294110 Localization of Control Rooms Using a MINLP Approach, Based On Accidental Explosion Scenarios
Localization of control rooms using a MINLP approach, based on accidental explosion scenarios
L. Sierra1, S. Rodríguez2, J. Cadena3, M. Ramos4, J. Gómez5, F. Munoz6
Chemical Engineering Department, Universidad de los Andes, Bogotá, Colombia
Facility layout for the process industry plants is an important element in the basic engineering stage, which is fundamental in the design of the process, influencing significantly the future safety performance of the operation and its related costs. A control room serves as an operations centre where the facility can be monitored and controlled, therefore, its location and design is a priority for the safety of the process unit operation. Given that the control room is the ‘core’ of control and monitoring of the process, any significant damage to it resulting from a process incident might jeopardize the whole process and generate subsequent incidents. This work presents a novel methodology intended to serve as an additional tool for aiding the decision-making process regarding the location of control rooms, taking into account the risk of explosion and several passive barriers against it. The methodology is based on a mathematical model that bonds the design and localization of a control room into a computational tool, which has been developed taking into account key characteristics of a process or plant layout for explosion scenarios. The tool combines stochastic estimation of atmospheric parameters through raw Monte Carlo simulations with a mixed integer nonlinear programing (MINLP) optimization to determine the control room location. This process also helps to choose the best construction strategy for the control room among a portfolio of strategies fed to the model by the user. The final results of the methodology are the optimal location for a control room in an existing facility and the minimum costs related to the risk level to which it is exposed.