Advances in High-Performance Computing for Static and Dynamic Modelling

Abstract In recent years, the oil and gas industry has increasingly leveraged high-performance computing (HPC) to tackle complex reservoir simulations. These simulations are crucial for understanding subsurface conditions, optimizing hydrocarbon recovery, and ensuring efficient resource management. Among the various simulation techniques, compositional modeling stands out due to its ability to capture detailed fluid behavior and interactions within the reservoir. This type of modeling is particularly valuable for predicting the behavior of multi-component fluid systems, which are common in many reservoirs. One of the primary challenges in compositional modeling is the need for high-resolution grids. High-resolution grids enable more accurate representation of geological features and fluid dynamics but come with significant computational demands. As the resolution increases, so does the number of active grid cells, which can quickly escalate into the billions. Managing such large-scale models requires not only advanced software capable of handling vast amounts of data but also powerful hardware to process and visualize the results efficiently. This study focuses on the development and implementation of a high-resolution compositional dynamic model with almost half billion active grid cells. The primary aim is to explore the capabilities and limitations of current software and hardware in managing these large-scale models.