Optimal construction site layout based on risk spatial variability

The construction site layout planning is a complex and important task conducted by project managers and planners. It must able to face the occurrence of potential hazards like fire and blast waves, for instance. However, minimizing risk resulting from natural or technological hazards is still a scientific challenge. In the present paper, a new methodology is developed in order to evaluate the risks within a construction site. It consists of: Modeling construction site components, for instance; electric generator, fuel storage, offices, equipment and material storages, in 2D layout. These components act as hazardous sources and potential targets at the same time, Modeling hazard interaction matrix: it shows the hazard interaction among site components and the attenuation of hazard with distance, Modeling vulnerability interaction matrix: it represents the potential weakness of whole targets to the hazard generated from each source. In the present research, the vulnerability is expressed as function of hazard intensity, Defining the utility function: it aims to afford an optimized site layout with minimum total risk in the construction site, finally Performing spatial analysis technique, utilizing space syntax principle, to realize space configurations in the construction site. As the evacuation process is considered in evaluating and visualizing the risk, the actual risk is amplified by utilizing penalty factor called mean depth. Geographic information system (GIS) is useful in visualizing the spatial variability of the risk within the site. It integrates the potential total impact of the facilities with the space configuration mean depth results. For illustration purposes, the methodology is employed in a case study consisting of several facilities acting as hazardous sources and potential targets in a 2D layout. The risk optimization considers the level of hazards at each source object, hazard attenuation and adopts conditional values for the vulnerability of the target objects. A differential evolution algorithm is adopted to minimize the global risk within the site. The results showed that the proposed methodology is efficient, due to its capability of generating site layout with safer work environment. This in turn leads to minimize work accidents, serious injuries and victims. In addition, the model is capable of highlighting the highest risk areas within a construction site. (C) 2016 Elsevier B.V. All rights reserved.