A case study on the variability of the coefficient of consolidation and its design reliability

Abstract

The consolidation characteristics of cohesive soils are estimated using established relationships between the coefficient of consolidation (cv) and index tests, as well as laboratory oedometer tests. While the design cv is preferred from the field dissipation tests, the conversion from a horizontal to vertical value needs to be considered. A trial load was used to verify the consolidation parameters during a Queensland Road upgrade, which involved both road widening and raising of the existing embankments over compressible soils. Construction was done in 4 stages, and with preloading and surcharging in selected areas. Settlement monitoring and Asaoka plots were used to validate the design, and “moderately conservative” design values were adopted. This case study is used to show the large variability of the cv by the various test methods. While 99% of the site settlement was within the magnitude and time predicted during design, a 25 m length was not consistent with the data and performance of the rest of this site within the flood plain. The back-calculated cv was below the lowest test value and even data from nearby settlement plate monitoring from adjacent stages. In situ tests were located within 25m of this unconforming area and given that stratigraphy was consistent then the cv value adopted may not be representative. The lessons learnt show the various verification and validation process required to envelope risks, but all conditions with a “moderately conservative” design may not be covered.