Predictive control of geothermal reservoirs based on early-stage, continuous hydromechanical characterization

Geothermal energy is a reliable source of energy that relies on the hot temperatures within the earth's core. The heat energy can be extracted through probes, closed loops or extraction and injection of water: Through a series of wells, hot water or steam are extracted from the ground while cold water is reinjected. While traditional geothermal energy requires sites where this heat is already close to the surface (e.g. near volcanoes, geysers or hot springs), enhanced/ deep geothermal energy taps into deeper rock formations. In hydrothermal geothermal energy, these rock formations already constitute reservoirs with sufficient permeability for the circulation of thermal waters. In petrothermal geothermal energy, the hot rocks are dry and do not allow for sufficient flow of water, such that an increase in their permeability is required. The main mechanism for increasing the permeability of petrothermal reservoirs is hydraulic stimulation, where water is injected into the rock under high pressure increasing the size of existing and creating new fractures.

A common side-effect of hydraulic stimulation is induced seismicity which has led to the interruption of geothermal projects in the past. Controlling induced seismicity during hydraulic stimulation is therefore of high importance. The joint research project Lerche aims to combine hydraulic and seismic methods for reservoir characterization with an adaptive predictive controller. During hydraulic stimulation, the hydromechanical state of a reservoir shall be derived from measurements and further changes shall be predicted. The predictive controller – based on surrogate system models or based on data – shall be able to take these changes into account in order to control current and future seismicity through adjustments of the hydraulic stimulation.