Spatial Modelling of Urban Physical Vulnerability to Explosion Hazards Using GIS and Fuzzy MCDA

Yasser Ghajari, Ali Alesheikh, Mahdi Modiri, Reza Hosnavi, Morteza Abbasi
2017 Sustainability  
Most of the world's population is concentrated in accumulated spaces in the form of cities, making the concept of urban planning a significant issue for consideration by decision makers. Urban vulnerability is a major issue which arises in urban management, and is simply defined as how vulnerable various structures in a city are to different hazards. Reducing urban vulnerability and enhancing resilience are considered to be essential steps towards achieving urban sustainability. To date, a vast
more » ... body of literature has focused on investigating urban systems' vulnerabilities with regard to natural hazards. However, less attention has been paid to vulnerabilities resulting from man-made hazards. This study proposes to investigate the physical vulnerability of buildings in District 6 of Tehran, Iran, with respect to intentional explosion hazards. A total of 14 vulnerability criteria are identified according to the opinions of various experts, and standard maps for each of these criteria have been generated in a GIS environment. Ultimately, an ordered weighted averaging (OWA) technique was applied to generate vulnerability maps for different risk conditions. The results of the present study indicate that only about 25 percent of buildings in the study area have a low level of vulnerability under moderate risk conditions. Sensitivity analysis further illustrates the robustness of the results obtained. Finally, the paper concludes by arguing that local authorities must focus more on risk-reduction techniques in order to reduce physical vulnerability and achieve urban sustainability. Sustainability 2017, 9, 1274 2 of 29 improving awareness of the degree of vulnerability in different locations and areas of the city. This study aims to analyze the vulnerability of buildings, as major physical urban elements, to intentional explosions, which are a common form of man-made hazard. The vulnerability of the building stock in District 6 of Tehran is evaluated under different risk conditions, varying from the most optimistic to the most pessimistic scenario. This study only considered the modelling of physical vulnerability to an explosion hazard (one of the man-made hazards). Among various physical components, only the vulnerability of buildings has been of interest in this research. Vulnerability is a function of several criteria; therefore, this study used a multi-criteria decision making (MCDM) model. MCDM or multiple-criteria decision analysis (MCDA) is a subdiscipline of operations research that explicitly evaluates multiple conflicting criteria in the decision-making process. MCDM is concerned with structuring and solving decision and planning problems involving multiple criteria [8] . The difficulty of the problem originates from the presence of more than one criterion. However, many of the criteria related to vulnerability have a spatial nature, such that they are a function of the geographical positions of buildings. Hence, geospatial information systems (GIS) have a very high potential for storing, editing, and analyzing spatial data, and providing multiple applications for modelling vulnerability. Therefore, in this study, a combination of MCDM and GIS (GIS-based MCDM), as one of the most common models for vulnerability evaluation, is considered. Examples of spatial criteria include the distance from the facilities network and the distance from petrol pumps; these criteria and others are described in the criteria extraction section. According to the above description, the vulnerability of buildings is a function of their geographical position. Therefore, this study used spatial modelling for evaluating vulnerability. Since many people may be living in the buildings, when they are damaged the population can be seriously affected. Such damage will also have many secondary consequences, including human loss and social and economic vulnerability. Although the purpose of this study is modelling physical vulnerability, the results can be used for modelling other types of vulnerability, (e.g., social, economic and environmental). For instance, based on the number of residents in each building, we can estimate the vulnerable population and the resulting human casualties. GIS-based, multi-criteria decision analysis techniques have been used extensively to evaluate disaster risk, prioritize risks, and inform land-use planning. Studies employing such techniques are mainly focused on the vulnerability of cities to natural hazards such as floods and earthquakes [9-17]. Many investigations have been conducted regarding combining the GIS-MCDM models and fuzzy set theory [18] [19] [20] [21] [22] [23] [24] [25] . Risk management is an integral part of urban crisis management [26] and many studies have been undertaken on the application of GIS in crisis management and modelling vulnerability in different risk conditions [27] [28] [29] [30] [31] [32] [33] . Risk management uses various different techniques and models; in this research, ordered weighted averaging (OWA) has been employed as one of the commonly-used aggregation operators. In recent years, there has been much progress in this area and many researchers have used the OWA operator [34] [35] [36] [37] [38] [39] . As previously stated, use of certain techniques for estimating risks and losses related to man-made and technological hazards such as explosions and blast loads is limited (with respect to natural hazards). Arguing that it is important to enhance the blast resistance of civilian structures, Tadepalli and Mullen [40] proposed a model for risk assessment, damage analysis, and loss estimation. Their GIS model incorporates building data such as the type of structure, occupancy level, number of storeys, type of glazing, economic value of the building and its contents, and the maximum number of building occupants. Damage levels were quantified as low, moderate, or high. Once the degree of damage in a building due to a blast event was determined, the performance levels of the building were classified into levels for immediate occupancy, life safety, and collapse prevention. The loss due to a blast event was computed using a methodology similar to that used for seismic loss estimation. The premise was that the range of damage states which might occur due to a blast is similar to those which occur during earthquakes. The results of a simulation analysis conducted for buildings located at the University Sustainability 2017, 9, 1274 3 of 29 of Mississippi were visualized in the GIS environment. The results of this type of simulation study can be used for blast-hazard mitigation. Kulawiak and Lubniewski [41] used a web-based GIS to spatially visualize the impact of a simulated blast on a railway station in Gdansk, Poland. Results indicated that this technique can reveal vulnerabilities in terms of the extent to which buildings and critical infrastructure can be damaged, and the costs and injuries that may occur. Ma and Cheng [42] developed a GIS-based quantitative risk analysis (QRA) model and simulated the potential damage to buildings as a result of a natural gas pipeline explosion. The process incorporated an assessment of the failure rates of integrated pipeline networks, a quantitative analysis model of accident consequences, and an assessment of individual and societal risks. Firstly, the failure rates of the pipeline network were calculated using empirical formulas influenced by parameters such as external interference, corrosion, construction defects, and ground movements. Secondly, the impacts of accidents due to gas leakage, diffusion, fires, and explosions were analyzed by calculating the area subject to instances of poisoning, burns, and death. Lastly, based on the previous analyses, individual risks and social risks were calculated. The potential damage was divided into four categories, i.e., minor damage, minor structural damage, major structural damage, and partial demolition. The proposed GIS-based QRA system proved to be promising for use in planning for urban risk management of gas pipeline networks on a macro-scale. By combining GIS index modelling and risk assessment methods, Armenakis and Nirupama [43] developed a spatial risk assessment method to estimate the spatial risks of hazards caused by a propane explosion. Their work investigated the propane explosion that occurred on the 10 August 2008 at 3:30 a.m. in a Toronto neighborhood, as a case study to estimate spatial risk in the neighborhood around the explosion site using a GIS-based approach. The total vulnerability was estimated based on social, physical, and economic vulnerabilities. This research demonstrated the utility of risk assessment using GIS techniques for prioritizing spatial risks and estimating the potential impacts based on proximity to the blast source. Integrating such techniques into land-use planning is essential for disaster risk reduction in urban areas. Matijosaitiene and Petriashvili [44], by investigating explosion-related terrorist events in different countries, concluded that terrorists' choice of targets was firmly based on a series of criteria. Therefore, the researchers identified these factors via questionnaires, and the most effective factors were identified using automatic regression analysis. The authors demonstrated that crime prevention through environmental design principles-such as minimizing vehicle access points to buildings, separating public and private domains, and ensuring that buildings with identical functions in a given area are highly distributed-can have a significant impact on whether a site is likely to be selected as a target for terrorist attacks. On the other hand, research on the vulnerability of the built environment to blast and explosion is scarce in the Iranian context. Integrating GIS and analytic hierarchy process (AHP) techniques, Azizi et al. [45] evaluated the vulnerability of urban blocks in the 11th district of Tehran Municipality. Vulnerability was evaluated using 13 criteria. Ultimately, a vulnerability map was generated for the studied structural blocks, in which 8% of structural blocks had low vulnerability, 64% had medium vulnerability, and 28% had high vulnerability. Results also indicated that antiquity of structures and high structural density were the main factors inducing vulnerability in the buildings of Tehran's 11th district. This study was developed based on the hypothesis that vulnerability to air raids can be modelled in a similar way to vulnerability to earthquakes. Accordingly, no specific criterion relating to explosions was incorporated into the study. Other limitations of the study were that the model was developed based on certainty conditions (no inclusion of uncertainty factors) and there was no sensitivity analysis. Torabi and Mehdi Nezhad [46] modelled road network vulnerability (among other physical components of cities) in the 6th district of Tehran Municipality, Iran. The Delphi method was used for extraction of criteria, along with the IHWP method for allocating weights to the different criteria. After identifying 11 vulnerability criteria and their corresponding weights, criterion maps were combined using weighted linear combination (WLC) to generate a final vulnerability map. According to the vulnerability map, 8% of road networks were of low vulnerability, 55% had medium Sustainability 2017, 9, 1274 4 of 29
doi:10.3390/su9071274 fatcat:c5wk76x42rexplngepys4ex6cm