1 Department of Environmental Engineering, Technical University of Denmark2 Water Resources Engineering, Department of Environmental Engineering, Technical University of Denmark
Climate change induced by anthropogenic CO2 emissions is widely accepted to be the greatest immediate threat faced by modern civilization. Carbon capture and geological storage (CCGS) is one of the most promising geoengineering technologies currently within reach by which to, at least partially, mitigate this threat. The capture, compression and injection of CO2 in supercritical state into deep saline aquifers is a technique which attracts criticism not least for its additional cost to energy production but more so for delaying transition to renewable energies and risks posed to the environment. Although considered highly unlikely following appropriate site selection, leakage of CO2 from CCGS forms a major concern for both scientists and the public. Leakage would potentially occur through faults or abandoned boreholes and ultimately result in upward migration and discharge to the atmosphere. During migration CO2 would dissolve into groundwater forming carbonic acid, induce water-rock reactions and thus change groundwater chemistry. Therefore prior to implementation of this potentially necessary technology, environmental risks associated with leakage must be understood. Over the past 10 years scientists have worked in earnest to understand the potential effects of leakage in order that an informed decision on CCGS implementation can be made. This research can be broadly described as aiming to answer two key questions; how deleterious is leakage of CCGS to groundwater resources? and can it be detected geochemically? Some common hydrochemical development is apparent from the literature however many aspects of hydrogeological and hydrogeochemical impact of leakage into shallow aquifers used in water supply remain unclear. In this Ph.D. study an integrated approach was employed in order to answer the two key questions regarding leakage of CO2 into shallow aquifers. Consequently a combination of laboratory and field investigations were conducted supported by numerical geochemical modeling in order to identify, constrain and quantify processes controlling groundwater chemistry evolution. The output is 4 journal articles (1 in press and 3 submitted) and 3 technical reports. In paper I and technical report I simple batch reactors were employed coupled to comprehensive sediment characterization to determine the likely effects of CO2 on water chemistry in a range of shallow aquifers. Results showed aquifers can be broadly divided into three types; carbonate dominated, silicate dominated and mixed. Each aquifer type showed distinct water chemistry evolution thus inherent risks vary. These studies also highlighted the complexity of risk assessment and detection caused by the range of formation types potentially overlying storage reservoirs. Investigations described in Papers II, III and technical report II increase applicability to real leakage by observing in situ effects including groundwater flow. A silicate dominated shallow aquifer in Vrøgum, western Denmark forms the focus of study upon which a series of investigations were conducted. The main field study involved injection of 1600 kg of gas phase CO2 into the shallow Vrøgum aquifer over 72 days with more than 770 water samples taken. In addition pre- and post-injection sediment cores were studied in order to assess sediment alteration and aid identification of controlling geochemical processes. Results show a significant lowering of pH and increase in electrical conductivity, but except for Al concentrations reaching up to 75 μmol/L, the detrimental effects on groundwater quality were limited. Groundwater chemistry evolves spatially and temporally during leakage, thus risks posed and the best methods for detection will also vary. In addition, 2 main phases of leakage were identified; a pulse phase of elevated ions moving with advective flow succeeded by increasing persistent acidification caused by buffering exhaustion (i.e. depletion of reactive minerals). Aqueous element concentrations were delineated into 4 generalized behaviors; 1. advective pulse (Ca, Mg, Na, Si, Ba and Sr,), 2. pH sensitive abundance dependent (Al and Zn), 3. complete removal (Mn and Fe) and 4. unaffected (K). Concentration behaviors were characterized by; 1. a maximal front moving with advective flow, 2. continual increase in close proximity to the injection horizon, 3. removal from solution to zero concentration and 4. no significant change. Paper IV describes geochemical modeling conducted in support of field and laboratory activities and proposes that gibbsite derived Al3+ driven cation exchange can explain the majority of the water chemistry evolution at Vrøgum. In addition buffering exhaustion/sediment depletion is corroborated explaining increasing acidification observed. Results infer risks associated with and how best to detect leakage will change with time and also increase with depth. Consequently water quality may become more deleterious as a leak matures and will vary with depth.