Paper No. 2
Presentation Time: 9:00 AM-6:30 PM

DETERMINING THE ORIGIN OF CO2 IN SHALLOW GROUNDWATER THROUGH MODELING


BERGER, Peter M., Illinois State Geological Survey, University of Illinois Urbana-Champaign, 615 E Peabody, Champaign, IL 61820, pmberger@illinois.edu

Shallow groundwater wells are becoming a common, and sometimes mandatory, method for monitoring potential brine and/or CO2 leakage at carbon sequestration sites. However, the chemical composition of groundwater can change in response to many different factors besides the introduction of CO2 into an aquifer. Geochemical modeling can be a valuable tool for differentiating different types of perturbations and determining whether brine or CO2 have actually migrated into shallow groundwater.

During routine monitoring at an Enhanced Oil Recovery demonstration pilot test in Indiana, a spike in the concentration of inorganic carbon was observed in a monitoring well at the project site. Changes in sulfur and iron concentrations were also observed. However, sodium and chloride concentrations remained fairly constant confirming that there was no evidence of brine migration from the oil reservoir. Additionally, there was no observed pressure decrease in the reservoir to indicate leakage had occurred. As the geochemical changes were not the result of leakage, an alternative theory to explain the observed changes was the perturbation and introduction of oxygen into the aquifer during the drilling, coring, and development of the monitoring well. The reaction of oxygen with pyrite in aquifer materials and dissolved iron followed by microbial activity could explain much of the observed chemical changes.

We used Differential Evolution software coupled with REACT to match the inorganic carbon changes using the introduction of oxygen, and then the same method to match the data with the introduction of CO2. Each gas interacted with the water, mineralogy, and in the case of oxygen, microbes in the model while optimization software adjusted the amount of gas added, mixing rate, and kinetic reaction rates.

The modeled introduction of oxygen yielded a better match to the observations than the introduction of CO2. The CO2 based model contained no mechanism to increase the amount sulfur in solution. Considering the model in conjunction with the stable sodium and chlorine values and the lack of significant pressure variation, the most likely mechanism for the inorganic carbon spike was the introduction of oxygen into the aquifer during the installation and development of the monitoring well.