Characterization of a Geothermal Reservoir Using a Transdimensional Inversion Method

Abstract

The Waiwera hydrothermal system is a geothermal reservoir in a fractured sandstone aquifer. Existing models simulate the aquifer using continuum approximations, neglecting the effect of individual geological features. In this paper we present a methodology using discrete fractures to model the thermal anomaly at the site. There are a number of challenges associated with modeling discrete features. For instance, the exact number of stratified layers of the sandstone aquifer and intersecting faults is unknown. This limits the applicability of any discrete fracture network model, as it requires a predefined number of model parameters for inversion. To resolve this issue, we apply a transdimensional fracture network inversion method. The approach relies on a reversible-jump Markov chain Monte Carlo algorithm that allows for variable number of model parameters, thereby simultaneously calibrating the number of fractures in the model. This technique has already successfully been applied on smaller scales, to interpret cross-borehole tracer tomography experiments. We demonstrate that the methodology is also applicable for reservoir scale aquifer characterization, with some modifications. We show how the discrete approach could reconstruct temperature observations and we demonstrate how this approach could reveal the discrete features (faults and layers) of the aquifer. The presented findings will be used for the preparation of new investigations, and to enhance existing 3-D continuum models.