No abstract

No abstract

This paper presents an investigation on damage to natural gas pipelines in Taichung City from the Chi-Chi, Taiwan, earthquake. This paper addresses damage due to ground shaking effects and does not address damage due to large ground deformations. Four parameters, that is, peak ground acceleration, peak ground velocity, Arias intensity, and spectral intensity, are used to represent ground shaking. Based on pipe repair data and recorded strong motion data, regression analyses of pipe repair rates were carried out to develop seismic vulnerability functions. From the regression analysis results, Arias intensity is considered as the best parameter for the derivation of seismic vulnerability function. The seismic vulnerability functions derived in this study are for steel gas pipelines with mechanical joints and the pipelines are located in firm soils and in the footwall area subject to ground shaking from an earthquake caused by a thrust fault.

Substantial damage to water supply systems, including water delivery pipelines, water treatment plants, reservoirs, and water storage tanks, was reported after the 1999 Chi-Chi Taiwan Earthquake. This paper first summarizes the damage survey and then presents the results of seismic fragility analysis for underground pipelines. Construction blueprints of the water delivery pipelines and repair work orders of 11 townships and cities in the disastrous area were digitized into a Geographical Information System (GIS) for analysis and assessment. With the aid of the GIS system, we found that PVC pipes made up 86% of water delivery pipelines while steel, cast iron, ductile iron, PE and others took the rest. Therefore, this paper focuses on the fragility analysis of PVC pipes. Three different methods were applied to derive the fragility relations between the PVC water pipes having nominal diameters (approximately inner diameters) greater than or equal to 65 mm and earthquake intensity parameters such as peak ground acceleration and peak ground velocity. The results were then examined with those of other countries. The discrepancy between our results and the empirical equation used by HAZUS, an earthquake loss estimation software developed by the Federal Emergency Management Agency was not significant.

No abstract