THM modelling of seismic velocities changes at DeepStor heat storage demonstrator

In Central Europe, the thermal energy required for heating and cooling represents a major CO2 emitter in the energy sector. Seasonal underground heat storage offers the option to store a large amount of excess heat in summer for usage in winter and, with that, decrease the need for conventional sources of energy. Today, high-temperature aquifer thermal energy storage (HT-ATES) systems are attracting large interest for securing a heat demand in a sustainable manner.

In HT-ATES systems, hot water is injected into a reservoir over the summer months while exchanged cold water is injected over the winter season. These changing temperatures and pressures will affect the geomechanical and thermo-hydraulic properties of the reservoir and the surrounding layers. Monitoring the changes in the reservoir properties is key to run a heat storage system safely and efficiently. We try to determine if active seismic imaging could be a suitable method to characterize the time-space evolution of the reservoir.

With view on designing future geophysical assessment and monitoring systems, we perform thermo-hydro-mechanical (THM) modelling, using MOOSE and TIGER applications, with characteristics based on the DeepStor demonstrator under development in the north of Karlsruhe (Germany), at KIT, to determine the changes in the poroelastic properties of the underground. The first three layers model includes different mechanical properties with one borehole. The simulation of hot water injection over a period of time allows to quantify its effect on the underground material properties. Besides the DeepStor demonstrator expected operational frame, we test additional injection schemes with varying underground properties to simulate the different ranges of porosity changes and look at their effects on the elastic properties.

The changes in the parameters from the THM model are linked to seismic sensitive variables, such as velocities and impedances, using empirical equations. Hence, we can quantify the effects of injection on such variables and determine if it would be possible to detect them with active seismic surveys.