International Journal of Greenhouse Gas Control 39 (2015) 302–317 Contents lists available at ScienceDirect International Journal of Greenhouse Gas Control journal homepage: www.elsevier.com/locate/ijggc The DEMO-CO 2 project: A vadose zone CO 2 and tracer leakage field experiment Jean Rillard a , Corinne Loisy a , Olivier Le Roux a , Adrian Cerepi a,∗ , Bruno Garcia b , Sonia Noirez b , Virgile Rouchon b , Philippe Delaplace b , Olivier Willequet c , Claude Bertrand d a EA 4592 G and E, ENSEGID, Bordeaux INP, 1, allée F. Daguin, 33 607 Pessac Cedex, France b IFP Energies Nouvelles 1 and 4, Avenue de Bois Préau, 92 500 Rueil-Malmaison, France c Dimelco S.A. 120 rue du Fort, 59175 Vendeville, France d ALGADE-1 Avenue du Brugeaud, BP 46 – 87250 Bessines sur Gartempe, France article info Article history: Received 4 August 2014 Received in revised form 19 April 2015 Accepted 22 April 2015 Available online 2 June 2015 Keywords: CO2 geological storage CO2 leakage Carbonates vadose zone Geochemical monitoring Noble gas tracers Carbon stable isotopes abstract The DEMO-CO 2 project is dedicated to develop an analytical method for CO 2 leakage detection within the vadose zone and to understand the behavior of CO 2 in the near subsurface carbonate environment during an induced CO 2 leakage. A gas mixture of 3 m 3 of CO 2 , He and Kr was released at 3.7 m depth in a borehole located in a carbonate vadose zone. Different detection techniques were used to follow the CO 2 , and noble gas migration within the vadose zone. A numerical simulation done with COORES TM code was used to predict gas flux at surface. Measurements performed in the field showed heterogeneous gas transfer in the subsurface as a result of contrast of permeability/porosity of the porous media at meter scale. Both advective and diffusive gas transfer could be observed due to the different time frames needed to reach maximum He, Kr and CO 2 concentrations, confirming that noble gases can be used as precursor of injected CO 2 leakage in case of diffusive transfer prevails. Despite the subsurface heterogeneity, sim- ulations results of CO 2 flux at the surface were in relatively good agreement with field measurements. The comparison of CO 2 concentration with measurements of ubiquitous gas such as N 2 , coupled with isotopic 13 C analyses could be used to clearly differentiate injected CO 2 from natural background CO 2 produced by biological activity. Lessons learned through this experiment could help to improve future near-surface gas geochemistry surveys for site assessment, identify CO 2 leakage, and leakage monitoring at active carbon capture and storage sites. © 2015 Elsevier Ltd. All rights reserved. 1. Introduction The geological storage of CO 2 is currently considered as a potential solution to limit the accumulation of greenhouse gas in the atmosphere. However, several technical and environmental concerns are now being considered as a part of Carbon Cap- ture and Storage project. A significant site characterization is needed as a part of CO 2 storage site selection. In particular the consequence of CO 2 leakage from the storage reservoir has to be evaluated (European Commission (2009); U.S. Environmental Protection Agency, 2010a, 2010b). Therefore, geochemical and geo- physical tools have to be developed to monitor and detect potential ∗ Corresponding author at: ENSEGID-IPB 1, allée Daguin, 33607 Pessac, France. Tel.: +33 557121011. E-mail address: adrian.cerepi@ipb.fr (A. Cerepi). CO 2 leakage in sub-surface environment and into the atmosphere. In particular, the gas migration within the vadose zone (i.e., unsatu- rated zone) appears to be very complex to understand and remains very case-specific. Soil gas concentration and gas flux through the soil to the atmosphere are dependent on numerous processes (Schlömer et al., 2014). Indeed, the physical gas transfer dynamics within vadose zone results from the combined action of diffusion and advective transport processes (Garcia-Anton et al., 2014). Dif- fusion is driven by spatial and temporal concentration gradients; advection is driven by spatial and temporal pressure gradients. In addition to these physical processes, chemical reactions that may occur between gas and aqueous phases may affect the gas mobil- ity and concentration, especially for highly reactive species such as CO 2 . Furthermore, these processes are highly influenced by envi- ronmental parameters (temperature, atmospheric pressure, rain infiltration, biological activity, etc.) (Loisy et al., 2013; Gal et al., http://dx.doi.org/10.1016/j.ijggc.2015.04.012 1750-5836/© 2015 Elsevier Ltd. All rights reserved.