Evolution of U.S. Geological Survey Methodology for Assessing Continuous Oil and Gas Accumulations

Ronald R. Charpentier and Troy A. Cook
U.S. Geological Survey, Denver, Colorado

The U.S. Geological Survey (USGS) methodology for assessing continuous oil and gas accumulations (such as coalbed gas, tight gas sands, and gas shales) has greatly evolved in the last few decades. Early efforts in the 1980s were primarily based on estimating in-place resources. Increased drilling and geologic understanding led to a new methodology based on well productivity. This methodology was used to assess the major continuous oil and gas accumulations of the United States for the 1995 U.S. Geological Survey National Assessment. Greatly increased drilling in the 1990s led to further improvements in geologic understanding and, around the year 2000, an improved methodology that better represented the concentration of resource into sweet spots. Future improvements in the methodology will come from better understanding of the spatial structure of the resource.

Early efforts to assess unconventional oil and gas resources in the late 1970s and into the 1980s were based on geologic and engineering calculations of in-place volumes. In-place volumes were estimated using areas and thicknesses of rock, porosities, fluid saturations, and other engineering parameters. Assessments were conducted for specific basins or plays, such as the Devonian gas shales in the Appalachian Basin or tight gas sands in individual western United States basins. There was no concerted effort to assess all such accumulations in the entire United States Estimation of recovery factors was commonly hampered by limited production data.

The methodology used for the 1995 U.S. Geological Survey National Assessment differed significantly in that it was based on well productivity data rather than in-place calculations. Much more production data were available by then, and a new well-based (or more precisely, cell-based) methodology was developed. The number of cells and their productivities were based on production well performance from the play being assessed or an analog play. Cell productivities were distributed in a lognormal distribution, but cells that fell below a minimum productivity value were treated as dry. Sweet spots were accommodated only when they could be mapped as a separate play. Otherwise, the productivities were assumed to be randomly distributed spatially within the play.

A review of the results of the 1995 assessment, including the subsequent economic analysis, indicated two main weaknesses of that methodology. First, the assessment of continuous resources included a fraction that was immediately relevant to supply, but also a fraction that was of such low quality as to not be relevant to supply in the next few decades without significant technological improvements. Second, the assumption of spatial randomness of well productivities used in the 1995 assessment did not impact the total volume of resource, but did have a significant impact on the economic assessment. A more sophisticated model that took into account these weaknesses was needed.

Increased drilling and production in the 1990’s also impacted the methodology at about this same time. Many of the 1995 continuous plays, especially the coalbed gas plays, had productivities based on analogy because many of the plays had not had significant production before then. New production data also showed that the assumption of spatial randomness was appropriate for some plays, but that most showed some amount of clustering into relatively high productivity sweet spots. The two main weaknesses of the 1995 methodology could be addressed by assessing the resources in sweet spots, rather than those in the entire play. This would leave out those areas that were unlikely to be relevant to supply in the next few decades. It would also make the economic analysis more realistic by taking into account the improved economics within sweet spots.

Current research in the USGS focuses on the spectrum concept for spatial variability within continuous accumulations. At the conventional end of the spectrum, resources are present in discrete accumulations with sharp boundaries, generally described by oil-water or gas-water contacts. At the other end of the spectrum, the classic continuous model, the resources are not in discrete accumulations, but are randomly distributed across extensive areas. Oil-water and gas-water contacts are not evident. Between these end members, there is increasing spatial organization and spatial correlation as one goes from the classic continuous toward the conventional end of the spectrum. This is reflected in the better-producing wells becoming more organized into sweet spots. The boundaries of sweet spots change from gradational to distinct as one approaches the conventional end. Better understanding of this spatial structure in continuous accumulations will lead to future methodology improvements that take that structure into account.