Sjoerd van Ballegooy

Earthquake Response of Underground Pipeline Networks in Christchurch, NZ

This paper explores key aspects of underground pipeline network response to the Canterbury earthquake sequence in Christchurch, New Zealand, including the response of the water and wastewater distribution systems to the MW6.2 22 February 2011 and MW6.0 13 June 2011 earthquakes, and the response of the gas distribution system to the MW7.1 4 September 2010 earthquake, as well as the 22 February and 13 June events. Repair rates, expressed as repairs/km, for different types of pipelines are evaluated relative to (1) the spatial distribution of peak ground velocity outside liquefaction areas and (2) the differential ground surface settlement and lateral ground strain within areas affected by liquefaction, calculated from high-resolution LiDAR survey data acquired before and after each main seismic event. The excellent performance of the gas distribution network is the result of highly ductile polyethylene pipelines. Lessons learned regarding the earthquake performance of underground lifeline systems are summarized.

Probabalistic Prediction of Severity of Liquefaction Surface Manifestation Using Geotechnical and Geospatial Models

The severity of liquefaction manifested at the ground surface is a pragmatic proxy of damage potential for various infrastructure assets, making it particularly useful for hazard mapping, land-use planning, and preliminary site-assessment. Towards this end, the recent Canterbury, New Zealand, earthquakes, in conjunction with others, have resulted in liquefaction case-history data of unprecedented quantity and quality, presenting a unique opportunity to develop fragility-functions for liquefaction-induced ground failure. Accordingly, this study analyzes nearly 10,000 liquefaction case studies from 23 earthquakes to develop functions that express the probability of exceeding specific severities of liquefaction surface-manifestation as a function of five different liquefaction damage measures (LDMs), of which three are based on geotechnical data and two are based on freely available geospatial data. The proposed functions have the same functional form, such that end-users can easily select the model coefficients for the particular damage state and LDM of their choosing. It should be noted that these functions are not to be used to predict lateral spreading, which requires LDMs other than those assessed herein. Lastly, the proposed functions are preliminary and subject to further development. In this regard, several thrusts of ongoing investigation are mentioned.