Optimization of gas detectors placement in complex industrial layouts based on CFD simulations

The optimal placement of hazardous gas detectors plays a relevant role in the prevention and mitigation of critical scenarios among chemical and process facilities. However, the operative performance of gas sensors networks is affected by the complexity of process facilities layout and sensors' reliability. Moreover, economic constraints may bias the selection of the optimal number of sensors. The present work deals with the development of a methodology to support the selection and positioning of hazardous gas detectors in complex industrial layouts. The methodology exploited preliminary computational fluid dynamics (CFD) simulations to quantify the extension of hazardous gas clouds. A comparison among gas dispersion studies based on CFD and integral models was also carried out to demonstrate the limitation of adopting simplified analyses on the evaluation of gas dispersion hazards. The optimization methodology was then developed based on a twofold algorithm. The first algorithm aimed at determining the optimal sensors positions, seeking for the locations that maximize the detection frequency for a given number of sensors. The second algorithm adopted a multicriteria decision analysis based on the reference point method to find the optimal economic number of sensors and, thus, the most cost-effective strategy. A case study located in a real gasoline storage plant was analysed. The results demonstrated the advantage of adopting a sound quantitative evaluation of the number of detectors integrating economic and risk-based evaluations.