An Industrial Area Layout Design Methodology Considering Piping and Safety Using Genetic Algorithm

Ruiqi Wang,Yan Wu,Yufei Wang,Xiao Feng
DOI: https://doi.org/10.1016/j.jclepro.2017.08.147
IF: 11.1
2017-01-01
Journal of Cleaner Production
Abstract:The industrial area layout can influence the economic benefit, safety and surrounding environment of an industrial area to a large extent, especially the length of piping and safety. Every year, the construction of new industrial areas requires thousands of tons of steel pipes, resulting in a high consumption of natural resource of iron and high emission of carbon dioxide. Additionally, the long pipeline will increase the energy consumption for material transportation. Meanwhile, accidents in industrial areas can usually result in serious damage to the local environment. So the optimization of industrial area layout holds great significance for natural resource, energy saving, and environment protection. But up to now, very few papers have been reported to consider the use of pipes and safety simultaneously to optimize the layout in industrial area level. Most works only focus on safety aspects, and the impact of connection is ignored. Accordingly, this paper proposes a methodology to optimize the relative position of each plant, whose objective function consists of piping cost and safety cost. In this paper, despite conventional consideration of material piping, steam piping, which features multiple-branches connected pipeline network, is also considered. For safety issues, it is firstly analyzed by qualitative principles to limit position for some specified plants. Then quantitative analysis, including explosion and toxic release, are optimized simultaneously with connection cost. A genetic algorithm is used to solve the proposed model. Kruskal algorithm and Arrangement & Combination are used to calculate the length of steam piping. Finally, a case study illustrates the effectiveness and applicability of the proposed methodology.
What problem does this paper attempt to address?