Using Field Data to Improve the Settlement Prediction Model of a Breakwater on Soft Soil

The need for an improved harbor access resulted in the construction of two new rubble mound breakwaters in Ostend, Belgium. In the past, a specific dredging technique was used to maintain the required depth of the access channel to the harbor. This has resulted in a complex soil layering underneath the northern part of the new western breakwater with locally very thick soft soil layers. As an alternative for an expensive soil remediation, a very strong geotextile was included into the breakwater core, reinforcing the structure. In addition a staged construction was necessary, allowing the soft layers to consolidate and preventing undrained failure during construction. Because of the complexity of the situation, a numerical model was used to get an estimate of the settlements and the consolidation process during and after construction of the breakwater. Extensive field measurements of the settlements were performed during construction and are still performed at this moment (early 2013). Comparison of the settlement measurements with the calculations results in a better understanding of the consolidation process and allows improvement of the model, resulting in a more accurate settlement (and thus consolidation) prediction curve. Because of the improved insight in the consolidation of the soft layer, a substantial gain in the progress of the construction of the dam is achieved. This paper shows, based on a case study of a breakwater constructed on soft soil, that existing models used to predict settlements can be improved using field data, leading to a clear benefit regarding the construction phasing.