| Abstract |
Geophysical interpretations using magnetotelluric, gravity, and micro-earthquake (MEQ) data are routinely employed for geothermal exploration and for delineating geothermal resources. During production and development phases, repeat precision gravity surveys and continuously recording MEQ arrays can be deployed to evaluate rock/fluid interactions, pressure/stress changes, fluid saturation changes, and production versus recharge balances as a function of time. If measurement arrays are permanently installed, then determining physical property changes with time is straightforward. However, in the case of MEQs, for example, the receivers are rarely permanently installed and are often moved during production, and the micro-earthquakes will be stimulated in different parts of the reservoir depending on injection well locations. We have therefore developed a 3D inversion algorithm that inverts two different epochs of data for a baseline model and the changes to that model over the intervening time period. This is done by parameterizing the inversion with the model at the first time period (m1) and the changes to that model (dm), such that the model at time period 2 is equal to m1+dm. Using synthetic data, we show how this algorithm works and we compare it to another method that solves for both models simultaneously (m1 and m2) while minimizing the difference between the two models. Finally, we apply the method to real data taken from an active geothermal resource, including 4D joint inversion of gravity changes and travel-time tomography over the same epochs. |