Uncertainties in models for hydrologic characterization of faults - implications from a study on the Wildcat Fault, Berkeley, California -
Abstract
The Nuclear Waste Management Organization of Japan (NUMO) will select a site for HLW and TRU waste repository through the three-staged program, namely, the Literature Surveys, the Preliminary Investigations and the Detailed Investigations. Areas that are susceptible to natural hazards such as volcanism, faulting and significant uplift/erosion will be eliminated in the early stages. Then, sites that have more favorable geological environment will be selected with respect to the repository design and long-term safety after closure. It is internationally acknowledged that hydrologic features of faults are of special concern in the above respects. It is highly likely from the experiences of site characterization worldwide that one could encounter numerous faults in an area of one hundred square kilometer assumed for the Preliminary Investigations. Efficient and practical investigation programs, and reliable models/parameters for the repository design and safety analysis are important aspects for implementers. Therefore, we studied a methodology for hydrologic characterization of faults through the multi-phased field investigations and modeling on the Wildcat Fault at the LBNL site. The field investigations consisted surface geophysics, geological mapping, trench surveys, hydraulic tests and monitoring both in vertical and inclined boreholes. In general, the degree of understanding of the site increased along with the progress of the investigation phases. At the end of the project, relatively large uncertainties remained in the hydrogeological model. The major reasons were that two different interpretations on the fault geometry were possible, and that hydrologic data on the fault and host rock were not enough because of the difficulties in hydraulic tests in the fracture zones and the borehole configuration under various constraints. Implications for reducing uncertainties in hydrologic characterization of faults based on our experience are summarized below. - More time, costs and flexibility are required than common investigation/modeling program. - Long-term hydraulic tests and monitoring in the vertical boreholes could better capture important features than penetrating the faults by inclined boreholes. - It is important to plan/modify investigation programs taking all possible conceptual models into account, not persisting on a single model.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.H24D..03G
- Keywords:
-
- 1829 HYDROLOGY / Groundwater hydrology;
- 1847 HYDROLOGY / Modeling;
- 1873 HYDROLOGY / Uncertainty assessment