Quantitative Evaluation of Water Flooding in a Low Resistivity Heavy Oil Reservoir with NMR and Conventional Logs

The classification of water flooding severity is crucial for planning reservoir production and improving the recovery ratio. In this paper, we study a siliciclastic heavy oil reservoir in Bohai Bay, with resistivity reading close to, or even lower than the wet zone (3~5Ω.m). In this environment, computing original reservoir Sw using Traditional hydrocarbon saturation equation is challenging. As a result, the displacement efficiency of a water drive cannot be accurately determined. In order to properly evaluate displacement efficiency, we must estimate initial reservoir Sw (Swirr) and the modern day Sw. Sw can typically be estimated from NMR data with a proper T2 time cutoff. However, in heavy oil reservoirs, the relaxation times of oil and capillary bound water overlap, leading to an over-estimation of Sw. We propose to compensate for the heavy oil effect by adjusting the cutoff until NMR Sw matches the Sw from core Mercury Injection for Capillary Pressure (MICP). As oilfield development proceeds, water displaces some oil in the pore space. Since the injected water has higher salinity than reservoir water, formation resistivity (Rw) becomes lower. Based on the material balance theory, the variable multiple water injection material balance equation is established, and the equation set is established by combining the material balance equation with the Simandoux equation and the calculation formula of mixed water resistivity (Rwz). According to the rock electricity experiment under different salinity, the dynamic rock electricity parameters are used in the Simandoux equation, and the mixed water resistivity and modern day Sw after water flooding are solved iteratively under the original SW constraint. The displacement efficiency is calculated as the difference between Sw and modern day Sw. The proposed method was applied to 10 wells and improved the Sw accuracy by 5%-15%. The continuous solution Rw from our method matches Rw measured in the lab. The calculated displacement efficiency is compared with actual production history and the accuracy improved from 68% to 80%.