Delaware Geological Survey Issues Report on Groundwater Monitoring and Water-Quality Impacts of Rapid Infiltration Basin Systems
B21C Groundwater Quality and Monitoring of Rapid Infiltration Basin Systems (RIBS), Theory and Field Experiments at Cape Henlopen State Park, Delaware
A rapid infiltration basin system (RIBS) consists of several simple and relatively standard technologies; collection and conveyance of wastewater, treatment, and discharge to an unlined excavated or constructed basin. By design, the effluent quickly infiltrates through the unsaturated or vadose zone to the water table. During infiltration, some contaminants may be treated by biological and/or geochemical processes and diluted by dispersion and diffusion. The combination of contaminant attenuation and dilution processes that may occur during infiltration and flow through the aquifer are termed soil-aquifer-treatment, or SAT. In the past decade, RIBS have been proposed more frequently for use in Delaware because they stop the direct discharge of treated effluent to surface water, can accommodate significant flow volumes typical of residential subdivisions, yet require much less land than options such as spray irrigation or sub-surface disposal systems.
Decades of research on the shallow Columbia aquifer of the Delmarva Peninsula have clearly identified the high susceptibility of the aquifer from land- and water-use practices, and the processes that control the fate and transport of contaminants from their origin at or near land surface to points of discharge in creeks, estuaries, and wells. The risk of aquifer contamination is great because it is highly permeable, has little organic matter in the aquifer matrix, and the depth to groundwater is very commonly less than 10 ft below land surface. USEPA guidance documents and several engineering texts that cover RIBS design clearly identify these same factors as increasing risk for groundwater contamination but do not provide much information on means to monitor and mitigate those risks. Further, design criteria are based on a small group of experiments conducted in the 1970s prior to development of current understanding of the processes that control groundwater contaminant transport.
Field and laboratory experiments to characterize the physical, chemical, and biological controls and processes associated with the rapid infiltration of treated sewage effluent through infiltration beds and the vadose zone were undertaken at a RIBS located at Cape Henlopen State Park (CHSP), Delaware. Field experiments to understand the geochemical effects of the long-term operation of a RIBS on ground and surface waters, and to evaluate monitoring systems were also conducted at the site. The CHSP RIBS has been in operation since the early 1980s.
Significant concentrations of nitrogen and phosphorus occur in groundwater from the point of effluent entry at the water table to distances greater than 150 ft from the infiltration beds. The high hydraulic, nitrogen (N), phosporus (P), and organic loading rates associated with the operation of RIBS overwhelm natural attenuation (e.g., sorption and precipitation) processes. Data are not sufficient to indicate whether denitrification is occurring. If there is denitrification, the rate is insufficient to remediate RIBS effluent at the site — despite a 25-ft thick vadose zone, an effluent with enough organic carbon to facilitate anaerobic conditions that permit abiotic denitrification and feed microorganism-driven denitrification processes, and hypoxic to anoxic groundwater.
Significant horizontal and vertical variability of contaminant concentrations were observed within the portion of the aquifer most impacted by effluent disposal. Despite the relatively small spatial extent of the disposal area in our study area, identification of the preferential flow zone and characterization of the vertical and temporal variability in the concentrations of contaminants required a multi-phase subsurface investigation program that included an analysis of data from samples collected at bi-monthly intervals from dozens of monitoring points and high frequency temperature monitoring in several wells. A well-designed monitoring system should be based on experimentally determined site specific evidence collected under conditions that duplicate the flow rates that are expected during full-scale operation of the RIBS. Conservative tracers should be used to determine if the monitoring wells are in locations that intercept flow from the infiltration beds.
The hydrogeologic framework of Cape Henlopen State Park (CHSP), Delaware was characterized to document the hydrologic effects of treated wastewater disposal on a rapid infiltration basin system (RIBS). Characterization efforts included installation of test borings and monitoring wells; collection of core samples, geophysical logs, hydraulic test data, groundwater levels and temperatures; testing of grain size distribution; and interpretation of stratigraphic lithofacies, hydraulic test data, groundwater levels, and temperature data. This work was part of a larger effort to assess the potential benefits and risks of using RIBS in Delaware.
The infiltration basins at CHSP are constructed on the Great Dune, an aeolian dune feature composed of relatively uniform, medium-grained quartz sand. The age of the dune, determined by carbon-14 dating of woody material in swamp deposits under the dune, is less than 800 years. Underlying the dune deposits are relatively heterogeneous, areally continuous, coarse-grained spit deposits of the proto-Cape Henlopen spit with interbedded and relatively fine-grained, discontinuous swamp and marsh deposits, and beneath, relatively fine-grained, continuous, near-shore marine deposits. The dune deposits can be 45 ft thick under the crest of the dune and nonexistent at the surface. Spit deposits range from 5 to 15 ft thick. Test drilling determined that the near-shore marine deposits are at least 10 ft thick in the vicinity of the infiltration basins. The complete thickness of these deposits was not determined in this study.
Hydraulic testing and grain-size data indicate that the dune and spit deposits are relatively permeable, with average hydraulic conductivities of 140 ft/day and that the swamp and marsh deposits are more than one order of magnitude less permeable, with average hydraulic conductivity of 25 to 10 ft/day. The water-table aquifer is present in the sandier dune and spit deposits. The swamp, marsh, and near-shore marine deposits form a leaky confining unit. The water-table aquifer is 15 to 20 ft thick under the thickest section of the Great Dune and nonexistent where the dune deposits are absent. The vadose zone is greater than 25 ft thick under the infiltration basins.
High-frequency groundwater level and temperature monitoring during periods of maximum wastewater disposal rates indicates that wastewater disposal causes increases in water-table elevations on the order of 1 ft. Groundwater elevations indicate that the water-table elevation is greatest under the infiltration basins and that most flow is directed southward toward a swampy discharge area.
Maximum disposal rates typically occur in summer months when the numbers of park users and water use are greatest. Coincident with greater disposal rates are higher wastewater temperatures. These higher wastewater temperatures are observed in groundwater and provide a means to track the flow of water from beneath the infiltration beds towards a nearby discharge area. Tracking of the warmer groundwater and modeling two-dimensional particle tracking both indicate that wastewater discharged to the infiltration basins reaches the nearby discharge area within 180 days.
RI79 Simulation of Groundwater Flow and Contaminant Transport in Eastern Sussex County, Delaware With Emphasis on Impacts of Spray Irrigation of Treated Wastewater
This report presents a conceptual model of groundwater flow and the effects of nitrate (NO3-) loading and transport on shallow groundwater quality in a portion of the Indian River watershed, eastern Sussex County, Delaware. Three-dimensional, numerical simulations of groundwater flow, particle tracking, and contaminant transport were constructed and tested against data collected in previous hydrogeological and water-quality studies.
The simulations show a bimodal distribution of groundwater residence time in the study area, with the largest grouping at less than 10 years, the second largest grouping at more than 100 years, and a median of approximately 29 years.
Historically, the principal source of nitrate to the shallow groundwater in the study area has been from the chemical- and manure-based fertilizers used in agriculture. A total mass of NO3- -nitrogen (N) of about 169 kg/day is currently simulated to discharge to surface water. As the result of improved N-management practices, after 45 years a 20 percent decrease in the mass of NO3- -N reaching the water table would result in an approximately 4 percent decrease in the mass of simulated N discharge to streams. The disproportionally smaller decrease in N discharge reflects the large mass of N in the aquifer coupled with long groundwater residence times.
Currently, there are two large wastewater spray irrigation facilities located in the study domain: the Mountaire Wastewater Treatment Facility and Inland Bays Wastewater Facility. The effects of wastewater application through spray irrigation were simulated with a two-step process. First, under different operations and soil conditions, evaporation and water flux, NO3- -N uptake by plants, and NO3- -N leaching were simulated using an unsaturated flow model, Hydrus-1D. Next, the range of simulated NO3- -N loads were input into the flow and transport model to study the impacts on groundwater elevation and NO3- -N conditions.
Over the long term, the spray irrigation of wastewater may increase water-table elevations up to 2.5m and impact large volumes of groundwater with NO3-. Reducing the concentration of NO3- in effluent and increasing the irrigation rate may reduce the volumes of water impacted by high concentrations of NO3-, but may facilitate the lateral and vertical migration of NO3-. Simulations indicate that NO3- will eventually impact deeper aquifers. An optimal practice of wastewater irrigation can be achieved by adjusting irrigation rate and effluent concentration. Further work is needed to determine these optimum application rates and concentrations.
To understand the effects of projected increased demands on groundwater for water supply, a finite-difference, steady-state, groundwater flow model was used to simulate groundwater flow in the Coastal Plain sediments of southern New Castle County, Delaware. The model simulated flow in the Columbia (water table), Rancocas, Mt. Laurel, combined Magothy/Potomac A, Potomac B, and Potomac C aquifers, and intervening confining beds. Although the model domain extended north of the Chesapeake and Delaware Canal, south into northern Kent County, east into New Jersey, and west into Maryland, the model focused on the area between the Chesapeake and Delaware Canal, the Delaware River, and the Maryland-Delaware border. Boundary conditions for these areas were derived from modeling studies completed by others over the past 10 years.
Compilation and review of data used for model input revealed gaps in hydraulic properties, pumping, aquifer and confining bed geometry, and water-level data. The model is a useful tool for understanding hydrologic processes within the study area such as horizontal and vertical flow directions and response of aquifers to pumping, but significant data gaps preclude its use for detailed analysis for water resources management including estimating flow rates between Delaware and adjacent states. The calibrated model successfully simulated groundwater flow directions in the Rancocas and Mt. Laurel aquifers as expected from the conceptual model. Flow patterns in the Rancocas and Mt. Laurel aquifers are towards local streams, similar to flow directions in the Columbia (water table) aquifer in locations where these aquifers are in close hydraulic connection.
Water-budget calculations and simulated heads indicate that deep confined aquifers (Magothy and Potomac aquifers) receive groundwater recharge from shallow aquifers (Columbia, Rancocas, and Mt. Laurel aquifers) in most of the study domain. Within shallow aquifers, groundwater moves toward major streams, while in the deep aquifers, groundwater moves
toward major pumping centers.
B16 Ground-Water Resources of the Piney Point and Cheswold Aquifers in Central Delaware as Determined by a Flow Model
A quasi three-dimensional model was constructed to simulate the response of the Piney Point and Cheswold aquifers underlying Kent County, Delaware to ground-water withdrawals. The model included the Magothy, Piney Point, Cheswold, and unconfined aquifers, and was calibrated using historical pumpage and water-level data. Model calibration was accomplished through the use of both steady-state and transient-state simulations.
The unconfined aquifer in central and southeastern Delaware occurs as a southward-thickening blanket of fine to coarse sand, and is recharged almost totally by precipitation and discharge is principally by seepage to streams, bays, and the ocean.