A "matrix" of sensitivity simulations was performed whereby combinations of key reservoir parameters and operational characteristics were evaluated. The resulting families of gas and water production curves provide an Improved understanding of coalbed methane behavior and a means to assess the impact of various operational strategies.
In May 1989 the Gas Research Institute (GRI) entered into a cooperative agreement with a 13-company industry consortium to conduct a reservoir engineering study of the Fruitland coalbed methane resources in the San Juan Basin. GRI's objective was to develop a better understanding of the relationships between coal seam gas producibility, reservoir characteristics and engineering practices. The consortium wished to develop technical information that could assist the New Mexico Oil Conservation Division (NMOCD) and the Colorado Oil and Gas Conservation Commission (COGCC) in their efforts to develop fieldwide rules for well spacing and resource conservation. These objectives were met by the reservoir characterization of selected field shes within the basin and a parametric study based on the application of reservoir simulation techniques. Results of the study were presented at the NMOCD Examiner Hearing in February 1991.
The major technical aspects of the study were:
to collect and examine geologic and engineering data that reflect the range of reservoir properties in the basin;
to develop reservoir characterizations of selected field sites by history matching production and pressure monitor well performance;
from the data analysis and history matching results, to define expected ranges of key reservoir properties; and
to determine the sensitivity of gas and water production rates to the key reservoir properties over their expected ranges.
Unlike conventional gas reservoirs, the traditional methods of analogy, decline curve analysis and material balance are not adequate to describe coalbed methane behavior. As a result, the methodology employed for this study is simulation-based as simulation provides a consistent and reliable way to account for the mechanisms of coal seam gas desorption and diffusion, and to assess the sensitivity of gas and water production to a wide range of reservoir properties and various operating methods. COMETPC 3-D, a two-phase, finite-difference coal seam gas simulator was used to perform the simulations. Prior to the study, the simulator was benchmarked against other black oil and coal seam gas models.
Following Ayers et al., the San Juan Basin was divided into three areas broadly defined by geologic and hydrodynamic similarities (Figure 1).