Parametric study of reservoir properties and their effect on energy recovery

Geothermal energy constitutes an important energy resource worldwide. Effective management of such a resource requires an understanding of a complex set of physical phenomena, including interphase mass transfer, convective transport of mass, and conduction and convection of energy. The coupled nature of these processes frequently requires that numerical simulation be used to investigate reservoir response to different management strategies. The first stage of designing a simulation study of a particular reservoir involves defining the boundaries of the reservoir itself. We must determine the reservoir structure, the ''edges'' of the field, select the appropriate boundary conditions to be used, i.e., whether pressure support from an adjacent aquifer is present, values of heat flux, etc. Having determined the three-dimensional extent and shape of the reservoir, we must identify the relevant fluid and petrophysical properties to be used in the simulation. In fractured geothermal systems, this data includes absolute and relative permeabilities for the fractures and rock matrix, fracture spacing and orientation, capillary pressure-saturation relationships, thermal conductivities, and others. While it is fairly straightforward to identify the data required for accurate simulation of a geothermal reservoir, data acquisition is a different matter. Reservoir engineering is unique in the engineering disciplines in that much of the data is inferred by indirect means rather than being collected or measured. The data that can be measured (relative permeability, for example) can be of questionable reliability, in that the rock is removed from its native state, and conditions altered. Much of the data is also fit to empirical relationships; some is even estimated from these relationships. Due to the inherent uncertainties is this data, results of simulation studies using these data must be used with caution. Of course, each of the variables noted above do not impact simulation results in the same way. In fact, errors in some of the data may exert little or no effect on our results. If that is the case, little effort need be expended by the geothermal operator to obtain this data. The problem is knowing which data exerts the most influence on simulation results, and therefore which should receive highest priority in acquisition efforts. The study presented in this paper seeks to answer this question by quantifying the effect changes in various parameters have on reservoir response to injection in a vapor-dominated reservoir. From a base case reservoir dataset we will examine differences in injectate recovery and steam energy produced as input data is varied. Parameters that are examined include relative permeability and capillary pressure relationships, fracture spacing, initial liquid saturations, and geologic structure