Title: The role of well testing in geothermal resource assessment

Abstract

In this thesis, different approaches and methods for analysing well test data to estimate the capacity of
a geothermal field are evaluated. The thesis lays emphasis on well completion and flow tests data, and
uses the Reykjanes geothermal field in southwest Iceland as an example for the application of the
selected methods. Analyses of data from step-rate injection tests, temperature and pressure profiles as
well as discharge tests conducted in the Reykjanes geothermal field are presented. Six wells, RN-12,
RN-13b, RN-17b, RN-18, RN-23 and RN-29 were selected for the study. The computer numerical
software, Well Tester, is used for the step-rate injection test analysis. Temperature and pressure
profiles are analysed to estimate the formation temperature and the initial reservoir pressure. Flow
characteristics are evaluated from the lip pressure method and the steam-water separator method,
respectively. The wellbore simulator, HOLA, is used to evaluate the generating capacity and
productivity indices of individual wells. The injectivity and productivity indices are compared with
results from other similar high temperature geothermal fields. The volumetric method based on well
test data is applied to predict the electrical generation capacity of the Reykjanes geothermal system.
The estimated values of permeability-thickness from the injection tests range from 1 to 30 Dm,
whereas the storativity ranges from 5·10-8 to 2·10-7 m/Pa.
The reservoir temperature estimate is between 280 and 340°C. The average capacity for producing
electricity is estimated around 10 MWe per well. The resource assessment is performed for two cases.
Case I considers the surface alteration area of the Reykjanes field of 2 km2 while case II considers the
low resistivity sheet at 1000 m depth b.s.l. of an area of 11 km2. The Monte Carlo simulation for case I
predicts with 90% confidence interval a generating capacity between 34 and 102 MWe for a
recoverable heat, with a most likely value of 65 MWe and between 20 and 61 MWe with a most likely
value of 39 MWe for 30 and 50 years, respectively. For case II, the simulation predicts with 90%
confidence interval a generating capacity between 38 and 290 MWe with a most likely value of 132
MWe and between 23 and 174 MWe with a most likely value of 79 MWe for 30 and 50 years,
respectively.