Because of their relatively low shear strength, high compressibility, and complex dynamic response, soft clay deposits routinely pose challenges to geotechnical engineers interested in improving the load-carrying capacity of deep foundations. Beyond increasing the dimensions of the foundation, the primary techniques that have been used to enhance the behavior of soft clays include preconsolidation with surcharge loading, electro-osmosis, or installation of vertical drains. While these soil improvement techniques have been shown to work well for clay deposits on land, they may be difficult to apply in offshore clay deposits. To address this issue, this study investigates a comprehensive approach to improve the properties of soft clays through in-situ heating to enhance the pullout capacity of an offshore foundation. In particular, this study will focus on two offshore foundations that incorporate internal heating devices – a cylindrical pile jacked into place and a torpedo pile installed under self-weight.

Soil improvement using in-situ heating in saturated, normally consolidated clays is a complex coupled thermo-hydro-mechanical process. Increases in pile temperature will leads to the generation of thermally-induced excess pore water pressure due to the differential thermal expansion of the clay particles and water. The generation of these pore water pressures will lead to a flow of water away from the heat source, leading to time-dependent, elasto-plastic volumetric contraction (consolidation). A further decrease in void ratio is expected during pile cooling due to elastic contraction. The decrease in void ratio of the clay due leads to an increase in undrained shear strength of the clay, and a corresponding increase in the pullout capacity of the pile.

To study this coupled problem, an empirical model was developed and applied to predict the distribution of thermally-induced excess pore water pressure during undrained heating of a saturated clay layer. This model was combined with a conduction analysis to estimate the changes in pore water pressure as a function of radius and depth in the clay layer surrounding a heated pile. Next, a poro-mechanics model was developed to estimate the amount of volumetric contraction of the clay as a function of the magnitude of pore water pressure. An experimental correlation between the undrained shear strength and the void ratio based on results from triaxial tests was then used to estimate the pullout capacity of the pile. Finally, a series of geotechnical centrifuge physical modeling experiments were performed to evaluate the different heat transfer, water flow, and volume change processes in a clay layer surrounding a heated pile, followed by measurement of the changes in pullout capacity of the pile. The pullout capacity estimated using the empirical models and undrained shear strength correlations were found to match well with the pullout capacities measured in the geotechnical centrifuge physical modeling experiments.