The Use of Simulation and History Matching To Determine Critical Coalbed Methane Reservoir Properties

Abstract

SPE Members Abstract Key coalbed methane reservoir properties can be determined using a logical history matching analysis procedure that incorporates laboratory, geologic and production data. Methane and water production data were history matched using a dual porosity, two-phase finite difference reservoir simulator modified to account for the storage and flow mechanisms unique to coal. Two field examples demonstrate the results and show the applicability of reservoir modeling to assessing the effectiveness of stimulation and predicting well performance. Introduction Coalbed methane is a recognized, significant natural gas resource. However, efficiently producing this resource has proven to be challenging, primarily due to the complex, "unconventional" nature of gas storage and flow. The bulk of the gas-in-place is adsorbed on internal coal surfaces, requiring a large pressure drop for the gas to become mobile. In the majority of the reported cases, the coal cleats are initially water saturated, and the water must first be produced to lower the formation pressure. The water and gas must flow through a complex set of orthogonal cleats and larger fracture systems (joints) to reach the producing wellbore. Finally, the system permeability and porosity change as reservoir pressure is lowered and gas production occurs. Commercial production of methane from coal is still in its infancy. Because of this, the essential data on reservoir characteristics that establish optimum well completion techniques, proper well spacing, and production economics are not yet available. In the past three years, the Gas Research Institute (GRI) has funded several projects to investigate these problems. CRI's work focused on establishing the key in-situ reservoir characteristics and developing reservoir simulation and analytic tools specifically for coal reservoirs. Based on work sponsored by GRI, InSitu, Inc., has shown that slug tests and interference tests can establish initial, near-well properties prior to well stimulation. However, in areas where these tests were not performed and where longer term data are required to describe the larger drainage area surrounding the well, the only method available to determine these critical reservoir properties is to history match production data, using a model that incorporates the complexities of gas storage in coal and accounts for two-phase water and gas flow through the fractured coal system. To determine critical storage and flow properties, this study used a finite-difference, dual-porosity simulator incorporating desorption, diffusion and compressibility to match production data from wells in the Warrior and San Juan Basins. The objective of this paper is to demonstrate how history matching, using a finite-difference simulator modified to include storage and flow mechanisms unique to coals, can be used to determine the essential reservoir characteristics, to assess the effectiveness of stimulation, and to predict performance. The collection of data necessary for history matching, the match analysis methodology, and discussion of parameter sensitivity are also presented in this paper. P. 307^