Optimal determination of chemical plant layout via minimization of risk to general public using Monte Carlo and Simulated Annealing techniques

• Optimum Industrial Plant Layout via Risk to General Public Minimization.
• Monte Carlo Sampling for Estimating Projection Areas of Accident Effect Ranges onto Population Polygons.
• Determination of Accident Effect Ranges via EFFECTS Software by TNO.
• Penalization of the Risk to General Public Minimization Objective to Account for Violation of Positioning Constraints of Units.
• Search of Risk to General Public Minimization by Means of a Simulated Annealing Strategy.

A new algorithmic approach is presented to optimally locate process or storage units in the plant area (layout) of industrial facilities. The proposed formulation defines a configurational optimization incorporating spatial constraints for locating units inside the industrial area and an objective measuring the consequences to near residential areas in the event of accidents. The Monte Carlo method is used to estimate superposing areas in order to check constraints and to evaluate the objective, which measures the superposition of accident effect areas onto population polygons. The method is fed with an initial feasible layout where the coordinates of all units are given. Then, a Simulated Annealing search randomly moves units throughout the industrial area, penalizing unfeasible configurations, until a feasible layout is found minimizing the consequences of accidents to general public. The method was validated through two hypothetical case studies: (i) a new marine fuel terminal; and (ii) the addition of a new LPG storage yard to an existing refinery. In each case, it was demonstrated that the method effectively reduced risks to the surrounding communities, since it achieved, in both cases, feasible plant layouts minimizing the populated area reached by the accident effect range of each unit in the installation.

Figure optionsDownload as PowerPoint slide