size distribution that used a fixed economic cut off. This was the USGS's finest level of specificity; it is analogous to the MMS's use of unidentified prospects. The committee recommends that the MMS method for assessing unidentified prospects be brought into line with the USGS 's procedures for extrapolating discovery histories.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures Sizes of Undiscovered Fields If a prospect survived the MMS risk screening and became an undiscovered field, it was assigned a field size. This field size was generated as a product of area, thickness, and recovery numbers, each randomly and independently drawn from distributions specific to that prospect. The attempt to break down the field size into more easily described physical quantities may seem reasonable. It allowed for separate assignment of distributions for the input area, thickness, and recovery of the selected prospect. However, one then needs to put these numbers back together somehow and this was done by assuming statistical independence of the three inputs. The committee recommends that the MMS compare field-size distributions produced under the assumption that area, thickness, and recovery values are independent with empirical field-size data. This comparison will show whether independence of field-size parameters is a reasonable assumption. MMS assessors typically specified probability distributions for quantities like thickness on prospect evaluation forms that differed somewhat from region to region. Evaluators familiar with the prospect were asked to specify minimum, most probable, and maximum values for thickness. Through an undocumented procedure, these three values were turned into one of several possible distributions. It appears that “minimum ” and “maximum” were to be interpreted as the lower and upper 5 percentage points of the distribution and “most probable” as the mean value. As discussed earlier, it is not clear whether individual evaluators were aware of these interpretations. Geologic Risk Factors Not all zones or prospects are undiscovered fields. The MMS assessed this uncertainty through a hierarchy of risk numbers associated with the geologic hierarchy of zones, prospects, basins, and areas. The procedure can be illustrated by considering the assignment of prospect and basin risk numbers. The prospect risk is defined as one minus the probability that the prospect is an undiscovered field. The basin risk is one minus the probability that the basin contains undiscovered fields. For discussion here, assume that zones and prospects coincide. The risk for each prospect was calculated on a work sheet that varied somewhat from one MMS region to another. The work sheet required the determination of risk factors for three geologic attributes: the presence of a trap,

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures a reservoir, and petroleum. The three risk factors were then mechanically combined as though they were statistically independent. For example, a 50 percent risk on each risk component produces an 87.5 percent risk [1 − (.5 × .5 × .5) = .875] for the prospect (i.e., an 87.5 percent chance that the prospect contains no petroleum). (Some MMS regions used a more detailed risk analysis work sheet in which each of the three basic components of prospect risk were broken down further.) The assignment of a basin risk factor was constrained so that it was never less than the combined risk of the component prospects considered as statistically independent trials. This created a degree of positive correlation among the outcomes of the prospects within a basin, which is good in principle. In typical situations with a large number of prospects in a basin, the lower risk limit for the basin is essentially zero; therefore, even small assigned basin risk numbers will introduce substantial correlation among the outcomes of prospects. The basin risk was also constrained so that it did not exceed the minimum risk for any component prospect. The problem with this method of assigning risk numbers is that it is susceptible to misinterpretation on the part of the dispersed group of risk evaluators. In assessing a specific prospect risk, one must to some extent anticipate basin-level and higher-level risk assessments. It could be argued that it is more natural to assign a prospect risk number conditionally on resources being present in the basin. In particular, the reservoir and hydrocarbon risk factors used in a prospect risk calculation should in many cases reflect properties of groups of prospects or even of the whole basin. The committee recommends that MMS assessors separate more clearly risk factors that are shared versus risk factors that behave independently. The risk numbers assigned to basins, and especially to areas, are extremely critical to the overall resource assessment. Great care in assigning prospect-level risks can be vitiated by misunderstood or carelessly assigned risk factors at higher levels. Detailed work sheets were not used for these higher level assessments but perhaps should have been to insure consistency. It is the committee's impression that the MMS set area-level risks at insupportably high levels. Many areas in the Atlantic and Alaska regions had area risks in the 90- to 98-percent range prior to any economic screening. A risk of 95 percent for an area implies that only one out of 20 similar areas would contain even a single undiscovered field of any size. One of the difficulties in assigning risk numbers to higher levels of the risking hierarchy is the avoidance of confounding the numbers with economic

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures screening. The committee's impression of exaggerated area-level risk numbers may derive from unintended but implied economic considerations. The committee recommends that the MMS explicitly and clearly separate economic considerations from basin- and area-level risk assignments to avoid unintended double discounting. The assignment of risk numbers, particularly at the higher levels of the risking hierarchy, is necessarily a nebulous task, but it nevertheless exerts a strong influence on the final undiscovered resource estimates. Therefore, the committee recommends that the MMS attempt to validate or calibrate risk numbers to increase the credibility of its assessments. The MMS should offer some assurance that assigned risk numbers are not systematically too high or too low, are not confounded with implied economic screening, and are appropriate to the hierarchical nature of risking. Providing such assurance should involve experimental work that compares multiple risk numbers across individuals and geographic units and that employs whatever historical information is available. It is inevitable that different individuals with comparable knowledge will assign different risk numbers in the same situation. The MMS should consult statistical literature on methods for combining diverse expert opinion. The committee believes that diversity of opinion on risk numbers should be propagated to the final reported range of resource estimates. Aggregation The MMS aggregated undiscovered resources across zones, prospects, and basins by using the assigned hierarchical geologic risk numbers to determine probabilistically which prospects (zones) contain undiscovered petroleum. The sizes of these undiscovered fields were also probabilistically assigned from prospect-specific size distributions, as described above. For the economic resource estimate, some or all of the converted prospects did not survive the hierarchical economic screening, which was prospect-, basin-, and area-specific. The net result was a realization of a probabilistically selected collection of undiscovered fields for a given area. The total of the selected field sizes was computed. This process of probabilistically estimating an area aggregate was repeated many times to generate a frequency distribution of area aggregate resources. Finally, a number of zeroes were added to the frequency distribution; the number of zeros corresponded to the area-level risk number. For example, if there were 1,000 simulated area aggregates and the area risk was 90 percent, then 900 zeroes were added to the distribution of aggregates. For each area of

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures each region, the MMS reported the mean value, together with the fifth and ninety-fifth percentiles, of this combined aggregate distribution. In the appendices of the DOI's assessment report, the MMS published “conditional” estimates of area resources. Conditional estimates represent an area's total resource estimate without the zeroes added for the area risk. In other words, conditional estimates assume that there is at least one petroleum-containing field in the area. Of course, these “conditional” resource estimates still contain substantial amounts of discounting for the geologic and economic risks applied at lower levels of the risking hierarchy. Assessment users may be confused by the meaning of conditional estimates. In future assessment reports, the MMS should provide a better interpretation of the reported conditional estimates. The MMS could compare the conditional estimates with past results and explain the causes of significant differences and implications for future production prospects. In addition to reporting resource estimates for each area, the MMS reported estimates for each region. Although the assessment report does not document the method for aggregating area resources to obtain regional estimates, it appears that area-specific distributions were repeatedly sampled to generate regional distributions and that areas were assumed to be statistically independent. As mentioned in the evaluation of USGS assessment methods, the MMS's assumption that areas are statistically independent conflicts with the USGS's assumption of complete conditional dependence among plays. For future assessments, the committee recommends that the USGS and MMS standardize their procedures for aggregating resources. REFERENCES Adelman, M. A., J. C. Houghton, G. M. Kaufman, and M. B. Zimmerman. 1983. Energy Resources in an Uncertain Future: Coal, Gas, Oil, and Uranium Supply Forecasting. Cambridge, Massachusetts: Ballinger Publishing Company. Association of American State Geologists. 1988 . Review of Geologic Information Utilized by the U.S. Geological Survey and Minerals Management Service in their Assessment of U.S. Undiscovered, Conventionally Recoverable Oil and Gas Resources. Tulsa, Oklahoma: Association of American State Geologists. Attanasi, E. D. 1988. Minimum commercially developable field sizes for onshore and state offshore regions of the United States. Pp. 90-117 in National Assessment of Undiscovered Conventional Oil and Gas Resources . U.S. Department of the Interior Open-File Report 88-373, Washington, D.C.Photocopy.

OCR for page 44
UNDISCOVERED OIL AND GAS RESOURCES:: An Evaluation of the Department of the Interior's 1989 Assessment Procedures Crovelli, R. A., R. F. Mast, G. L. Dolton, and R. H. Balay. 1988. Assessment methodology for estimation of undiscovered petroleum resources in play analysis of the United States and aggregation methods. Pp. 30-48 in National Assessment of Undiscovered Conventional Oil and Gas Resources U.S. Department of the Interior Open-File Report 88-373, Washington, D.C. Photocopy. Energy Information Administration. 1990. U.S. Oil and Gas Reserves by Year of Field Discovery. Report CIA-0534. Washington, D.C.: U.S. Department of Energy. Energy, Mines and Resources, Canada. 1977. Oil and Gas Resources of Canada, 1976. Report EP77-1. Ottawa, Canada. Fisher, W. L. and W. E. Galloway 1983. Potential for Additional Oil Recovery in Texas. Circular. 83-2. Austin, Texas: University of Texas Bureau of Economic Geology. Houghton, J. C., G. L. Dolton, R. F. Mast, C. D. Masters, and D. H. Root. 1988. The estimation procedure of field size distributions for the U.S. Geological Survey's national oil and gas resource assessment. pp. 56-72 in National Assessment of Undiscovered Conventional Oil and Gas Resources U.S. Department of the Interior Open-File Report 88-373, Washington, D.C. Photocopy. Minerals Management Service. 1985. Estimates of Undiscovered, Economically Recoverable Oil and Gas Resources for the Outer Continental Shelf as of July 1984. OCS Report MMS 85-0012. Washington, D.C.: U.S. Department of the Interior. Morton, R. A. and D. Nummedal, eds. 1989. Shelf Sedimentation, Shelf Sequences and Related Hydrocarbon Accumulation: Proceedings, Seventh Annual Research Conference, Gulf Coast Section of the Society of Economic Paleontologists and Mineralogists. Austin, Texas: Earth Enterprises, Inc. National Research Council. 1988. Energy-Related Research in the U.S. Geological Survey. Washington, D.C.: National Academy Press. Podruski, J. A., J. E. Barclay, A. P. Hamblin, P. J. Lee, K. G. Osadetz, R. M. Procter, and G.C. Taylor. 1988. Conventional Oil Resources of Western Canada. Paper 87-26. Ottawa, Canada. Tyler, N., W. E. Galloway, C. M. Garrett, Jr., and T. E. Ewing. 1984. Oil Accumulation, Production Characteristics, and Targets for Additional Recovery in Major Oil Reservoirs of Texas. Circular 84-2. Austin, Texas: University of Texas Bureau of Economic Geology. U.S. Department of the Interior, Geological Survey and Minerals Management Service. 1988. National Assessment of Undiscovered Conventional Oil and Gas Resources Open-file Report 88-373, Washington, D.C. Photocopy. U.S. Department of the Interior, Geological Survey and Minerals Management Service. 1989. Estimates of Undiscovered Conventional Oil and Gas Resources in the United States—A part of the Nation's Energy Endowment. Washington, D.C.: U.S. Government Printing Office. U.S. Geological Survey. 1975. Geological Estimates of Undiscovered Recoverable Oil and Gas Resources in the United States. Circular 725. Washington, D.C.: Department of the Interior. White, D.A. 1980. Assessing oil and gas plays in facies-cycle wedges. American Association of Petroleum Geologists, Bulletin 64: 1158-1178.