Application and variance based sensitivity analysis of surfactant–polymer flooding using modified chemical flood predictive model

We study the performance and behavior of surfactant–polymer (SP or micellar–polymer, MP) flooding enhanced oil recovery (EOR) using an analytical chemical flood predictive model (CFPM). The research has two parts based on the deterministic and the stochastic nature of the problem.In a deterministic study, the SP flood performance (ultimate recovery efficiency and oil-rate vs. time) of TORIS reservoir database (the Tertiary Oil Recovery Information System) was predicted using the modified CFPM. Results helped to determine the best candidates for SP flooding based on each reservoir's rock and fluid properties. From there we can determine the effect of different parameters (reservoir rock and fluid properties, injection design variables) on ultimate recovery efficiency and peak oil rate of an SP flood, which gives good clues about the sensitivity of output results to different input parameters.Stochastic study helps to recognize the behavior of the model under uncertain inputs. We used a variance-based sensitivity analysis (SA) method known as Winding Stairs (WS), which needs much fewer runs than traditional Monte-Carlo (MC) and Latin Hypercube methods. The results of the SA method facilitate identifying the most important sources of uncertainty of SP floods either through direct influences or interactions with other parameters. Based on these results we can reduce the uncertainty of output results of SP flood significantly by reducing the uncertainty of the input parameters that cause the largest uncertainty.