Publish and Present
The Port of Alaska has embarked on a multi-phase modernization program that includes the development of a new Petroleum and Cement Terminal (PCT) Petroleum Oil and Lubricants (POL) Berth 1. The new terminal is being designed for a minimum 75-year service life as an Essential Facility, critical to maintaining immediate post-earthquake serviceability to support the region’s post‐earthquake response.
The new terminal is located in an aggressive marine environment with numerous design challenges including a magnitude 9.3 subduction zone design earthquake, weak foundation soils, and shoreline conditions that are vulnerable to large-scale seismic displacements. This paper will present generalized subsurface conditions, in situ and cyclic laboratory testing data, and numerical modeling results which consisted of nonlinear, effective stress, time history analysis of the proposed pile supported terminal structure and ground improvement zone.
The subsurface conditions at the Port of Alaska PCT site predominately consist of liquefiable, non-plastic, fine grained Tidal Silt deposits and the Bootlegger Cove clay formation (BCF), which can exhibit significant stiffness and strength reduction under cyclic loading. The dynamic behavior of these fine grained soils was evaluated with cyclic direct simple shear (CDSS) testing which was used to calibrate soil behavior in the 2D numerical model by using pore pressure generation models for the Tidal Silt and an iterative strength reduction scheme for the BCF. Several ground improvement alternatives were evaluated with deep soil mixing (DSM) proving to be the most effective at reducing kinematic loads on the pile foundations. The analysis and design of the DSM required close coordination with the structural model which applied ground displacements into p-y soil springs to develop kinematic loads. The interaction of the pile and DSM panels required modification to the p-y springs within the ground improvement zone which accomplished using a local pile and DSM numerical model to determine p-multipliers.