As a part of a full scale production test, a long term tracer test was performed in the Thelamork low temperature geothermal system, in N-Iceland. The tracer test was aimed at recovering the transport properties of fractures connecting the injection and production wells. Hence, the estimated parameters might be used in determining the performance of the system under various injection schemes. A qualitative evaluation the tracer return profile showed the presence of strong recirculation effects. In addition, the return profile indicated that the medium appears to be highly dispersive. Earlier modelling studies employed a one-dimensional two path model to match the return profile and substituted the properties of the major path in the Lauwerier model to estimate the thermal breakthrough time. However, the two path model estimates a very large dispersive tiansport almost equal to the convective transport. This large dispersivity necessitates adding a dispersive heat transport term in the Lauwerier model and as a result reduces the Lauwerier thermal breakthrough time almost to half. Considering the injection and production rates, we used a more accurate one-dimensional five-path model in this work. This model infers a smaller dispersivity and leads to a greater breakthrough time than the two path model, …

Until recently, geothermal reservoir simulators use flat interface thermodynamics to determine the thermodynamic state of the reservoir. Development of new simulators and the modification of existing ones has now incorporated the physics of curved interface thermodynamics. These simulators account for the effects of sorption and capillary forces. The simulators GSS and TETRAD were used to simulate the performance of a hypothetical vapordominated geothermal reservoir. GSS is a simulator specifically developed to account for adsorption by using adsorption isotherms. On the other hand, TETRAD is a commercial simulator that was modified to account for vapor pressure lowering by using capillary pressure relations. GSS and TETRAD yielded similar results. Thus, the two formulations being used to account for curved interface thermodynamics are practically equivalent. Areas for improvement of both GSS and TETRAD were identified. The hysteresis and temperature dependence of sorption and capillary properties are issues that are needed to be addressed.

A long term interference test was conducted under conditions of multiwell variable flow rate at Takigami for about ten months in 1987. The test data have been analyzed with an on-line analysis method on the basis of the linesource solution. This method employs Kalman filtering to process the data and then provides the best estimates of reservoir transmissivity and storativity when a new pressure data at an observation well becomes available. The pressure changes measured at seven observation wells have been analyzed with the present method using an infinite reservoir model. The data from one observation well have been further analyzed assuming a presence of a linear boundary. Performances of the parameters estimated for different reservoir models are compared. Fairly good estimates of reservoir parameters are obtained on the basis of an infinite reservoir model for two wells using the entire pressure data whereas for other five wells using a part of the pressure data.