SQUARE: Stress predictions - QUAntification and REduction of uncertainties with geomechanical-numerical subsurface models: GFZ

SQUARE: Stress predictions - QUAntification and REduction of uncertainties with geomechanical-numerical subsurface models

Duration : 2023-2026

Funding: BMWK

Principle Investigators: Moritz Ziegler, Oliver Heidbach

Project partners: Andreas Henk, Karsten Reiter (TU Darmstadt); Florian Wellmann, Denise Degen (RWTH Aachen)

For the planning of a deep geological repository for high-level radioactive waste, precise knowledge of the stress state in the subsurface is necessary. Due to the often insufficient data available, the stress state in the volume is simulated with 3D geomechanical-numerical models. In this way, the complete stress tensor can be predicted. However, generally no reliable statements about the prediction quality of the model are possible. In general, it must be assumed that the uncertainties of a model are large. Especially for applications with such high safety requirements as the final disposal of high-level radioactive waste, the uncertainties must not only be known, but also be within a clearly limited range.

The SQuaRe project is dedicated to the development of new methods for quantifying and reducing uncertainties in geomechanical-numerical models. For this purpose, uncertainties in the model output are considered on the one hand as the result of uncertainties in different input parameters: 1) the stress magnitude data for model calibration, 2) the material properties in the lithological units and 3) the subsurface geometry. On the other hand, concepts are developed to comprehensively quantify the complete uncertainties, for example with the help of surrogate models. At the same time, a reduction of uncertainties is developed by considering indirect data and observations of the stress state.

Project SQuaRe — The goal of the SQuaRe project represented in a Mohr-Coulomb diagram. Normal stress (X-axis) is plotted against the maximum shear stress (Y-axis) to show the stress state as a Mohr circle in relation to a failure criterion (straight line). Different stress states are possible due to the data situation (left). The methods developed in SQuaRe allow the stress states to be weighted and thus the prediction to be more precise (right). *(modified from Ziegler & Heidbach 2020, 2023)*

References:

Degen, D., Veroy, K., Freymark, F., Scheck-Wenderoth, M., Poulet, T. & Wellmann, F. (2021) Global sensitivity analysis to optimize basin-scale conductive model calibration – A case study from the Upper Rhine Graben. Geothermics 95. https://doi.org/10.1016/j.geothermics.2021.102143

Degen, D., Veroy, K. & Wellmann, F. (2022) Uncertainty quantification for basin-scale geothermal conduction models. Scientific Reports 12, 4246. https://doi.org/10.1038/s41598-022-08017-2

Hergert, T., O. Heidbach, K. Reiter, S. Giger, and P. Marschall (2015) Stress field sensitivity analysis in a sedimentary sequence of the Alpine foreland, northern Switzerland, Solid Earth, 6, 533-552. http://doi.org/10.5194/se-6-533-2015

Reiter, K., and O. Heidbach (2014) 3-D geomechanical-numerical model of the contemporary crustal stress state in the Alberta Basin (Canada), Solid Earth 5(2), 1123–1149. http://doi.org/10.5194/se-5-1123-2014

Ziegler, M.O. (2022) Rock Properties and Modelled Stress State Uncertainties: A Study of Variability and Dependence. Rock Mechanics and Rock Engineering 55, 4549–4564. https://doi.org/10.1007/s00603-022-02879-8

Ziegler, M. O., and O. Heidbach (2023) Bayesian Quantification and Reduction of Uncertainties in 3D Geomechanical‐Numerical from the Upper Rhine Graben. Geothermics 95. https://doi.org/10.1016/j.geothermics.2021.102143