Abstract:
This thesis describes an investigation into the characteristics and performance of geothermal doublet systems in the Rotorua geothermal field, New Zealand. The term doublet applies to a system which includes two wells, the first well being used for production and the second one being used for injection of the water after extraction of its heat. The investigation includes experimental work and a theoretical study. The aim of this investigation is to obtain a better understanding of the doublet system behaviour in order to use the Rotorua geothermal resource more effectively. The experimental work was to carry out tracer tests in order to determine the likelihood of thermal breakthrough in some Rotorua doublets; seven tracer tests (3 tests in winter and 4 tests in summer) were carried out by the author and co-researchers during 1991 and 1992 in the Rotorua geothermal reservoir. Only two of these tracer tests, both in the same doublet system, produced breakthrough and the amounts of tracer recovered were very small. Therefore there does not appear to be a direct connection (or a strong connection) between the injection and the production wells in those tested doublet systems in Rotorua. The theoretical study is to obtain a suitable computer model for interpreting doublet tracer tests. Analytical methods for solving both one-dimensional and two-dimensional geothermal doublet problems were developed. They can be used to calculate the travel time of tracer and heat from the injection well to the production well, the tracer concentration and temperature changes versus time at the production well. In order to determine what kind of block structures are required for the more complex three-dimensional numerical doublet simulations, the simple analytic models are compared with a semi-analytical method (RESSQ) (Javandel et al., 1984) and a numerical method (MULKOM) (Pruess, 1988 and 1990). This comparison concludes that models using smaller blocks near the injection and the production wells of the doublet and larger-sized blocks away from the wells can be used to obtain acceptable results with MULKOM. The MULKOM results are approximately 10% different from the analytical or semi-analytical results. The tracer tests conducted in Rotorua, New Zealand, by the author were simulated. To examine the method of doublet simulation, the Klamath Falls tracer test carried out by other investigators (Sammel et al., 1984 and Gudmundsson et al., 1984) was selected and simulated as another example. Computer simulations were conducted to test various models and to estimate reservoir parameters by matching computational results to field tracer breakthrough curves. The simulations show that multi-layered models consisting of alternate layers of high permeability fractures and low permeability matrixes give the best-fit between the simulation results and field tracer test data. The same method was used to simulate the Klamath Falls tracer test, and satisfactory results were obtained. The Klamath Falls results give strong support to this simulation method. The simulation results demonstrate that the production wells are not suffering thermal drawdown due to regional flow effect in the aquifers for the tested doublets in the Rotorua geothermal reservoirs. The simulation results for the Rotorua winter and summer tracer tests show that there is a seasonal change in the regional flow pattern due to the geothermal utilization load. The final stage of this research work is an assessment of the doublet systems for the Rotorua geothermal field. This assessment is based on the reservoir characteristics, utilization experience, tracer tests and tracer test simulation. The environment impact, doublet economics and engineering techniques were also considered. It is concluded that the use of doublet systems is feasible in the Rotorua geothermal field.