Selection, instrumentation and characterization of a pilot site for CO2 leakage experimentation in a superficial aquifer

CO2 geological storage is one of the options for mitigation of GHG emissions into the atmosphere. Although storage is provided for several millennia, leakage can occur. Therefore the risk that the fugitive CO2 reaches a shallow aquifer cannot be excluded [1]. That is why the consortium launched the research program CIPRES, funded by the French Research Agency, concerning the potential impacts of CO2 leakage on groundwater quality. This program includes the realization of a shallow CO2 release experiment in a carbonated aquifer and its monitoring in the saturated and unsaturated zones, the soil and the soil-atmosphere interface. The experimentation consists in drilling a well into the aquifer, injecting dissolved CO2 in water [2] and tracking its impact downstream. Prior to this experiment, a site had to be selected, instrumented and characterized. Hydrogeological target is the Paris basin chalk which is the largest French aquifer. The selected area is a large parcel, formerly cultivated in intensive monoculture but left fallow for several years, located in Catenoy (Oise) about 50 km north of Paris. The geological log shows the chalk is covered by 6-7 m of sands but water table is located at 12-13 m depth, i.e. integrally in the chalk strata. The site has been equipped with 10 wells to characterize the groundwater flow (hydraulic gradient similar to 5 10(-3)). 6 wells are aligned in the groundwater flow direction: the PZ2, planned for CO2 injection, the PZ1 located 20 m upstream and 4 monitoring wells located downstream between 10 and 60 m from PZ2. There are also 4 lateral piezometers located near (PZ7, PZ8) and far (PZ9, PZ10) from the injection site: they are dedicated to the lateral plume. These 25 m depth wells are equipped with tubing fully slotted in the chalk. Alongside, 4 wells dedicated to gas monitoring in unsaturated zone have been drilled at 11m depth. To characterize the site, the following operations were performed prior to the CO2 injection between March and September 2013: pump test into PZ2 to estimate the hydraulic conductivity (10(-3) m. s(-1)) and the storage coefficient (1.6%); the hydraulic conductivity is higher in the first 3 m of saturated zone (similar to 5 10(-3) m. s(-1)); hydrogeochemical baseline with monthly water analysis in each well (physicochemical parameters, major and minor ions, metallic trace elements); electrical conductivity is 714 mu s.cm(-1) and groundwater has a calcium-bicarbonated facies with high nitrate concentration (45 mg. l(-1)) and low-level presence of trace metals; gas baseline with continuous O-2 and CO2 measurement in vadose trough 11 m depth boreholes and measurements campaigns to determine gas concentration in the soil and gas flux on the surface; flowmeter heatpulse logging at PZ2, PZ3 and PZ4 in static and dynamic conditions; only the area situated from 15 m to 20 m depth is productive and a vertical natural downflow is measured in PZ3 (similar to 1-2 mm. s(-1)) and PZ4 (1 cm. s(-1)); furthermore, no flow is measured from 20 to 25 m depth; this confirms the vertical anisotropy of the chalky aquifer; Following this experiments, a tracer test has been done in June 2013 from PZ2 with fluorescent tracer (amino-acid G) and a dissolved gas (He), in the way to calibrate the future CO2 injection and its monitoring. 2 m(3) of water from the aquifer were pumped the day before the injection and stored in tanks. Then 2 kg of tracer were dissolved and water was saturated with He. This water was injected into the PZ2 the next day during 8 h. The monitoring was conducted for a month by water sampling, tracer and He analyzes. The peak of fluorescent tracer arrived at the 2nd day following injection in the PZ4 (20 m downstream) but only one week later in the PZ3 (10 m downstream), due to the site anisotropy. Low He concentrations have been detected in the unsaturated zone of PZ4 before PZ3, in correlation with tracer migration. The main result is the existence of an high aquifer anisotropy in such a small area: i) the first 3 m of the saturated zone shows a higher permeability and the last 5 m a lower one, ii) the tracer arrives quickly and with higher concentration at PZ4 located 10 m farther than PZ3. However, the dilution ratio is rather important and may induce, during CO2 leakage experimentation, a slight pH decrease at the downstream wells: in order to increase the impact of the CO2 leakage, we have therefore planned to inject a higher volume (10 m(3)) of CO2 saturated water. (C) 2014 The Authors. Published by Elsevier Ltd.