| Title | The Effect of Bottom Boundary Conditions on Predictions of Steam Production from Geothermal Reservoir Models |
|---|---|
| Authors | John O'SULLIVAN, Mike O'SULLIVAN |
| Year | 2016 |
| Conference | Stanford Geothermal Workshop |
| Keywords | Geothermal reservoir modelling, boundary conditions, TOUGH2 |
| Abstract | A very large numerical model of a synthetic geothermal system has been created using the University of Auckland's improved air/water supercritical AUTOUGH2 code. The model extends from the surface to the brittle/ductile transition region at depth of 8 km and encompasses the system’s entire geothermal plume. At the bottom boundary a heat anomaly is applied which creates a hot, liquid-dominated system similar to those found in New Zealand. A production scenario equivalent to 120MW was then run using the model and the results treated as the true history for the system. Two smaller models were then created using the standard version of AUTOUGH2 with bottom boundaries above the supercritical region. Each model applied a different, common reservoir modelling approach at the bottom boundary. In both approaches a distributed constant mass flow is applied at the bottom boundary to represent the hot inflow from the deeper plume. However, for heat input the approaches differ. The first approach applied heat uniformly to the bottom boundary representing the background heat flux. The second approach applies a constant temperature (hot-plate) boundary condition using inactive blocks at the base of the model. Both models accurately reproduced the natural state temperatures of the hypothetical system. However, for the production scenario both approaches underpredicted the total steam production with the hot-plate approach having a slightly larger error. The underprediction could be reduced when using either approach by calibrating the model parameters away from their true values. |