Geochemical modeling-based prediction of water-rock interaction during aquifer storage and recovery utilizing selected Colorado Front Range aquifers
Date
2023
Journal Title
Journal ISSN
Volume Title
Abstract
This study characterizes the Fountain Formation, Ingleside Formation, and sandstones of the Dakota Group and considers the potential of these three formations as hypothetical Aquifer Storage and Recovery (ASR) targets. Compositional data from surface rock samples, including major, minor and trace elements from bulk rock geochemical analysis and mineral identification from petrography are used to infer a generalized mineral suite to represent each of the formations of interest. Similarly, compositional analyses from domestic water well samples, including major anions and cations and selected metals, were used as generalized representations of native water from each formation of interest. Finally, compositional data from treated city water was obtained and used as a generalized representation of injection water. The generalized rock data along with the generalized native water data represent a hypothetical injection environment while the treated water composition represents a hypothetical injection water. All water and rock data were used to populate a Single Pass Mixing equilibria Model that simulated an ASR system using the USGS geochemical modeling computer program PHREEQC (PH REdox EQuilibrium). Model results include mixed solution compositions, mineral saturation indices and estimates of mineral mass precipitation during simulated injection. Results of modeling suggest there is limited geochemical water-rock interaction during ASR in the hypothetical environment in this study. Model results indicate that the mixed solution composition is controlled more by the injected solution than by reactions occurring between the injection fluid and aquifer host material. Specifically, as greater volumes of hypothetical injection water are introduced with each model step, the compositions of the resulting mixed solutions increasingly resemble those of the injected water. The model predicted the precipitation of hematite, kaolinite and quartz during injection of the hypothetical injection water. Because aluminum was below detection in the water analyses and an arbitrary value less than the detection limit was used in the model, the prediction of kaolinite precipitation is not meaningful. Further, the model was constrained to not permit mineral dissolution, limiting the applicability of the model only to the consideration of mineral precipitation. In addition, benchtop leaching experiments were performed on rock samples to provide additional information about potential water-rock interaction. Benchtop experiment results are presented, but the focus of the study is primarily on geochemical modeling results. Water analysis results presented here suggest that the formations of interest currently contain good quality water. Modeling results suggest that injection of treated water would likely not lead to volumetrically important precipitation of minerals in the formations.