Impact of Rock Heterogeneity on Reactive Flow During Acid Stimulation Process

Cunqi Jia,Saif Alkaabi,Jinchuan Hu,Kamy Sepehrnoori,Jun Yao,Lei Zhang
DOI: https://doi.org/10.1016/j.ijheatmasstransfer.2024.125560
IF: 5.2
2024-01-01
International Journal of Heat and Mass Transfer
Abstract:Acid stimulation within carbonate rocks represents a crucial aspect of reactive flow applications in subsurface porous media and has been widely used in carbonate reservoir development attributed to its rapid effectiveness and cost-efficiency. This study is motivated to investigate further the impact of rock heterogeneity on the acid stimulation process. The two-scale continuum model is employed for capturing mass, momentum, and energy transfer during the acid stimulation process. The core-scale model is expanded to consider the specific composition of solid minerals, including reactive minerals such as calcite and dolomite and non-reactive ones like quartz and clay. A heterogeneity generator is developed by coupling with the open-source geostatistical program to produce different heterogeneous parameters using random seeds with varying spatial correlation lengths. Sensitive cases are performed to present an investigation into the influence of mineral composition and matrix heterogeneity on the acid stimulation process. Results show that the typical results of acid stimulation are closely tied to different heterogeneity conditions. The optimum conditions for acid stimulation are sensitive to the concentration of reactive components. The optimum acid injection rate declines as dolomite content increases, while the corresponding acidizing breakthrough volume reduces with higher calcite levels. The presence of mineral layers, especially in horizontal distributions, complicates acid fluid efficiency due to the competitive reaction between calcite and dolomite. Compared to the heterogeneous distribution of minerals, matrix heterogeneity plays a critical role in acid fluid consumption efficiency. Increasing heterogeneity leads to significant reductions in breakthrough volumes, largely governed by the correlation length.
What problem does this paper attempt to address?