Measuring a Low Horizontal Hydraulic Gradient in a High Transmissivity Aquifer.

Abstract

High transmissivity aquifers typically have low hydraulic gradients (i.e., a flat water table). Measuring low gradients using water levels can be problematic because measurement error may be greater than the true difference in water levels (i.e., a low signal-to-noise ratio). In this study, the feasibility of measuring a hydraulic gradient in the range of 10-6 to 10-5 m/m was demonstrated. The study was performed at a site where the depth to water from land surface ranged from 40.1 to 94.2m and the aquifer transmissivity was estimated at 41,300 m2 /d (hydraulic conductivity of 18,800 m/d). The goals of the study were to reduce measurement error as much as practicable and assess the importance of factors affecting water level measurement accuracy. Well verticality was the largest source of error (0.000 to 0.168 m; median of 0.014 m), and geodetic survey of casing elevations was the next most important source of error (0.002 to 0.013 m; median of 0.005 m). Variability due to barometric pressure fluctuations was not an important factor at the site. Hydraulic heads were measured to an accuracy of ±0.0065 m, and the average hydraulic gradient was estimated to be 8.0 × 10-6 (±0.9 × 10-6 )m/m. The improvement in accuracy allowed for two reversals in the groundwater flow direction to be identified, after which the gradient averaged 2.5 × 10-5 (±0.4 × 10-5 )m/m. This study showed it is possible to sufficiently control sources of error to measure hydraulic gradients in the 10-6 to 10-5 m/m range.