Open Collections

Methodologies for the use of electrical and electromagnetic methods in complex, subterranean environments

University of British Columbia

In underground mining, water inflow poses a serious danger that can result in loss of life and the destruction of infrastructure. In some situations, flooded mines have to be abandoned entirely. The DC resistivity field data used in this study comes from a potash mine in Saskatchewan which has been battling water inflow problems for over three decades. To help mitigate water related risks my research focuses on developing new data quality control and survey design methodologies for surveys collected in underground environments. Both methodologies improve the reliability of inversion results, which help engineers identify and monitor dangerous water-bearing regions. Mines are often noisy environments for data collection because of infrastructure. Since conventional data quality control methodologies are poorly adapted for large 3D, tunnel-based datasets I developed a new methodology. To reduce subjectivity and semi-automate the procedure I combine a manual search for noise sources with statistical analysis and classification tools, including singular value decomposition and k-means clustering. The methodology performed well on both field and synthetic datasets. In underground environments, conventional linear array DC resistivity surveys produce fundamentally non-unique inversions. A 3D ring array, with at least one electrode in each tunnel face, is required to constrain the around tunnel location of anomalies. With more electrodes and possible measurements, the full ring array survey is costly to collect. To balance model resolution and survey size I developed a physics-based survey design methodology where transmitters are selected using secondary charge accumulations on a series of test blocks. Results show that resolution can be retained with less than 5% of the ring array measurements. In search of further model resolution improvements, I extended the survey design methodology to a grounded source frequency domain electromagnetic survey. For comparison, the same electrode locations were used for the DC and EM surveys. Synthetic examples show that EM surveys better resolve the shape and location of many conductive anomalies. However, EM datasets are more complicated and expensive to invert. For most applications, a DC ring array reconnaissance survey is recommended. If necessary, a targeted EM survey can then be collected to improve resolution wherever needed.

Text

2020-04-14

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/74006

Geophysics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/