Quantitative evaluation of hydrothermal fluids and their impact on diagenesis of deep carbonate reservoirs: Insights from geochemical modeling

Abstract

Hydrothermal dolomitization is an important diagenetic process that affects the quality of carbonate reservoirs. While geochemical modeling is a useful tool to improve our understanding of this process, comparatively little research has been focused on producing reservoir intervals. This study examines deeply buried (over 4500 m deep), Lower Ordovician dolomitized limestones (more than 50 m revealed by the borehole) in an exploration well in the Tazhong Uplift, Tarim Basin. The well penetrated a fault zone that was affected by extensive hydrothermal activity during the Permian. Batch geochemical modeling and reactive transport modeling (RTM) were used to evaluate the potential sources of hydrothermal fluid and the associated mechanisms that resulted in a relatively restricted distribution of dolomitization in the study area. Simulation results show that the RTM, which can include geological heterogeneity, provides more reliable constraints for the volume of hydrothermal fluids required for the observed scale of dolomitization, compared to batch geochemical modeling. According to the simulation results, the interstitial fluids released from underlying Cambrian formations due to high pressure and high temperature are not the only source of hydrothermal dolomitization. Other mechanism like long-lasting recharge of seawater-derived basement brines may be took place. Scenario simulations also indicate that the patterns of hydrothermal dolomitization and porosity distributions are strongly related to permeability heterogeneity in the formation. This study shows the ability and feasibility of geochemical modeling as a tool to quantitatively evaluate and further understand hydrothermal dolomitization processes, which are also helpful for other similar geological process research. However, reliable and detailed geological information, geological constrains, especially geological heterogeneity distributions, are crucial to achieve beneficial conclusions.