Yong Fu

Natural gas hydrate (NGH) are a promising unconventional energy resource, but reservoir parameter uncertainty significantly affects gas production and formation deformation predictions. Focusing on spatial permeability variability, this study constructs conditional permeability random fields constrained by field data and develops a fully coupled thermal-hydraulic-mechanical-chemical (THMC) model for depressurization-based NGH exploitation. The model is solved with a fully implicit finite-element formulation, incorporating dynamic porosity evolution due to hydrate dissociation and skeleton deformation. Permeability structures are characterized by different coefficients of variation (COV) and spatial correlation lengths. Results show that most stochastic realizations yield higher cumulative gas production than the deterministic model. Changes in correlation length generally induce larger dispersion in cumulative gas production and subsidence than changes in COV, indicating the strong control of permeability spatial structure. Spatial descriptors, i.e., near-well average permeability, high- and low-permeability area proportions, and connected high-permeability structures, are constructed and correlated with responses. Path-resistance descriptors showed stronger associations with gas production and subsidence than near-well mean permeability. Evolution of pore pressure, gas saturation, and hydrate saturation further indicates that effective connectivity between preferential pathways and the wellbore explains response differences better than the high-permeability area proportion alone. The proposed field-data-constrained conditional random field-THMC-Monte Carlo framework provides a probabilistic basis for productivity evaluation, deformation risk assessment, and optimization of NGH depressurization.

The installation of large-diameter monopiles for offshore wind foundations faces growing risks from pile running, which refers to a sudden and uncontrolled penetration event causing substantial safety hazards and economic losses. This risk is amplified by the spatial variability of seabed soils, which is often oversimplified in deterministic design, leading to non-conservative predictions. This study proposes a probabilistic framework to quantify the influence of soil spatial variability on pile running risk. By combining the Coupled Eulerian–Lagrangian method for large-deformation analysis with random field theory and Monte Carlo simulations, we model pile running in spatially variable clay, incorporating strain-softening and rate effects. Results show that soil randomness alters soil flow mechanisms, strongly controlling penetration velocity and resistance. Deterministic analysis underestimates the maximum running velocity, which follows a lognormal distribution. The framework establishes a direct link between failure probability and design safety factor, recommending design safety factor ≥ 1.5 to achieve a failure probability below 1%, in line with major offshore standards. Application to a Zhuhai offshore wind farm case study confirms the framework’s practicality and highlights the critical role of stratigraphic uncertainty. This work advances reliability-informed design by explicitly accounting for soil spatial variability to improve safety in monopile installation.

This study introduces a dynamic model analysis framework to simulate the hydrodynamic installation of dynamically installed anchors (DIAs) under the coupled effects of ocean current and operational mooring lines. The key innovations of this framework lie in its dynamic evaluation of the anchor's directional stability during descent and its integrated treatment of environmental and operational mooring loads. Unlike traditional static approaches, the proposed dynamic model updates the hydrodynamic forces, center of gravity, and applied moment of DIAs at each time step through a six-degree-of-freedom motion calculation module, thereby enabling continuous tracking of the restoration moment coefficient and instantaneous inclination angle. The MoorDyn library, based on the lumped-mass method, is applied to simulate mooring dynamics, and the overset grid technique is employed to facilitate the transfer of information related to fluid–structure interaction. The framework yields critical performance indicators, including directional stability, inclination angle, offset distance, impact velocity, and drag coefficient under the coupling influence of ocean currents and mooring lines. Validation cases confirm the model's robustness and reliability. The verified framework is subsequently applied to investigate the installation characteristics of torpedo- and plate-shaped dynamically installed anchors, with a specific focus on the impact of mooring line configurations and current-line coupling effects on the torpedo anchor performance. Results show that the mooring line reduces the falling velocity of the DIAs and increases their drag coefficient. Current action increases the inclination angle and offset distance of the DIAs, affecting their directional stability. The coupling effect between the two further exacerbates this adverse impact.