This paper describes the results of an analysis of data obtained from a series of heat-sweep experiments performed in the Stanford Geothermal Reservoir Model using the Lawrence Berkeley Laboratory reservoir simulator. The physical reservoir model is an experimental system consisting of a pressure vessel which contains a granite rock matrix with production and recharge capabilities to simulate the heat-sweep process in a fractured hydrothermal reservoir under liquid-phase conditions. Arrangements were made with the Lawrence Berkeley Laboratory to test their geothermal reservoir simulator on the physical model data. The objectives were to provide insight into the detailed physical processes occurring in the relatively complex physical system and to provide feedback to LBL on the capability and possible improvements to the LBL reservoir simulator to model a complex physical system. The overall conclusion of this work is that the LBL simulator does an excellent job of predicting the physical processes in the Stanford Geothermal Reservoir Model experiments for extreme thermal gradient conditions and for a system with very complex boundary conditions. The analysis demonstrates the importance of specifying relevant parameters accurately to provide adequate modeling for the important physical processes.
Read full abstract