ENSEMBLE BASED CHARACTERIZATION AND HISTORY MATCHING OF NATURALLY FRACTURED TIGHT/SHALE GAS RESERVOIRS

Tight sand and shale gas reservoirs are becoming bigger players in natural gas supply system as easily developed gas resources become scarcer, natural gas prices go higher, and more efficient technologies become available to develop these hard to access reservoirs. Although these reservoirs are really hard to develop but better hydraulic fracturing techniques are making them more productive. Production and development of these types of reservoirs depend heavily on the hydraulic fracturing and the existence of natural fractures in the formations, since the permeability of the matrix is so low that no commercially feasible gas production can be achieved without fracturing. Therefore, characterizing and understanding the fractures have become of great importance. We have proposed using different methodologies for fracture characterization and simulation in tight gas reservoirs. Dual Porosity, Dual Permeability modeling has been used for simulation purposes. For a commercial simulator to be able to simulate the fluid flow in a reservoir, properties such as permeability and porosity should be known which are usually unknown because enough observations are unavailable. The only known values are measurements such as bottom‑hole pressures and production rates. Such a system requires solving an inverse problem. Ensemble Kalman Filter has been used for our estimation purposes. EnKF is a Monte Carlo based, MMSE estimation tool that generates multiple realizations based on our prior knowledge of the reservoir and improves their performance using data assimilation. Our proposed methodology has been tested on a synthetic 2D reservoir model containing a couple of fractures, some intersecting the wells. The results have proven the applicability and the advantages of this methodology. We have shown that EnKF can be used efficiently to characterize a fractured reservoir when the approximate orientations of fractures are known which may be obtained from the stress‑strain field data in the reservoir.