Improving Exploration for Geothermal Resources with the Magnetotelluric Method

• Main Author: Lee, Benjamin M
• Other Authors: Unsworth, Martyn (Physics)
• Format: Thesis
• Language: English
• Published: University of Alberta. Department of Physics. 2020
• Subjects: geo; socio; Krafla; Iceland
• Online Access: https://era.library.ualberta.ca/items/7dff1f0b-53f1-4ae2-bb11-9cf6f35a06cf

Specialization: Geophysics Degree: Doctor of Philosophy Abstract: This thesis investigates improvements in methods used for exploration for geothermal resources with the magnetotelluric (MT) method. Geothermal energy is a renewable resource that provides heat and electricity with low carbon emissions. Targets in conventional geothermal exploration are geothermal reservoirs, which are underground regions with hot fluids which can be extracted for direct use of heat or electricity production. Geothermal reservoirs are detected by geophysical methods such as MT, a technique utilizing naturally occurring electromagnetic signals to image the subsurface electrical resistivity. MT is useful for locating geothermal reservoirs because high-temperature hydrothermal alteration can make the reservoir relatively resistive. Improvements in the strategy used for MT exploration are needed to make this method more effective. In this thesis two case studies are described where conventional analysis and interpretation of MT data was inadequate for assessing the potential geothermal resource. The first study considered the Krafla geothermal field in Iceland. Supercritical fluids beneath the geothermal field could increase electric power output by an order of magnitude per well if used instead of steam. The IDDP-1 well was drilled in the year 2009 to reach the supercritical fluids at a depth of 4 to 5 km. However, drilling ended prematurely when magma was unexpectedly encountered at 2.1 km depth. This study investigates why the magma was not imaged with the existing MT data. First, improvements to the 3-D inversion of the Krafla MT data were implemented, including: (1) a 1-D resistivity model as a constraint on the final resistivity model; (2) full impedance tensor data instead of only the off-diagonal elements used by previous authors; and (3) model cells with horizontal dimensions of 100 by 100 m, which is a finer discretization than used by previous authors. The most prominent feature in the 3 D resistivity model is the low ...