Single-well and inter-well dual-tracer tests for CCS pilot site characterization at Heletz (Israel): scope and limitations

A semi-analytical approximation to tracer signals is proposed for evaluating the chances of unambiguous tracer test interpretation for the CCS pilot site at Heletz (www.co2mustang.eu/Heletz.aspx), given the multiple constraints imposed upon tracer test execution at any deep geological formation regarding maximum pressure buildup admissible, fluid turnover rates, injection-fluid availability and conditioning capabilities, fluid disposal capacity, maximum test duration, tracer quantities/species available, tracer metering costs, etc. At the Ketzin site in Germany (www.co2sink.org), such constraints acted prohibitively towards conducting any tracer tests prior to CO2 injection (cf. EGU2009-11625, EGU2011-2823-1). In contrast, at the Heletz site within the MUSTANG project, a particular sequence of single-well (SW) and inter-well (IW) tracer tests to be conducted prior to CO2 injection (Bensabat et al. 2011, Niemi et al. 2011) is being regarded as a critical, indispensable contribution towards the goal of georeservoir characterization. An advantage of the proposed approach, expressing tracer signals U in terms of fluid volume recovery fraction f (rather than time), is its independence upon IW distance and upon fluid injection/production rates (which may be difficult to prescribe in advance). It roughly predicts conservative- and sorptive-tracer signals, fluid mixing and temperature evolution in SW and IW tests, for stratified georeservoirs whose high- and low-permeability layers have thicknesses of same magnitude order. Thus, besides the CCS context, it can be extended to a range of geothermal applications, including some aquifer-dominated systems in the Upper Rhine Rift Valley, S-German Malm/Molasse Basin, N-German Sedimentary Basin, but excluding fracture-dominated (petrothermal) systems. IW tracer signals are sensitive to storage capacity (measured by transport-effective porosities n), but suffer from ambiguity between longitudinal heterogeneity (Peclet number Pe) and vertical heterogeneity (hydrostratigraphy Prf(z)), and from all-parameter ambiguity at early times (f