| Title | USE OF GRAVITY, MICROEARTHQUAKE, AND TRACER INJECTION TEST DATA IN CALIBRATING THE TIWI NUMERICAL RESERVOIR MODEL |
|---|---|
| Authors | F.R. Cinco, H.L. Marcuap, and A.J. Menzies |
| Year | 2019 |
| Conference | New Zealand Geothermal Workshop |
| Keywords | Tiwi, reservoir model, numerical modeling, geothermal model, precision gravity, microearthquake, tracer injection, conceptual model |
| Abstract | The Tiwi Geothermal Field, operated by Philippine Geothermal Production Company, Inc. (PGPC) in Albay, Philippines, has been in production since 1979 and the numerical model of the reservoir has been adequately matched to the collected production and injection history, including changes in subsurface pressure and temperature data as well as discharge enthalpies. The numerical reservoir model is based on the TOUGH2 simulation code and uses a double porosity formulation, with various levels of matrix block sub-division (MINC’ing) in the reservoir section. In addition to matching the usual production and downhole survey data, matching of the changes in gravity, which have been precisely measured across the field for the past 40 years, is also routinely conducted. It has proven to be useful in providing additional insights into the reservoir processes taking place, particularly related to phase change, movement of the steam-water interface in the reservoir and the degree of mass influx, either from external aquifers or internally from matrix to the producing fractures. The matrix to fracture interactions are controlled in the model by parameters such as the fracture spacing, degree of MINC’ing of the matrix blocks and the permeabilities of the matrix and fractures. Another geoscientific data set that is used to constrain the model is microearthquakes (MEQs). These are very low intensity earthquakes that can only be detected by precision instruments. Some of the MEQs are induced by injection, and to a lesser extent by production. We believe the distribution of these events indicates where rock-fluid heat exchange is occurring, and the existence and/or stimulation of the fracture network. Production and geochemical data during the past 20 years indicates a strong connection between edgefield injection wells in the southeast Naglagbong sector (where about 70% of the total brine is injected) and production wells in the South Kapipihan sector. Both the data and the model indicate that the injection in Naglagbong provides long-term pressure support to South Kapipihan with no negative thermal effects. The great depth and lateral extent of the MEQ activity readily explains this favorable behavior, as the result of a large volume of hot, fractured rock that allows for ample heat exchange before the injected brine reaches the production wells. The model has been adjusted to account for this volume, with consequences for reservoir volume and reserves. More recently, a binary project has been proposed which will lower injection temperatures. Although the match in the model to pressure and temperature changes in the area has been generally good, it was found by tagging the injection water that the velocity of fluid movement in the model was not consistent with available tracer test results. To improve the calibration of the model and make it more consistent with tracer arrival times and concentration trends, it was necessary to add a streak of high permeability blocks that corresponds to a mapped fault (Cale Fault) to act as a conduit of fluids between the injection and production wells used in the tracer test. The inclusion of both production and geoscientific data in calibrating the numerical model of Tiwi has resulted in improvements in the history matching and the simulation of reservoir processes, increasing confidence in the model’s output and the use of the model to provide input to resource development decisions. |