Scholarlink

<p>Fast geomechanical simulations of fault reactivation induced by the exploitation of porous subsurface reservoirs play an important role in (upfront) screening, seismic hazard assessment, and data assimilation approaches. Here, a 2D semi-analytical model framework is presented to compute stress changes on faults offsetting two depleting reservoir compartments, building on existing closed-form expressions for stress changes. The model was extended by incorporating a simplified vertical pressure diffusion function to the seal and base formations. The closed-form expressions were further extended to include thermo-elasticity and vertical temperature diffusion. The model was applied to generic depleting and cooling reservoir offset by a fault, and the effect of diffusion on different fault reactivation metrics was assessed. It was shown that diffusion regularized the stress distribution along the fault, smoothing the stress concentrations at the reservoir boundaries and eliminating cell-size dependence of the results. It was also shown that the assumed depletion of the fault zone itself had a fundamental effect on the stress path, with lower fault zone depletion than the reservoir compartments leading to more unstable fault stresses. Depending on the assumed fault zone depletion, cooling can have a similar (in case of no fault zone pressure change) or fundamentally different (in case of full fault zone depletion) stressing effect compared to depletion. In the latter case, cooling resulted in higher stressing rates due to the reduction in effective normal stress. Furthermore, we used a gradient-boosted machine-learning tree regression to fit the model results, creating a surrogate model that was used to rank the relative influence of input parameters on fault reactivation metrics. For stressing rates the most important parameters were reservoir height, fault throw, and the poro-elastic or thermo-elastic parameters, but for reactivation metrics relating to the absolute state of stress on the fault, the initial horizontal stress, fault dip, and fault strength became the dominant parameters. Both the semi-analytical and the surrogate model are useful tools for probabilistic screening methods or hazard assessment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

2D Semi-Analytical Modeling of Fault Reactivation: The Effect of Pressure and Temperature Diffusion into the Seal and Base Formations

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月30日

查看全文
  • Loes Buijze

摘要

Fast geomechanical simulations of fault reactivation induced by the exploitation of porous subsurface reservoirs play an important role in (upfront) screening, seismic hazard assessment, and data assimilation approaches. Here, a 2D semi-analytical model framework is presented to compute stress changes on faults offsetting two depleting reservoir compartments, building on existing closed-form expressions for stress changes. The model was extended by incorporating a simplified vertical pressure diffusion function to the seal and base formations. The closed-form expressions were further extended to include thermo-elasticity and vertical temperature diffusion. The model was applied to generic depleting and cooling reservoir offset by a fault, and the effect of diffusion on different fault reactivation metrics was assessed. It was shown that diffusion regularized the stress distribution along the fault, smoothing the stress concentrations at the reservoir boundaries and eliminating cell-size dependence of the results. It was also shown that the assumed depletion of the fault zone itself had a fundamental effect on the stress path, with lower fault zone depletion than the reservoir compartments leading to more unstable fault stresses. Depending on the assumed fault zone depletion, cooling can have a similar (in case of no fault zone pressure change) or fundamentally different (in case of full fault zone depletion) stressing effect compared to depletion. In the latter case, cooling resulted in higher stressing rates due to the reduction in effective normal stress. Furthermore, we used a gradient-boosted machine-learning tree regression to fit the model results, creating a surrogate model that was used to rank the relative influence of input parameters on fault reactivation metrics. For stressing rates the most important parameters were reservoir height, fault throw, and the poro-elastic or thermo-elastic parameters, but for reactivation metrics relating to the absolute state of stress on the fault, the initial horizontal stress, fault dip, and fault strength became the dominant parameters. Both the semi-analytical and the surrogate model are useful tools for probabilistic screening methods or hazard assessment.