A NEW CHARACTERIZATION METHOD FOR COAL BED METHANE

Abstract

Coal bed methane reservoirs are hard to characterize due to their small scale heterogeneities which often vary from seam to seam. Based on recent developments in image analysis techniques and fracture generation software, we propose a methodology aiming at a better quantitative characterization of the macro-cleat system of coal, for better prediction of petro-physical properties. The methodology is based on four main steps:  Image analysis aiming at identifying to the best of the CT scan resolution, the macro-cleat system for large cores.  A statistical analysis of the main features characterizing the network: orientation, length distribution, frequency, density, aperture distributions and average distance between fractures. Different families are then identifi ed and their respective statistics recorded.  Generation of realistic images at a large scale (10m*10m*10m), from the statistics of one core, using a dedicated software developed at IFP.  Calculation of porosity and permeability values for up to 30 block of 1m*1m*1m, selected randomly from the large block. The generation of such a petro-physical data-base allowing the development of K-PHI plots can improve log analysis performed in CBM fi elds leading to fl ow-unit identifi cation, since permeability attributes are considered along with porosity. Problems encountered and use of those statistics for other purposes than characterization are discussed. Ex- amples of applications and of the results leading to this characterization are shown, using coals from France.