Stochastic analysis on the lateral buckling of a HPHT pipeline considering the spatial variability of seabed

Complex geological processes and depositional environments could bring great spatial variability of the engineering properties of seabed sediments, which may further cause uncertainty in the global buckling predictions of high-pressure high-temperature (HPHT) pipelines. Unlike previous investigations with the deterministic method, a stochastic finite element (SFE) model is established for predicting the lateral buckling of a pipeline considering the spatial variability of the seabed. A flow chart of the Monto Carlo simulation for the lateral buckling of pipeline is provided. The SFE results for the simplified homogeneous seabed are compared with the existing analytical solutions for model verifications. Monte Carlo simulation results indicate that the mean values of the critical temperature rise for triggering the lateral buckling of a pipeline on the spatially varied seabed are generally larger than the deterministic predictions. It is found that the critical temperature rise may reach its maximum value when the horizontal scale of fluctuation ranges from the value of the discrete-element size to that of the imperfection wavelength. As the horizontal scale of fluctuation increases, the buckling probability increases significantly, which could be up to about 50% while the horizontal scale of fluctuation approaching a quite large value or even infinity.

Automated summary

The spatial variability of seabed sediments can cause uncertainty in global buckling predictions of HPHT pipelines due to complex geological processes and depositional environments. A stochastic finite element model is used to predict lateral buckling of pipelines considering the spatial variability of seabed sediments. Monte Carlo simulations show that the mean values of critical temperature rise for triggering lateral buckling are generally larger than deterministic predictions on spatially varied seabeds. The buckling probability significantly increases as the horizontal scale of fluctuation increases, reaching up to about 50% for large values or even infinity.