A multiscale approach to determine hydraulic conductivity in thick claystone aquitards using field, laboratory, and numerical modeling methods

• Published in: Water resources research 2016-07, Vol.52 (7), p.5265-5284
• Main Authors: Smith, L. A., Barbour, S. L., Hendry, M. J., Novakowski, K., van der Kamp, G.
• Format: Article
• Language: English
• Subjects: Aquifers; Aquitards; Bentonite; Clay; Computer simulation; Concrete; Cores; Coring; Cretaceous; Current distribution; Drawdown; Drilling; Finite element method; Fractures; Groundwater flow; Grout; Hydraulic conductivity; Hydraulics; in situ; Laboratories; Laboratory tests; Mathematical models; modeling; Modelling; Multiscale analysis; Numerical methods; Numerical models; Piezometric head; Pore pressure; Potash; Potassium carbonate; Pressure; Pressure measurement; Pressure recovery; Pressure reduction; Pressure transducers; Stress concentration; Transducers; Two dimensional models
• ISSN: 0043-1397; 1944-7973
• DOI: 10.1002/2015WR018448

Characterizing the hydraulic conductivity (K) of aquitards is difficult due to technical and logistical difficulties associated with field‐based methods as well as the cost and challenge of collecting representative and competent core samples for laboratory analysis. The objective of this study was to produce a multiscale comparison of vertical and horizontal hydraulic conductivity (Kv and Kh, respectively) of a regionally extensive Cretaceous clay‐rich aquitard in southern Saskatchewan. Ten vibrating wire pressure transducers were lowered into place at depths between 25 and 325 m, then the annular was space was filled with a cement‐bentonite grout. The in situ Kh was estimated at the location of each transducer by simulating the early‐time pore pressure measurements following setting of the grout using a 2‐D axisymmetric, finite element, numerical model. Core samples were collected during drilling for conventional laboratory testing for Kv to compare with the transducer‐determined in situ Kh. Results highlight the importance of scale and consideration of the presence of possible secondary features (e.g., fractures) in the aquitard. The proximity of the transducers to an active potash mine (∼1 km) where depressurization of an underlying aquifer resulted in drawdown through the aquitard provided a unique opportunity to model the current hydraulic head profile using both the Kh and Kv estimates. Results indicate that the transducer‐determined Kh estimates would allow for the development of the current hydraulic head distribution, and that simulating the pore pressure recovery can be used to estimate moderately low in situ Kh (