Immiscible and Near-Miscible Gas Flooding in Tight Chalk: Laboratory Experiments and Compositional Simulation

Abstract

History matching methods are widely used to extract relative permeability curves as well as other uncertain parameters, which cannot be measured accurately through laboratory analysis. This study presents the results of gas flooding experiments in composite chalk cores and seeks a systematic approach toward overcoming the challenges encountered during history matching of the performed experiments. Two vertical core flooding experiments are conducted on tight chalk composite cores at two different pressures providing immiscible and near-miscible conditions for the natural gas and live oil used. An EoS (equation of state) model tuned with routine PVT tests as well as swelling data is used to simulate the experiments by Eclipse compositional simulator E300. Difficulties encountered in the process of history matching are addressed, and a proper method to be implemented to resolve each of the problems is proposed and investigated in detail. A common drawback in compositional simulation of gas injection processes is the excessive vaporization of oil into gas due to local equilibrium assumption, which leads to over-predicting the oil production. It is shown that using a proper technique such as the Sorm method (available in Eclipse via SOR keyword) can be an efficient solution to overcome this issue. It is also established that in the absence of enough reliable data for absolute and relative permeability, these parameters can be subjected to modification and improvement based on experimental observations such as gas breakthrough time and the pressure difference across the core. Furthermore, the change of relative permeability due to the reduction of IFT (interfacial tension) at near-miscible conditions is studied in detail, and the contradicting findings in this area reported by various authors in the literature are elaborately discussed. Different approaches for correcting the relative permeability of the wetting and non-wetting phases are examined in the history matching process and the obtained results are evaluated by being compared to the experimental results of this study. The findings of this work can help to identify and resolve some of the most common problems in compositional simulation of gas injection processes. These results should specifically be taken into consideration in upscaling the reservoir characteristics and performing field-scale simulations in order to obtain reliable results for the future performance of the field.