Assimilating Microseismic and Welltest Data Using EnKF for Accurate Reservoir Characterisation

The pressure trainsient data in welltest can only resolve the thickness-weighted average permeability but are more sensitive to the high productive (high permeability) layers. The mircroseismic data between fracture simulation treatment well and the monitor wells can only resolve the average velocity along its wave path but are more sensitive to the layer (or region) with high wave velocity (low productivity). Therefore, these two types of data are complimentary to each other in reservoir characterization. In this paper, we assimilate these two types of data using the state-of-the-art ensemble Kalman filter (EnKF method. Layered homogeneous and heterogeneous reservoir examples verified the complimentary nature of these two types of) data. The porosities and permeabilities in the layered reservoir obtained after assimilating both types of data are comparable assimilating pressure trainsient and layer rate data. As EnKF is a stochastic process, we generated 10 different ensembles fotor each example for better uncertainty quantification. Assimilating only one type of data may yield biased estimates in porosity and log-permeability and hence biased prediction in the layer rates, while assimilating both pressure and microseismic data yield good estimates in reservoir properties with smaller uncertainty bound and hence more accurate layer rate prediction.