Modeling Arsenic Behavior During Artificial Aquifer Recharge
Loading...
Authors
Bardet, Benjamin M
Issue Date
2022
Type
Thesis
Language
Keywords
Arsenic , Arsenic Modeling , Geochemical Modeling , Groundwater Recharge , Modeling
Alternative Title
Abstract
Increased variability in precipitation coupled with increases in population across the large portions of the western United States has resulted in greater demand being placed on this region's water resources. As a result, water conservation and storage methods are being explored to mitigate the impact that potential drought may have on these expanding population centers. Aquifer storage and recovery projects (ASR) in which excess surface water is captured and injected in local aquifers for storage, are becoming more widespread and important. As part of these programs, the use of highly treated wastewater or reclaimed water as a potential source for groundwater injection is becoming more feasible. This wastewater stream is typically underutilized if utilized at all. A hurdle in “traditional” ASR projects and particularly in reclaimed water ASR projects is understanding how these injected water’s chemistries will interact with the native groundwater and geology of the aquifer into which they are being introduced. This study looks to identify how a numerical geochemical model framework can be used to identify potential compatibility issues between injected and native groundwaters in which degradation to overall water quality may occur. This framework is tested by looking at a case study in Reno, Nevada in which a small-scale aquifer injection test was conducted utilizing highly treated reclaimed water. In this case study, unanticipated changes in arsenic levels were observed in the aquifer during the injection test. A geochemical model was created utilizing the USGS PHREEQC modeling software to replicate batch experiments performed utilizing mineralogy collected from the Reno-Stead site. This model was used to confirm the mechanism of arsenic release at the site as well as demonstrate its ability to understand potential mitigation measures or investigate other potential contaminants.
The results were matched with the lab and field observations and iron-oxide sorption was identified as the source and driver of arsenic mobility at the site. The developed geochemical model was able to match the observed arsenic variability at the Reno-Stead site in response to changes to injection pH values. Injection pH values in the range of 6.7-7.2 were shown to decrease arsenic concentrations below background levels. At pH values in the range of 7.1-7.4 which correspond to the actual pH values of the injection water at the site, arsenic concentrations remained relatively stable or increased slightly in response to desorption from iron-oxide surfaces caused by shifts in pH. Models run without the inclusion of iron-oxide surfaces showed no arsenic concentration changes in response to changes in injected pH and did not match the field observations from the site. This validates the hypothesis that iron-oxide sorption and desorption was the primary driver of arsenic variability observed at the Reno-Stead site.
The ability of this relatively simple geochemical framework to accurately identify and replicate observed chemistry changes at this test site proves its viability as a potential method to identify and test future sites for any compatibility issues between injected and native groundwater before significant investments of time and resources occur. While further work is needed to create a more rounded and widely applicable model, this modeling pathway allows for better understanding of the geochemical mechanisms present at recharge sites and helps direct future investigations.