A mixed integer quadratic programming formulation of risk management for reverse osmosis plants

There are many critical factors affecting the operation of reverse osmosis (RO) plants such as scheduled maintenance, response to warnings or accidents. Risk management (RM) is an area that is attracting a lot of interest from the scientific and industrial community. The main objective of RM is to provide policies to obtain better tradeoffs in safety and productivity. This paper shows a method that uses risk metrics to forecast and optimize the freshwater, the benefits and the costs associated to choose a set of risk mitigation actions. Mitigation actions reduce the risk impacts that may affect the system. A Model Predictive Control approach is used to determine the set of mitigation actions to be executed and the original control variables of the plant over a finite time period. The mitigation actions to be carried out can be discrete or continuous. Therefore, the resulting optimization problem is formulated as a mixed integer quadratic problem (MIQP). Several scenarios are simulated and also a Monte Carlo simulation is undertaken. ►Risk management and desalination plant planning are joined. ►Optimization method based on risk management. ►Model Predictive Control is used to compute the optimization problem. ►Mitigation actions are executed to reduce the risk impacts. ►A Monte Carlo simulation is undertaken to show the feasibility of the method.