Physical Characterization of two fractured sedimentary rock aquifers in New Brunswick, Canada with emphasis on the development of self-potential methods
University of New Brunswick
Heterogeneity and anisotropy associated with fractured rock aquifers can make predicting fluid flow pathways difficult using traditional hydraulic testing methods alone. Surface-based self-potentials (SP) measured during pumping and hydraulic testing have been shown in this thesis to provide valuable information used to infer water table drawdown, preferential flow paths and hydraulic properties. To provide a basis for investigating the SP approach in fractured rock, conceptual models were first developed for two separate aquifers. Fracture patterns, combined with pump test and geophysical logging data allowed for flow directions to be assessed within a confined fluvial sandstone-shale aquifer underlying the Springdale wellfield in the Carboniferous Moncton Subbasin. In the Ordovician to Silurian Matapédia Basin, fracture characteristics were measured within the folded turbidite sequences underlying the Black Brook Watershed, and combined with hydraulic conductivity estimates derived from packer testing to provide a better understanding of heterogeneity responsible for anisotropic groundwater flow conditions. Both fractured aquifers were subsequently investigated through SP monitoring and numerical modelling. Transient SP signals recorded during pumping in the Springdale wellfield, combined with measurements of the electrokinetic voltage coupling coefficient, allowed spatial and temporal variations in drawdown to be inferred below electrodes positioned around the pumping well. SP-derived drawdown was fitted to the Theis model to obtain transmissivity and storativity estimates at electrode locations. Numerical modelling showed surface SP measurements to be an excellent proxy for hydraulic head at the top of an underlying confined aquifer even in the presence of hydraulic and electrical heterogeneities. SP signals recorded in the Black Brook watershed during a constant head injection test at a depth of 44 m were used to infer azimuthal anisotropy in fluid flow in directions consistent with dominant fracture set orientations. Numerical modelling showed that fracture transmissivity, length, and frequency all contribute to the magnitude and shape of SP anomalies recorded on surface resulting from flow from the highly transmissive fractures into the surrounding rock matrix. Results from the SP monitoring approach, combined with the conceptual aquifer models, demonstrated an inexpensive and non-invasive means of assessing water table drawdown pattern, preferential flow directions, and hydraulic properties estimates without the need for additional wells.