(56o) Probabilistic Risk Assessment Tool Applied in Facility Layout Optimization

Facility siting and layout configuration are critical factors during design and expansion of any industrial installation. In the recent past, the severity and propagation of various chemical incidents can be attributed to improper layout arrangement or proximity of chemical facility to a densely populated area. To address this problem, several studies have been reported incorporating safety into the layout optimization. However there is still a need for a method that combines layout reformulation and detailed risk assessment accounting uncertainties. Based on that, the objective of this study was to develop a risk-based layout optimization tool to support decision-making process during the design phase by providing safer layout options that are also cost-efficient. The proposed methodology is divided in three steps: risk calculation, determination of safety distances, and layout optimization.

In the first stage, a risk assessment program is developed in MATLAB to estimate risks associated with human life losses and structural damage in a chemical plant. Analytical models for fire and explosion scenarios and toxic chemical releases were included in the code. Monte Carlo simulation was then employed to propagate uncertainties attributed to environmental conditions and release parameters. The proposed program generates risk maps and risk distributions at a particular point of interest. Subsequently, domino effect concepts are included in the resulting code to obtain minimal separation distances between process units necessary to prevent escalation events. These distances vary as a function of the targeted unit type, escalation vector (overpressure or fire impingement) and the risk acceptability criteria. In the last stage, risk maps and safety distances are accounted in a mixed-integer linear programming (MILP) for layout optimization. The objective function is set to minimize the total capital cost associated with structural damage risk, fatality risk, pipeline interconnection, and protective devices. A grid-based approach is selected to maintain the linearity of the system, and individual risk criteria is applied as an additional constraint for high occupancy buildings, meaning that the overall risk for buildings such as control room or lab may not exceed this criterion.

The applicability of the proposed methodology is demonstrated through a case study. Sensitivity analysis is performed varying the risk values, safe distances, and risk acceptance criteria. Even though different layout options are generated, a trend is observed towards placing the equipment by its type: high occupancy buildings are placed in regions with low wind probability while storage units are located nearby the main processing unit, requiring protective devices to prevent escalation events.