Large scale fluid transport within the Earth’s crust is predominantly controlled by planar void space like fractures and crack networks. Characterizing the time dependent hydro-mechanical properties of such structures, therefore, is of paramount importance for an economic use of geo-reservoirs for energy (power and/or heat) provision. This particularly applies to Enhanced Geothermal Systems (EGS), where single or multiple fractures are created by engineering means, called reservoir stimulation. In the framework of the EU Horizon 2020 MEET project, researchers of the GFZ German Research Centre for Geosciences, in cooperation with colleagues from other partner institutions, investigate the sustainability of fractures to transport fluids in different geothermal reservoir rock types in the laboratory.
Using apparatuses that simulate reservoir conditions (Fig. 1a), both short and long-term experiments are performed to investigate effects of physical and chemical boundary conditions (e.g., stress, pressure, temperature, fluid chemistry) as well as time on changes in the hydraulic performance of (artificially) fractured rock samples (Figs. 1b and c). Emphasis is placed on revealing the micro-mechanisms that change the fracture opening with time (Fig. 1d) and rock types are investigated that, to date, have not or only rarely been targeted for EGS, e.g., metamorphic rocks like slates. The primary physical parameter investigated is the fracture transmissivity (or permeability) being a measure of the fracture’s ability to transmit flow. The experiments show that the ability of fractures to conduct fluids over the lifetime of an EGS is strongly influenced by, both, physical and chemical factors.
Regarding the physical contribution, this was found to be expressed by (1) a reduction of fracture transmissivity by several orders of magnitude when increasing the confining pressure (pc) or differential stress (?) acting perpendicular to the fracture surfaces, depending on host rock composition, (2) a minimum in fracture transmissivity at elevated confining pressures and differential stresses, suggesting a change of fluid flow pattern from flow over the entire fracture surface at low pc and ? towards localized flow through channels at higher pc and ? due to crushing of fracture surface asperities, and (3) a pronounced positive effect of fracture surface roughness on fracture transmissivity.
With regard to the influence of long-term chemical fluid-rock interactions on the hydraulic properties of fractures it shows that (1) fracture transmissivity decreases with fluid flow over time but any such reduction terminates when fluid flow is stopped, (2) elevated temperatures enhance the former process, and (3) the long-term fracture evolution is closely correlated with fluid composition, which controls the competition of pressure solution and fracture wall dissolution (Fig. 1d). The former leads to a retreat of the propping asperities that keep the fracture open and the latter causes the enlargement of the void space that allows fluid flow.
In summary, the experimental results demonstrate that fracture surface roughness, the mechanical properties of the rock matrix, the respective fluid and rock compositions, as well as temperature are important factors controlling long-term fracture sustainability. This study implies that the utilization of unconventional reservoir rocks for running an EGS successfully may be feasible, if the present key findings are being considered.

Figure 1. a. One of the three apparatuses at GFZ used to perform flow-through experiments at simulated reservoir conditions; b. disk-like samples prepared for investigating the influence of physical boundary conditions, e.g., surface roughness, rock composition, effective stress, etc. on fracture transmissivity; c. cylindrical samples used to investigate chemical effects by fluid-rock interactions on long-term fracture transmissivity; d. schematic illustrating potential mechanisms governing fracture deformation and transmissivity evolution.