Abstract Archive Select a year below to view:
Field testing and simulation of vadose-zone recharge wells as an artificial recharge method in the Mississippi River Valley Alluvial aquifer
Proceedings of the 2020 Mississippi Water Resources Conference
Year: 2020 Authors: O'Reilly A.M., Kwak K., Rigby J.R.
Past studies in the Delta region of Mississippi document substantial groundwater losses from the Mississippi River Valley Alluvial aquifer (MRVAA) and indicate limited potential for infiltration and recharge due to fine-grained, low permeability surficial sediments. An artificial recharge technique not dependent on permeable surficial soils is a vadose-zone well. A vadose-zone well is a borehole, which does not intersect the saturated zone, excavated through low permeability surficial sediments and completed as a dry well into underlying higher permeability sediments. Water directed to the well flows by gravity into the native sediments of the vadose zone.
Data were collected at a field site near Ruleville, Mississippi, consisting of four vadose-zone wells, six monitor wells, and one production well. From pumping test data, transmissivity of the MRVAA at the site is 5,700 m2/day and storativity is 0.33. Despite being considered an unconfined aquifer, a distinct inverse correlation existed between barometric pressure and water level in the wells, indicating a barometric efficiency of approximately 60%. During a 50-hour injection test, well recharge caused small water-table rises ranging from 4 cm at the nearest monitor well (6.1 m) to 1 cm at the most distant well (35 m). Small rises likely are due to the high hydraulic conductivity of the MRVAA, vertical heterogeneity, screen location of the monitor wells, or some combination of these factors. Laboratory analyses included measurement of saturated hydraulic conductivity of vadose-zone soil cores. Additionally, wetting/draining curves were determined using the hanging water column method, representing some of the first measurements of capillary hysteresis in the Delta.
A three-dimensional numerical variably-saturated model of four vadose-zone wells was developed using HYDRUS-3D software. Pressure-head changes were reported at five observation nodes located 0.17 m below the water table. Head rise beneath the vadose-zone well was 2 cm and dropped to 0.6 cm at a distance of 6.3 m. Different water-table responses between the field test and model simulations are likely due to differences in the amount of injected water and lack of data on aquifer heterogeneity relative to monitor well screen locations. A total of 272 m3/day of water was injected during the field test, whereas only 88 m3/day was simulated in the HYDRUS model. This research provides understanding of the hydraulic properties controlling operation of vadose-zone wells. Challenges include clogging leading to limited well life and potential water-quality impacts caused by source water for the vadose-zone well bypassing shallow soil-aquifer treatment processes.