Tracer testing at the Habanero EGS site, central Australia

Abstract

The Habanero Enhanced Geothermal System (EGS) in central Australia has been under development since 2002, with several deep (more than 4000m) wells drilled to date into the high-heat-producing granites of the Big Lake Suite. Multiple hydraulic stimulations have been performed to improve the existing fracture permeability in the granite. Stimulation of the newly-drilled Habanero-4 well (H-4) was completed in late 2012, and micro-seismic data indicated an increase in total stimulated reservoir area to approximately 4km2. Two well doublets have been tested, initially between Habanero-1 (H-1) and Habanero-3 (H-3), and more recently, between H-1 and H-4. Both doublets effectively operated as closed systems, and excluding short-term flow tests, all production fluids were re-injected into the reservoir at depth. Two inter-well tracer tests have been conducted: the first in 2008, and the most recent one in June 2013, which involved injecting 100kg of 2,6-naphthalene-disulfonate (NDS) into H-1 to evaluate the hydraulic characteristics of the newly-created H-1/H-4 doublet. After correcting for flow hiatuses and non-steady-state flow conditions, tracer breakthrough in H-4 was observed after 6 days (compared to ∼4 days for the previous H-1/H-3 doublet), with peak breakthrough occurring after 17 days. Extrapolation of the breakthrough curve to late time indicates that approximately 60% of the tracer mass would eventually be recovered (vs. approximately 80% for the 2008 H-1/H-3 tracer test). This suggests that a large proportion of the tracer may lie trapped in the opposite end of the reservoir from H-4 and/or may have been lost to the far field. The calculated inter-well swept pore volume is approximately 31,000m3, which is larger than that calculated for the H-1/H-3 doublet (∼20,000m3). A simple 2D TOUGH2 tracer model, with model geometry constructed based on the current conceptual understanding of the Habanero EGS system, demonstrates good agreement with the measured tracer returns in terms of timing of breakthrough in H-4, and observed tracer dispersion in the tail of the breakthrough curve.