Assignment of Groundwater Table in Liquefaction Analysis of Soils

This paper discusses the issue and potential influences of miss-assignment of groundwater table in the analysis of soil liquefaction. In viewing that the groundwater table (\( \varvec{GWT}_{0} \)) during subsurface exploration or testing for evaluating cyclic resistance ratio (\( \varvec{CRR} \)) of soils is sometimes mistakenly assumed the same as the groundwater table (\( \varvec{GWT} \)) for computing cyclic stress ratio (\( \varvec{CSR} \)) due to seismic shaking, the results of liquefaction analysis may thus be erroneous. If the \( \varvec{GWT}_{0} \) is assigned higher than the actual level, the \( \varvec{CRR} \) and the associated factor of safety (\( \varvec{F}_{\varvec{L}} \)), would be overly predicted. Alternatively, if the \( \varvec{GWT}_{0} \) is assigned lower than its actual one, the \( \varvec{CRR} \) and \( \varvec{F}_{\varvec{L}} \) would be underestimated. If the groundwater table during exploration or testing is mistakenly assigned the same as the groundwater table for computing cyclic stress ratio (i.e., \( \varvec{GWT}_{0} = \varvec{GWT} \); or “one-groundwater-table, \( \varvec{OGT} \), scenario”), then the variation in the groundwater tables will lead to the changes in \( \varvec{CRR} \) and \( \varvec{CSR} \) in the same sense. Owing to different rates of change, however, the computed factor of safety (\( \varvec{F}_{\varvec{L}} = \varvec{CRR}/\varvec{CSR} \)), and the associated liquefaction potential index (\( \varvec{LPI} \)), may sometimes result in an unexpected situation. Namely, a rise in the groundwater tables would cause an unanticipated increase in the computed factor of safety and a decrease in the associated liquefaction potential index. Based on results of current study with assumption of \( \varvec{OGT} \) scenario, the \( \varvec{LPI} \) could be reduced by 10–30% if \( \varvec{GWT}_{0} \) is 3 m higher than the actual level; or alternatively, the \( \varvec{LPI} \) would be increased by 5–45% if \( \varvec{GWT}_{0} \) is 3 m lower than the actual one.