Novel Evaluation of Foam and Immiscible Gas Flooding in Glass-Silicon-Glass Micromodels

Abstract

We present a systematic workflow to facilitate the visualization of different fluids behavior/performance. Fluids study in this work are brine, gas and specially foam, during flooding experiments in glass-silicon-glass micromodels. These allow for the detailed evaluation and comparison of the individual flooding experiments. This workflow can then be used as part of screening processes to evaluate the fluid-fluid interaction in a porous medium. The experimental setup consists of a cabinet-dryer with a camera mounted on top. Micromodels are placed inside the cabinet and PTFE Teflon® tubings are used as connection lines. Fluids are injected using a syringe pump. Pressure is measured via a differential pressure transducer. This setup allows to visualize the entire pore space every 10 seconds. To achieve more comparable results, a black oil is used as the displaced fluid. Brine flooding is used as a benchmark to which results can be compared to. Foam was generated before injection with a mixture of a commercial surfactant and 3 g/l NaCl brine. The observed behaviour for the three different flooding were in line with the reported in the literature. First, gas flooding depicted the lowest final recovery with 22% of the OIIP produced, with viscous fingering clearly visible. The weak performance of gas was also displayed in the recorded differential pressures. No effect on pressure due to gas injection was observed during the flooding. Second, the brine flood performed better than the gas flood, where 36% of the OIIP was ultimately recovered. Due to the more favorable mobility ratio of brine and oil this improvement was to be expected. Third, foam flooding achieved the best oil recovery with 58% of the OIIP produced. Pore blocking and the thus increased areal sweep efficiency is the reason for this improvement. Differential pressure behavior for foam and brine flood was similar: A steep pressure decrease after entering the model until breakthrough was observed, although foam had a higher initial differential pressure than brine before entering the model. The high initial pressure difference to the brine flood, is assumed to be due to the compressibility of the individual foam bubbles present in the tubing. The workflow presented in this paper, could lead to a fast and economical addition to EOR screening processes. Due to only small volumes of fluids being required to get qualitative and quantitative results. This, in turn could provide relevant insight for foam and immiscible process understanding and modelling.