Reliability of Lumped Parameter Modelling of Pressure Changes in Geothermal Reservoirs

Abstract

Lumped parameter models have been used extensively to simulate data on pressure changes in geothermal systems in Iceland as well as in the P.R. of China, Central America, Eastern Europe, The Philippines, Turkey and other countries. Lumped models can simulate such data very accurately, if the data-quality is sufficient. The properties of the lumped models provide information on the corresponding properties of the geothermal system in question. Yet the principal purpose of such modeling is, of course, to estimate the production potential of geothermal systems through pressure response predictions and to estimate the effects of different management options. Because of how simple the lumped parameter models are, their reliability is sometimes doubted. Experience has shown that they are quite reliable, however, as examples presented demonstrate. The examples involve comparing pressure responses of geothermal systems, calculated by lumped parameter models, developed some years ago on basis of production histories available at the time, with the pressure responses observed since then. In addition, examples are presented of predictions for the same geothermal systems, calculated by open and closed lumped models, which demonstrate the precision, or sensitivity, of lumped modeling. Future pressure changes are expected to lie somewhere between the predictions of open and closed models and it is argued that the differences between these predictions do not demonstrate the unreliability of lumped parameter modeling, but simply the inherent uncertainty in all such predictions.