Several models have been developed of the Svartsengi geothermal field and used to predict the future pressure drawdown due to exploitation. These are based upon different reservoir geometries and recharge patterns. All of them neglect the effect of a two-phase zone overlying the main reservoir. A numerical model is developed that includes the effect of the two-phase zone and the reservoir parameters determined and predicted pressure drawdown are compared to previous estimates. It is found that the assumed recharge patterns and two-phase zones have significant impact upon the lone-term pressure drawdown predictions.
INTRODUCTIONIn the evaluation of geothermal systems, modeling studies are necessary to develop the proper exploitation scheme for a resource. On the one hand, it is generally advisable to use the simplest model possible that considers the important physical processes occurring in the system [Bodvarsson et al., 1985]. More complex models introduce a greater degree of freedom, so that problems of solution uniqueness may arise. On the other hand, an oversimplified model will not consider all the important processes, nor their coupling. The choice of a mod. e•l depends upon the available data base, the questions to be addressed, and the field complexities that will affect answers to these questions.At present, we do not completely understand how detailed a model of a geothermal system must be in order to possess reasonable predictive capabilities. Rather simple material balance models have been used successfully in the oil and gas industry; less experience is available for geothermal systems. Various material balance (lumped parameter) models have been applied to the Wairakei geothermal field in New Zealand [e.g., Fradkin et al., 1981] starting with the model of Whiting and Ramey [1969]. It can be argued that these models can readily match the production history, but that they lack reliable predictive capabilities. It can also be argued from the Wairakei experience that simple lumped parameter models are comparable in performance to the more sophisticated numerical models [Grant, 1983]. It is therefore of interest to investigate the predictive capabilities of various models of different complexity for a given geothermal system.The field data from the Svartsengi geothermal field in Iceland are well suited for testing exploitation models and methods. The field has been under exploitation since 1976 and has a documented production and drawdown history [Gudmundsson and Olsen, 1987; Gudmundsson and Thorhallsson, 1986]. Fluids from the reservoir are used to heat freshwater which is piped to nearby villages for space heating (capacity 125 MWt); the geothermal fluids are also used for electric power production (capacity 8 MWe; Thorhallsson, 1979]. Material balance analysis of the field data has been carried out by Kjaran et al. [1979, 1980], Regalado [1981], and Gudmundsson and Olsen ['1987]. In the present work we develop a rather simple model of the reservoir that includes thermal and two-phase effects. THE SVAR...