Research on the timing for subsequent water flooding in Alkali-Surfactant-Polymer flooding in Daqing Oilfield based on automated machine learning.

Determining the optimal timing for subsequent water flooding in Alkali-Surfactant-Polymer (ASP) flooding is essential to maximizing both the technical and economic outcomes of oilfield blocks. This study identified eight critical parameters that influence the benefits of ASP flooding and established parameter ranges based on data from completed blocks and actual field measurements. The optimal timing for subsequent water flooding was determined by evaluating cumulative net profit variations throughout the ASP flooding lifecycle. Given the complexity and high-dimensional nature of evaluating multiple parameters across diverse blocks, a machine learning-driven optimization model was developed. This model enhances work efficiency by automating complex analyses. However, predictive uncertainties and limitations remain due to the variability in oilfield development and the potential for unpredictable changes in reservoir conditions, external market factors and so on, which may affect the model's results. The model was applied to six blocks in the Daqing oilfield currently in the chemical flooding phase, where injection schemes, such as extending the polymer slug, were adjusted according to the model's optimized results. These adjustments yielded an increase in cumulative net profit of 224.9 million CNY compared to the original scheme, with a potential total increase of 752.1 million CNY by the end of the flooding process.