Integrated Simulation for Microseismic Fracture Networks with Automatic History Matching in Tight Oil Development: A Field Case from Block Y2 in Ordos Basin, China

Abstract

Abstract The reliability of forecasts, fracture design, and recovery enhancement strategies in tight oil reservoirs is significantly compromised by the substantial uncertainties associated with fracture characterization. This article introduces an integrated simulation workflow for modeling microseismic fracture networks in tight oil reservoirs, incorporating automatic history matching, as illustrated through a field case study from block Y2 in the Ordos Basin, China. The model stochastically generates the geometry of complex fracture networks (CFNs), including parameters such as length, aperture, inclination and azimuth angles, and spatial positioning, constrained by data from hydraulic fracturing, core analyses, and microseismic monitoring. It employs a stochastic parameterization model to produce an ensemble of initial CFN property realizations and utilizes an advanced Green function-based hierarchical fracture model to accurately depict CFN morphology. The model is further refined, and its uncertainty in fracture characterization is minimized through calibration with an innovative Ensemble Kalman Filter-based assisted history-matching algorithm. Evidence suggests that this comprehensive approach effectively leverages all available geological data, substantially reduces uncertainties in the production process, and aids in identifying the optimal development strategy.