In this study, a simulation-optimization model is presented to minimize bioremediation cost in an oil-contaminated aquifer. Flow and contaminant transport simulations were conducted using MODFLOW6 and MT3D-USGS models, while bioremediation optimization was performed using the differential evolution algorithm. The accuracy of modeling was validated through a test problem. Subsequently, the Shiraz refinery was selected as a case study, and remediation of BTEX contamination (benzene, toluene, ethylbenzene, and xylene) resulting from a potential leak was examined under four scenarios.The results of Scenario 1 indicated that natural bioremediation alone could not reduce benzene concentration to the permissible level. The other three scenarios were defined based on location, number of bioremediation wells, and implementation duration. In all three cases, the contaminant concentration at the end of the ten-year period fell below the standard threshold.Optimization results showed that Scenario 2 (with two injection and two pumping wells over four years) had a total cost of $157,634, making it 12.1% cheaper than Scenario 3 and 9.5% cheaper than Scenario 4. In contrast, Scenario 3 had the fastest remediation rate.In summary, Scenario 2 had a lower remediation cost with lower pumping and injection rates due to its longer duration, whereas Scenario 3 demonstrated faster cleanup efficiency.