Optimization of In-Situ Bio-Surfactant Production in Microbial Improved Oil Recovery

Abstract Microbial Improved Oil Recovery (MIOR) has the potential to be one of the most cost-effective IOR methods. However, uncertainties about different MIOR mechanisms and their immediate returns on investment (ROI) have prevented their wider adoption. Here, we demonstrate that applying state-of-the-art optimization algorithms to the in-situ bio-surfactant production process can result in outstanding financial benefits. Furthermore, sensitivity studies show consistently high ROI with varying microbial conditions and costs for a synthetic reservoir model. We performed numerical experiments on a reservoir simulator with well-calibrated in-situ bio-surfactant production functionality. Different costs for MIOR (including costs of microbes and nutrients) as well as oil prices were considered in calculating the net present values (NPV) for various MIOR processes. In addition, a novel Ensemble Optimization (EnOpt) algorithm (originally developed to optimize injection and production rates in waterflooding) was used to adjust microbe and nutrient injection concentrations dynamically during the MIOR processes to further maximize NPV. We compared the final NPV's after (1) only waterflooding, (2) MIOR without optimization, and (3) MIOR with optimization. The in-situ bio-surfactant production showed remarkable 5- to 10-fold NPV increases for MIOR without and with optimization, respectively, compared to only waterflooding. In addition, the cumulative oil production increased with an extra 15 to 20% of original oil in place (OOIP) for the synthetic reservoir model after MIOR. Finally, sensitivity studies showed that the NPV increases were consistently above 10-fold for a wide range of microbe prices, indigenous reservoir microbe concentrations, and oil prices for MIOR with optimization compared to only waterflooding. To the best of our knowledge, this paper is the first study that quantifies the ROI of MIOR measured in NPV with realistic assumptions of the costs and rewards in an exemplary in-situ bio-surfactant production process. The observed high NPV increases were promising and encouraging for further developments in this direction. Finally, successful application of the EnOpt algorithm to the MIOR problem has not been reported in literature and is worth highlighting here.