Keywords: industrial facilities, earthquake risk assessment, seismic vulnerability assessment, dynamic analysis, vulnerability analysis.

Due to the increasing urbanization and industrialization as well as the increasing susceptibility and capital intensity of modern technologies in industrialized countries, earthquakes cause ever increasing losses in such regions. Structural damage to the industrial facilities can also trigger consequential fire, explosions or the release of dangerous materials which leads to further casualties and financial loss. Moreover, suspension of production especially that of water and energy supply facilities, might magnify the dimensions of the catastrophe and impose more financial loss. Therefore, evaluating seismic vulnerability of industrial facilities is of great importance. Reliable vulnerability assessment methods are the sine qua non for predicting the likely risk of loss. In recent years, developing methods for vulnerability analysis were the subject of numerous research activities. But as a result of the heterogeneous configuration of industrial plants a systematic procedure for the assessment of their vulnerability has not been provided yet.

In this present paper we introduce a systematic procedure for the vulnerability analysis of industrial facilities at different levels of accuracy. The first step is acquiring a classification of industrial facilities by exploiting the classifications existing in technical literature and rearranging them, if necessary, with respect to similarities in the seismic vulnerability and subsequent losses. The facility classes will be further subdivided into individual components. Typical components of industrial plants are silos, tanks, pipelines and so forth. Though some components, e.g. production units of chemical industries are a complex mixture of steel or concrete frames and multiple installations such as tanks, reactors, etc., rather than a set of discretizable components. Thus, components of each industry type should be selected so that they are general enough to be employed for modeling similar facilities, yet in a way that their integration truly represents the target facility.

The next step is to evaluate the vulnerabilities of these components through appropriate damage curves or matrices, which can be extracted either from existing empirical evaluations or calculated by simulations. In this sense we can imagine three levels of accuracy in which the first level is fastest and requires the minimum amount of input data while Level III demands a detailed precise analysis of components or even the entire facility. The overall vulnerability of the facility is then determined by superposing the vulnerabilities of individual components as a result of the configuration of the industrial plant and the importance factors of individual components.

This paper does not really cope with secondary and socio-economic losses of earthquake damage to industrial facilities, yet they have been considered to a certain extent through the significance factors of components. Essential input parameters, particularly at the first and second level of precision should be determined preferably with a simple and short questionnaire but they could also be extracted from aerial photos. The vulnerability relations of industrial facilities will be further incorporated to an already developed Earthquake Risk Assessment System [1], which couples the vulnerability of desired structure with the site hazard likelihood to compute the likelihood distribution of loss.

We have performed the aforementioned procedure for the case of chemical facilities in level I. As it was indicated before, the vulnerability analysis at this level is not so precise but it is fast and requires the minimum of input data. Hence, it is quite useful for insurers and reinsures as well as strategic planning for loss mitigation. Vulnerability of individual components has been extracted from the respective technical sources [2,3]. Nonetheless, as pointed before, no vulnerability curves or matrices are available for the production units and thus its vulnerability has been investigated at two levels. First a simple "screening and scoring" approach has been introduced referring to the established standards [4] in which an on-site inspection of the unit and a scoring of the quality of influential features of the unit would lead to an appropriate estimation about its safety and whether a more detailed investigation is required. Furthermore an approach for investigation at level II for acquiring the corresponding vulnerability relation based on simple simulations has been presented.

1

I. Karimi, E. Hüllermeier, "A Fuzzy-Probabilistic Risk Assessment System for Natural Hazards", Proc. of 11th International Fuzzy Systems Association (IFSA) Congress, 2005.

2

Applied Technology Council, ATC 13, "Earthquake Damage Evaluation Data for California", 1985.

3

Federal Emergency Management Agency, "HAZUS MH-Technical Manual", FEMA, Wachington, D.C., 2003

4

Federal Emergency Management Agency, FEMA 154, "Rapid Visual Screening of Buildings for Potential Seismic Hazards: A Handbook", Washington D.C., March 2002

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £135 +P&P)