Towards an integrated risk assessment: a method for understanding na-tech in urban areas

Territorial systems have often to face more than one hazard. In these cases, aiming at safer communities, a multi-risk assessment is required. The need to take into account in land use planning all the hazard factors an area is prone to has been declared since the early ‘80s. Nevertheless, the all-hazard approach has been often reduced to the “overlapping” of individual hazard maps: this can effectively support the singling out of “critical areas”, prone to more than one hazard, but it is generally ineffective to describe the complex interactions among different hazards and their impacts. Complex hazardous events could depend both on the spatial coincidence of individual hazards, which may occur separately in time, and on interactions among different hazards, which occur simultaneously or in strict connection. Namely, this kind of events could be composed by linked natural hazards or technological accidents triggered by natural hazards. The scientific community is paying increasing attention to these latter, the so-called na-tech events (Showalters and Myers 1992). This work focuses on na-tech events suggesting a methodology to “understand” and, consequently, to prevent and mitigate them. A more in-depth attention will be paid on technological accidents triggered by earthquakes, such as major accidents (as defined in the European Commission Directive 96/82/EC) and interruptions of lifelines systems. The earthquake occurred in the Turkish Kocaeli industrial region, in 1999, is one of the most documented na-tech event. This accident clearly highlights that, in case of linked hazards, interactions among them modify the coping capacity of the hit area. This complexity could not be faced just “overlapping” individual risk maps and so undervaluing the vulnerability, and consequently overvaluing the resilience, of the exposed area. Many are the differences between facing a single hazard and a na-tech. First of all, as stated in many studies (Lindell and Perry 1996; Cruz 2005), facing at the same time a natural disaster and a technological emergency requires more resources and a specific training for emergency personnel. Besides that, it is probable to have shortage of utilities (interruption of water, power and communication supply) and lack of personnel and infrastructure. Moreover, technological accidents that could be triggered by a natural event are usually more than one, as it normally happens in case of a major accident in ordinary time, and, besides that, they could probably be much more different than a “ordinary” major accident for magnitude, typology and triggering point. Hence, to be efficacious and to represent an effective supporting tool for the definition of mitigation actions, especially at local scale, multi-risk and, in detail, na-tech risk assessment have to cope with synergies among hazard factors and their complex impacts, taking into account possible chains of hazardous events and their effects on the vulnerability of exposed areas. Nevertheless, up to now, natural and technological hazards have been mainly faced separately and no methodologies are available to face na-techs, which represent one of the most spreading complex hazardous event. As Menoni states (2005), urban planners could contribute to understand disasters with their attitude to consider more systems at the same time and to look at their mutual relationships. Nevertheless, they have the capacity to deal with systems, relations and objects evolving tridimensionally in the space. This skill is more needed if dealing with complex disasters, and, namely, with na-techs, where environmental, physical, functional and social systems are closely connected in chains of failures and damages. Basing on a systemic approach, this paper provides a methodology to analyze na-tech risk and to define urbanistic mitigation strategies against na-tech events. A method both for provincial and local scale has been defined. In detail, the “overlapping” of natural and technological risk factors maps is adopted to quickly single out the most critical areas on a provincial scale. A multi attribute decision making method is applied to rank each municipality of the investigated province, according to attributes referred to their exposure, vulnerability and hazard features. At local scale, a scenario setting method is worked out in order to forecast the possible chains of events and to define na-tech mitigation strategies for the critical areas singled out at provincial level. The method is structured in three main phases: the identification of elements to be involved in the scenario and their mutual relations; the singling out of the scenario key elements on which damages depend; the definition of mitigation strategies. Elements belonging to environmental, physical, functional and social system and their relations are singled out basing on scientific literature and case studies (Turkey, Northridge and Kobe earthquake, etc.) and on civil protection emergency plans and emergency fact sheets for what concerns natural and technological disasters. Case studies and literature help figuring out elements that have effectively been involved in chains of failure and damages during past na-techs and their vulnerability features. In addition, the study of civil protection emergency plans and fact sheets gives information on which are elements (e.g. roads, communication networks) people have to interact with in case of each single emergency (both to the technological accident and to the triggering natural hazard). The comparison among fact sheets allows to figure out which are the behaviours that population should be called to observe in case of a single hazard, but that, otherwise, could expose them to the other occurring hazard, in case of na-tech. Once elements and relations have been defined, they are structured in a cognitive map based scenario. A cognitive map is a support tool to set up and to explore the scenario. Through the analysis of hierarchies and relations of the map, it is possible to rank elements and to recognize which are the key ones from which the most of the other elements depend on. Moreover, it is possible to highlight hidden indirect connections among elements and the main chains of failures and damages during a na-tech event that, otherwise, should not be easy recognized. In this way, it is possible to select elements on which to concentrate resources to maximize the effectiveness of mitigation actions and to avoid triggering of failures and damages, improving the resilience of the whole urban system. Actions are grouped in an abacus and classified if they refer to environmental, physical, functional and social system and if they are mainly oriented to land uses modification, variation of the intensity of uses and to actions aimed at improving accessibility during the emergency phase. Different mitigation strategies can be defined. For instance, we could chose to implement actions on the elements belonging to one of the main chains of failures and damages or it is possible to chose to accomplish only those referring to key elements. Outcome-oriented scenarios are set up to verify the efficaciousness of strategies and to sort out the most appropriated ones. To support findings, the method has been tested on the Province of Naples (Campania Region) and on a middle size Municipality, classified as 2nd class seismic area (Ordinance 3274/2003) and affected by the presence, within the city core, of a LPG storage plant, classified as dangerous by the Seveso II Directive (art. 9).