Stochastic characterization of discrete fractures in rock by hydraulic and tracer tomography

• Applicant: Professor Dr. Peter Bayer
• Subject Area: Hydrogeology, Hydrology, Limnology, Urban Water Management, Water Chemistry, Integrated Water Resources Management ; Palaeontology
• Term: from 2018 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 401048478

• Final Report Year: 2022

Final Report Abstract

This project was dedicated to resolving the three-dimensional (3D) hydraulic and structural properties of fractured rock. For this purpose, hydraulic tomography experiments are utilized. For those experiments, several intervals of adjacent boreholes are isolated by packer systems in different depths. A perturbation of the steady-state pressure is created by maintaining a predefined pumping rate in one of the intervals and the resulting pressure responses are recorded in the other nearby intervals. By repeating this procedure for each interval, a set of pressure signals is measured that serves as basis for the inversion of the hydraulic and structural properties of the fractured rock. In comparison to previous studies that utilize mainly continuous inversion methods, the properties of the fractured rock are integrated more directly by the discrete fracture network (DFN) model. Thereby, a DFN model is implemented for the execution of the forward simulations, the structural and hydraulic properties of a DFN are inferred, and the results are evaluated as fracture probability. The inversion is accomplished by a stochastic method based on the Bayesian equation that defines a posterior distribution of the DFN parameters given the measured data. The posterior distribution is characterized by sampling from it according to Markov chain Monte Carlo (MCMC) methods. An initial DFN configuration is updated iteratively by adding or deleting fractures or by updating the position, length, and hydraulic aperture of a fracture. This results in several posterior DFN realizations of approximately equal probability. The structural properties of the DFN realizations are processed to a fracture probability map over the investigated region and a mean hydraulic aperture indicating transmissivity and storativity of the fractures. In the project, the 3D inversion of discrete fractures was successfully accomplished by utilizing a streamlined fracture flow model as forward simulation tool. It was integrated in the iterative Bayesian inversion procedure to successfully reconstruct the DFNs in synthetic and field case applications. In the future, the developed inversion concept can be applied to facilitate the planning and design of an enhanced geothermal system, the characterization of an excavation-induced damaged zone, and monitoring of groundwater reservoirs.

Publications

• Comparison of Hydraulic and Tracer Tomography for Discrete Fracture Network Inversion. Geosciences, 9(6), 274.
  Ringel, Lisa Maria; Somogyvári, Márk; Jalali, Mohammadreza & Bayer, Peter
  
• Inversion of Three- Dimensional Discrete Fracture Networks Using Hydraulic Tomography. In AGU Fall Meeting Abstracts (Vol. 2020, pp. H129-03)
  Ringel, L. M., Jalali, M., & Bayer, P.
  
• Stochastic Inversion of Three‐Dimensional Discrete Fracture Network Structure With Hydraulic Tomography. Water Resources Research, 57(12).
  Ringel, Lisa Maria; Jalali, Mohammadreza & Bayer, Peter
  
• Characterization of the highly fractured zone at the Grimsel test site based on hydraulic tomography. (2022, 6, 29). American Geophysical Union (AGU).
  Ringel, Lisa Maria; Jalali, Mohammadreza & Bayer, Peter
  
• Estimation of hydraulic and geometrical characteristics of fractured geothermal reservoirs using insitu tomographic methods. European Geothermal Conference Proceedings, 5 pages
  Ringel, L. M., Jalali, M., & Bayer, P.
  
• Inversion of Hydraulic Tomography Data from the Grimsel Test Site with a Discrete Fracture Network Model. (2022, 3, 27). American Geophysical Union (AGU).
  Ringel, Lisa Maria; Jalali, Mohammadreza & Bayer, Peter