Monte Carlo simulation-based dynamic mixed integer nonlinear programming for supporting oil recovery and devices allocation during offshore oil spill responses

Abstract

In response to an offshore oil spill, many devices such as skimmers are deployed to recover spilled oil. The lack of effective decision support for device allocation and operation can usually result in a compromise of recovery efficiency and/or waste of resources and manpower. It is therefore much desired to optimize such processes by integrating the simulation of oil recovery and weathering processes, an optimization module, and an uncertainty handling approach. However, limited studies have reported on such integration. Furthermore, no studies have considered the allocation and management of oil recovery devices. To help fill the gaps, this study developed a Monte Carlo simulation-based dynamic mixed integer nonlinear programming (MC-DMINP) approach to provide sound decisions for devices allocation and recovery operation in a fast, dynamic and cost-efficient manner. In a case study, regression models were developed to simulate the efficiencies of three types of drum skimmers based on the past performance evaluation tests. The models were further integrated with the simulation of oil weathering processes and the optimization method. Finally, the uncertainties in slick area, temperature, and wind speed were also involved in the case study. The optimization results without the consideration of uncertainty indicated a 79.3% of oil recovery efficiency. Meanwhile, 18.5% of the spilled oil was evaporated and 2.1% was dispersed. With the consideration of uncertainties, the mechanical collection of oil still had a major contribution to the transport and fate of oil. Negative effects on mechanical collection and positive effects on evaporation were observed from the uncertainties associated with slick area and temperature. The uncertainties of wind speed had positive effects on dispersion. The results demonstrated that the developed MC-DMINP approach could help making timely, sound decisions on the allocation and operation of oil recovery devices and therefore ensure more efficient response actions under dynamic and uncertainty.