Abstract
Process plants should be designed to be economically viable and environmentally friendly, operable and maintainable when projects are implemented. Process synthesis is a rather challenging task since results obtained should be feasible and acceptable. The safety of processes is among the most important considerations in obtaining more acceptably realistic results. Safety issues have mostly been addressed as retrofit analysis for projects that have already been implemented. There is a reasonably well-developed field of research focusing on deviation events and on predicting their possible consequences. This approach is well suited for making action plans to control deviation events. However, designers are aware that every control system has its weaknesses and gaps. For this reason, safer design can be obtained when inherent safety is considered. Inherent safety can successfully be enhanced at the early stages of the design. There have been some studies dealing with safety metrics at the design stage. The aim of this study is to develop a quantitative risk assessment method and to incorporate it into process synthesis, using a mathematical programming approach. A mixed-integer nonlinear programming (MINLP) model has been used for the synthesis of a methanol production process including risk assessment during the synthesis. The superstructure is extended using a variety of process units in order to enable the selection of safer designs. Risk assessment is performed simultaneously with MINLP process synthesis, where the risk is determined for each unit of the process individually and, after summing up, for the whole process. Since improving the safety of individual units can lead to decreased overall risk, it is obvious that safety at both the individual and overall levels should be considered simultaneously. The results obtained using the method developed are inherently safer, while still obtaining economic viability.
