Field scale evaluation of the In Situ Permeable Flow Sensor and assessment of river-aquifer interaction at the Brazos River Hydrologic Field Site / by Andrew Scott Alden

Abstract

Two In Situ Permeable Flow Sensors (ISPFS), recently developed by Dr. Sanford Ballard at Sandia National Laboratories, were field tested at the Brazos River Hydrologic Field Site near College Station, Texas. The Flow Sensors use a thermal perturbation technique to quantify the magnitude and direction of ground water flow in three dimensions. In the first phase of testing, Flow Sensor results and piezometric data from monitoring wells at the site were used to monitor interactions between the Brazos River and the adjacent flood plain aquifer. Components of groundwater flow as determined from a piezometer gradient analysis of monitoring wells at the site were compared with results from the Flow Sensors. In the second phase of testing, two aquifer pumping tests lasting 8 and 13 days were used to field test the accuracy of the Flow Sensors. Saturated hydraulic conductivity values derived from previous, long-term analyses were used in the piezometric gradient analysis. The changes in velocity and direction of groundwater flow induced by the pump were evaluated using Flow Sensor data and piezometric analyses. Flow Sensor performance was assessed by comparison to piezometric analysis results. Statistical analyses of long-term Flow Sensor data and piezometric data from the first phase of testing resulted in saturated hydraulic conductivity values of approximately 28.9 m/day at a depth of 13.7 m and 16.5 m/day at a depth of 18.3 m. Assessment of river-aquifer interactions indicates that a direct and measurable linear relationship exists between river stage and groundwater flow components at the site. Differences in the response of vertical groundwater flow to river stage at each Flow Sensor location suggests that lower portions of the aquifer may be less permeable than upper portions of the aquifer. During pump tests in the second phase of testing, groundwater flow vectors measured using piezometric and Flow Sensor data correlated well. The azimuths obtained from both methods predicted the approximate position of the pumping well at both Flow Sensor locations accurately.