Ensemble-based history matching and its application in estimating reservoir petrophysical properties

The reservoir simulation model plays an important role in reservoir management. Accuracy of reservoir simulation is dictated by the quality of the reservoir model used to represent the petrophysical and geological characteristics of a given reservoir. In spite of extensive research, calibration of a static geologic model to dynamic production data is still a challenging task. In this work, an ensemble-based history matching technique using the ensemble Kalman filter (EnKF) method has been developed and successfully applied to improve estimation of reservoir petrophysical parameters, such as absolute and relative permeabilities and porosity.;The ensemble-based assisted history matching technique has also been applied to estimate relative permeability for two-phase and three-phase flow conditions, respectively. Both the power law and the B-spline representative models have been utilized in the estimation process. For the two-phase flow case, the cubic B-spline model is modified and utilized to generate relative permeability curves. It is found that the estimated B-spline controlling parameters for relative permeability curves are improved progressively to their reference values and variance of the parameters is decreased as time index increases. Good estimations of the relative permeabilities are obtained by assimilating the bottomhole pressure or oil production rate of producers. It is shown from the three-phase case study that both the endpoints and the shape of relative permeability curves are estimated with good accuracy by history matching the observation data of the production wells. When endpoints are involved in the estimation process, a larger ensemble size is needed to avoid the filter divergence.;Finally, the standard test case, PUNQ-S3 reservoir model, is used to further investigate the performance of the newly developed ensemble-based history matching technique in a more comprehensive way. Four different testing scenarios, with different combinations of the tuning petrophysical parameters, have been examined. It is found that the ensemble-based technique is capable of estimating petrophysical parameters by conditioning the reservoir geological models to production history. Compared to the results available in the literature, good production performance predictions have been achieved by using the technique developed in this study.;The reference case is found to be located within the uncertainty ranges of results generated from all four data assimilation scenarios. Case studies show that the selection of parameters will affect the parameter estimation and production prediction performance. Relative permeability has a profound impact on the history matching and performance prediction. In this study, relative permeability can be estimated with good accuracy when absolute permeability fields are known; however, history-matched models may not provide good parameter estimation due to the non-uniqueness of the history matching.;A method of integrating the ensemble-based history matching with the geostatistical technique has been developed and successfully applied to improve estimation of the absolute permeability and porosity. The newly developed technique is validated with a synthetic reservoir. Compared to the existing implicit estimation approaches, the newly developed technique does not require the gradient of the objective function, and thus it is easy to implement.