Optimization procedure to select an inherently safer design scheme

Abstract

There are different well-established strategies for making a process plant inherently safer. The benefits of applying these strategies on reducing the overall risk inside a plant are obvious.However, some of these changes are rejected many times because they appear to be too costly.But if the effects of applying inherently safer design strategies are investigated not only on the processing costs of a plant but also on the potential accident costs, the decision would in fact be different. In this paper an optimization procedure is proposed which integrates both processing and accident costs for different design schemes. In this procedure, some of the design variables are chosen with regard to inherently safer design strategies. The objective function is the sum of accident costs and plant lifecycle processing costs. For assessing accident costs, consequence modeling techniques and probit functions are applied. Consequence modeling formulas and an objective function are codified in an optimizer package (MATLAB) and to accomplish the optimization process a process simulator (called HYSYS) is coupled with this package. The application of the proposed procedure is demonstrated by selecting an optimum process scheme for a Refrigeration plant as a case study.

Inherently safe and economically optimal design using multi-objective optimization: The case of a refrigeration cycle

“…10 inherent safety strategies are incorporated into the optimization algorithm by considering 184 decision parameters associated with each strategy. The novelties of the present study in 185 comparison to the previous research(Eini et al, 2015) are as follows: 186  Multi-objective optimization programming is applied in order to simultaneously optimize 187 economics as well as inherent safety measures 188 Detailed quantitative risk assessment is applied in order to measure inherent safety 189 considering incidents frequencies and consequences 190  The combination nature of accident scenarios is formulated in the optimization program.191  Layout optimization considering land cost in the economic analysis as well as safe 192 distances in the case study 193 Accidents frequencies have not been considered in the procedure presented by Eini et al (2015). 194 Their defined objective function comprised of summing plant processing costs and accident 195 costs.…”

“…The value of optimization variables for the base case design (Figure 4), P=676 kPa, NPP=1), the 639 solution reported by Eini et al (2015) (P= 600kPa, NPP= 11), and the optimum solution found in 640 this study (P= 350 kPa, NPP= 6) are shown in Table 6. Table 6 suggests that, in comparison to 641 the base-case, for 19 percent increase in the TAC, it is possible to increase the level of process 642 safety by 1 order of magnitude.…”

“…Table 6 suggests that, in comparison to 641 the base-case, for 19 percent increase in the TAC, it is possible to increase the level of process 642 safety by 1 order of magnitude. With respect to the previous results by Eini et al (2015), the new 643 solution is much more realistic as the number of parallel paths is almost halved (i.e., less 644 complexity) but the associated risk is only increased by 17%, In addition, the new solution 645 considers the costs of land-use, piping, and instrumentation that were ignored in the previous 646 study. 647 Saeed Eini, Hamid Reza Shahhosseini, Majid Javidi, Mahdi Sharifzadeh, Davood Rashtchian.…”

“…Quantification of the opportunities for ISD based on consequence modeling approach provides a 117 clearer understanding of the risk (Eini et al, 2015). modeling-based methods in order to select the inherently safer design.…”

“…Such a 156 framework conforms to optimization programming. 157 Eini et al (2015) proposed an optimization procedure which integrates both plant economics and 158 accident costs based consequence modeling for different design schemes. They applied the sum 159 of accident costs and processing costs all through the plant lifespan, as the objective function.…”