Florian Konrad, Alexandros Savvatis, Kai Zosseder
Technische Universität München, Deutschland
Fault zones represent a major component of geothermal exploration concepts for the Upper Jurassic aquifer. The majority of these areas of complex deformation accrued through multiphase normal faulting processes during the Alpine orogeny (Moeck et al. 2015). Due to the possibility of their increased hydraulic properties compared to the surrounding host rock they often represent targets for a geothermal well’s path design (Böhm et al. 2013). But as the genesis of these fault zones varies due to a multitude of different factors (e.g. local stress field history, orientation, mechanical host rock properties) their effect on the hydraulic situation can also change from “highly permeable conduit” over “transparent to fluid flow” to “barrier behavior”(Cacace et al. 2013). Furthermore, if the hydraulic properties (specific storage and permeability) of the host rock are relatively high the effect of a developed highly permeable fault zone can be invisible in the pressure signal of the geothermal well (Konrad et al. 2019). Based on these factors there is an ongoing discussion of the importance of fault zones compared to the aquifer host rock for the productivity of a well.
To investigate this question, a realistic calibrated numerical model of a fault zone in the Upper Jurassic aquifer is used for the exploration of the naturally possible hydraulic parameter space. Here, the fault zone is treated as an integrative area of a certain thickness (representing fault core and damage zone together) where its properties are applied as average value and which is surrounded by a matrix portraying the aquifer host rock. By expanding this Finite-Element (FE) model through the Reduced Basis (RB) method and lowering numerical calculation times by a factor of 1000 while maintaining high accuracy, a comprehensive study of the governing large parameter space in the reservoir becomes possible (Degen et al. 2019). Through pressure data comparison of simulated pumping tests for each parameter combination first with and subsequently without a fault zone, the hydraulic effect of fault zones can be isolated. The fault zone influence on the geothermal well is then described and quantified in terms of the productivity index (PI) change. This methodology is on one hand used to explore the entire possible parameter space of the Upper Jurassic aquifer and additionally in a spatial analysis of the greater area of Munich where currently (June 2020) available hydraulic information from geothermal wells is gathered, regionalized and applied as model input. Through using spatial information in the form of high-resolution grids as model input the model results do not need additional interpolation and can directly be visualized as maps. The influence especially of fault zone properties (permeability, specific storage and thickness) on well flow regimes and PI enhancement is therefore spatially illustrated.
The comparison of these calculated flow regimes with in actual geothermal wells locally observed flow regimes suggests that fault zone thicknesses of 100 meter or higher are unlikely for the Upper Jurassic aquifer. Studying the fault zone induced PI change in dependency of the natural parameter space shows that fault zones (with increased hydraulic properties compared to the host rock) which are invisible in pumping tests can still contribute significantly to well productivity. Furthermore, in the southern region of the analyzed area where hydraulic matrix properties are decreasing, fault zones become the dominating source of well productivity up to a point where economic production rates can only be reached by developing a highly permeable fault zone.
