Qingjing Hong

Based on the results of 2D visualization oil displacement experiment, this study uses a gray image recognition algorithm, which relates the recognized gray data with the oil saturation, and then quantifies the oil saturation in each area of the 2D visualization experiment. The EOR effects of different composite modes in different stages in different regions during multi-component composite steam flooding are analyzed. The microscopic displacement characteristics during the experiment were photographed by macro lens.It is found that in the process of pure steam flooding, the phenomenon of steam channeling is obvious, and the injected steam mainly flows along the dominant channel between injection and production wells, and the sweep efficiency is only 52.4%. After combining foam and viscosity reducer successively, the overall sweeping efficiency has been greatly improved, and the final sweeping efficiency has reached 97.97%. The picture taken by the macro lens shows that the foam formed by the foaming agent and the non-condensate gas and the emulsified oil droplets formed by the viscosity reducer and crude oil block the mainstream channel to a certain extent through the Jamin effect, causing the subsequent injection fluid to divert, which is the main reason for the improvement of sweeping efficiency. The acquired images are partitioned and combined with image recognition algorithm, then the oil saturation of each area is identified, and the recovery effect of multi-component composite steam flooding process is evaluated by the evaluation parameters of enhanced oil recovery factor. In the process of pure steam flooding, the main displacement is concentrated in the near-well zone and main flow channel, and the final recovery rate is only 28.01%. In the process of succession foam composite steam flooding, the displacement area extends to the middle of the swept zone and the side channels, and the recovery factor is increased by 36% on the basis of steam flooding. Finally, the viscosity reducer composite steam flooding further extends the displacement area to the front of displacement zone and corner zones, and increases the oil recovery by 10.71% on the basis of foam composite steam flooding. Multi-component composite steam flooding was used to achieve the multiple compounding of steam-gas agent, and the final recovery factor reached 74.72%, which was 46.71% higher than that of pure steam flooding.This study introduced an image recognition algorithm that offering a more accurate, intuitive, and efficient alternative to traditional physical analysis methods. The experimental results reveal the EOR mechanism of multi-component composite steam flooding, and this information is helpful to understand the oil displacement mechanism deeply. By overcoming the limitations of traditional steam flooding methods, this method provides a valuable reference for optimizing the steam flooding process and promoting future research in the field of heavy oil recovery.