The Lusi mud eruption in East Java has been active since May 2006. Magma emplacement at depth, clay dehydration, and mud liquefaction during seismic wave propagation have been invoked as mechanisms fueling this eruption. However, the respective roles of these processes are still poorly constrained. In this focused study, we numerically investigate the influence of clay dehydration, mass and heat transport on fluid outflow at the Lusi site using a fully coupled 3D model for this active system. Using a multi-GPU parallel processing algorithm, we propose an estimate of the 3D time evolution of pressure, temperature, porosity, permeability and water liberation in a large-scale (9 km - 14 km - 5.5 km) deep hydrothermal system at high-resolution. Simulations indicate that high-pressure fluids generated by dehydration reactions are sufficient to induce hydro-fractures that would significantly influence the porosity and permeability structures. Dehydration is an essential component for understanding the Lusi system, because the fluids generated contribute to the outflow and may have a considerable impact for the maintenance of the infrastructure required to keep the Lusi site safe. High-Performance Computing (HPC) offers high-resolution simulations for studying time evolution of such natural systems, and potentially for geothermal resource development for the surrounding population.

东爪哇鲁西泥浆喷发自 2006 年 5 月以来一直活跃。深部岩浆侵位、粘土脱水和地震波传播过程中的泥浆液化被认为是推动这次喷发的机制。然而，这些过程各自的作用仍然受到很大限制。在这项重点研究中，我们使用该活动系统的完全耦合 3D 模型，数值研究了粘土脱水、质量和热传输对 Lusi 场地流体流出的影响。使用多 GPU 并行处理算法，我们提出了一个大规模（9 公里 - 14 公里 - 5.5 公里）深部热液系统中压力、温度、孔隙度、渗透率和水释放量的 3D 时间演化估计。解析度。模拟表明，脱水反应产生的高压流体足以引发水力压裂，从而显着影响孔隙度和渗透率结构。脱水是了解 Lusi 系统的重要组成部分，因为产生的流体有助于流出，并且可能对维护 Lusi 站点安全所需的基础设施产生相当大的影响。高性能计算 (HPC) 提供高分辨率模拟，用于研究此类自然系统的时间演变，并可能为周围人口开发地热资源。因为产生的流体会导致外流，并且可能对维护 Lusi 场地安全所需的基础设施产生相当大的影响。高性能计算 (HPC) 提供高分辨率模拟，用于研究此类自然系统的时间演变，并有可能用于周围人口的地热资源开发。因为产生的流体会导致外流，并且可能对维护 Lusi 站点安全所需的基础设施产生相当大的影响。高性能计算 (HPC) 提供高分辨率模拟，用于研究此类自然系统的时间演变，并有可能用于周围人口的地热资源开发。