A p-y Analysis of Laterally Loaded Offshore-Well Conductors and Piles Installed in Normally Consolidated to Lightly Overconsolidated Clays

Abstract

This paper presents a cyclic equivalent spring (p-y) model describing lateral soil resistance on offshore well conductors and piles in normally to lightly overconsolidated clays subjected to fatigue cyclic motions. Improved understanding and characterization of cyclic lateral soil response is critical in fatigue assessment of well conductors and piles subjected to dynamic fatigue loads. Key features of the model include nonlinear load-displacement behavior with stiffness degradation during cyclic loading. The model provides a full description of soil resistance during lateral loading, including an initial short-excursion monotonic loading stage, a transient stage of progressive degradation in stiffness from the first excursion, and a steady-state stage involving minimal changes in soil stiffness after a large number of load cycles. The model input parameters obtained from back-analysis of data derived from centrifuge tests on model conductors subjected to harmonic lateral loads are presented in this paper. This model has capabilities of simulating random load sequences. Fatigue damage in well conductors and piles arises from changes in axial and bending stresses, with the latter being more dependent on lateral soil response. Accordingly, the proposed model is evaluated primarily in terms of its ability to accurately predict bending moments when the spring model is used in conjunction with a laterally loaded soil-structure interaction model. The model successfully predicts the maximum change in cyclic bending moment (change in moment during a load reversal) and the location of the maximum cyclic moment along the conductor depth approximately within a 20% range. This performance evaluation was derived by comparing the model computations/predictions to the test data from the centrifuge program and validated within the tests displacement range of 0.1D. The current form of the model does not consider consolidation effects, which may significantly affect long-term loading predictions used in fatigue life assessments.