A new approach of history matching coalbed methane pilot wells

Abstract

History matching was usually conducted on pilot wells for calibrating reservoir parameters in static model. How to complete matching work with limited geological data is a big challenge, especially for coalbed methane (CBM) reservoir with high heterogeneity. In this paper, a new workflow was proposed to overcome it. Study area lies in Surat Basin, Australia, and the targets for six pilot CBM wells are low rank coal. Totally, five parameters, including permeability, gas content, gas saturation, fracture porosity and variogram length, were selected for testing their uncertainties. For the former two parameters, the correlations with depth were established for predicting their distribution separately, and then, percentile (P10, P90) methods were used to define their high and low variation ranges. For the later three parameters, experience methods were used to define their high and low variation ranges. Then, experimental design method was adopted to generate 43 realizations of simulation cases with aforementioned five parameters. For gas rate, water rate and bottom hole pressure of each well, 10 more better matching cases were selected separately, and then compared to find the common cases. After combing all the common cases of each well together, a general understanding about geological conditions was achieved. It was found that all cases have high gas saturation and most cases have high fracture porosity and high permeability. These results agree with previous regional geological studies.