Response curves for phosphorus plume lengths from reactive-solute-transport simulations of onland disposal of wastewater in noncarbonate sand and gravel aquifers

Abstract

Surface-water resources in Massachusetts often are affected by eutrophication, excessive plant growth, which has resulted in impaired use for a majority of the freshwater ponds and lakes and a substantial number of river-miles in the State. Because supply of phosphorus usually is limiting to plant growth in freshwater systems, control of phosphorus input to surface waters is critical to solving the impairment problem. Wastewater is a substantial source of phosphorus for surface water, and removal of phosphorus before disposal may be necessary. Wastewater disposed onland by infiltration loses phosphorus from the dissolved phase during transport through the subsurface and may be an effective disposal method; quantification of the phosphorus loss can be simulated to determine disposal feasibility. In 2003, the U.S. Geological Survey, in cooperation with the Massachusetts Department of Environmental Protection, initiated a project to simulate distance of phosphorus transport in the subsurface for plausible conditions of onland wastewater disposal and subsurface properties. A coupled one-dimensional unsaturated-zone and three-dimensional saturated-zone reactive-solute-transport model (PHAST) was used to simulate lengths of phosphorus plumes. Knowledge of phosphorus plume length could facilitate estimates of setback distances for wastewater-infiltration sites from surface water that would be sufficient to protect the surface water from eutrophication caused by phosphorus transport through the subsurface and ultimate discharge to surface water. The reactive-solute-transport model PHAST was used to simulate ground-water flow, solute transport, equilibrium chemistry for dissolved and sorbed species, and kinetic regulation of organic carbon decomposition and phosphate mineral formation. The phosphorus plume length was defined for the simulations as the maximum extent of the contour for the 0.015 milligram-per-liter concentration of dissolved phosphorus downgradient from the infiltration bed after disposal cessation. Duration of disposal before cessation was assumed to be 50 years into an infiltration bed of 20,000 square feet at the rate of 3 gallons per square foot per day. Time for the maximum extent of the phosphorus plume to develop is on the order of 100 years after disposal cessation. Simulations indicated that phosphorus transport beyond the extent of the 0.015 milligram-per-liter concentration contour was never more than 0.18 kilogram per year, an amount that would likely not alter the ecology of most surface water. Simulations of phosphorus plume lengths were summarized in a series of response curves. Simulated plume lengths ranged from 200 feet for low phosphorus-concentration effluents (0.25 milligram per liter) and thick (50 feet) unsaturated zones to 3,400 feet for high phosphorus-concentration effluents (14 milligrams per liter) discharged directly into the aquifer (unsaturated-zone thickness of 0 feet). Plume length was nearly independent of unsaturated-zone thickness at phosphorus concentrations in the wastewater that were less than 2 milligrams per liter because little or no phosphorus mineral formed at low phosphorus concentrations. For effluents of high phosphorus concentration, plume length varied from 3,400 feet for unsaturated-zone thickness of 0 to 2,550 feet for unsaturated-zone thickness of 50 feet. Model treatments of flow and equilibrium-controlled chemistry likely were more accurate than rates of kinetically controlled reactions, notably precipitation of iron-phosphate minerals; the kinetics of such reactions are less well known and thus less well defined in the model. Sensitivity analysis indicated that many chemical and physical aquifer properties, such as hydraulic gradient and model width, did not affect the simulated plume length appreciably, but duration of discharge, size of infiltration bed, amount of dispersion, and number of sorption sites on the aquifer sediments did affect plume length appreciably. Because simulation of plume length in carbonate-mineral sediments indicated that the plume would be substantially longer than in noncarbonate-mineral sediments, the application of the response curves in locations with carbonate-mineral sediments would be inappropriate. The effect of carbonate minerals in sediments is to increase pH, which causes decreased sorption of phosphorus on aquifer sediments. Phosphorus removal from solution by precipitation onto aquifer sediments is more efficient at high concentrations of disposed phosphorus than at low concentrations. At very low phosphorus concentrations, the solubility product of phosphorus minerals is not exceeded and no phosphorus mineral forms. An important consequence is that removal of dissolved phosphorus from the plume by processes in the subsurface is decreased the more that removal efforts are applied in treatment before wastewater is disposed. Model simulations indicate that removal of phosphorus from wastewater disposed through septic systems would have the advantage of efficient phosphorus removal in the subsurface because phosphorus concentrations are high in septic-system effluent. Short plume lengths result from wastewater disposal through septic systems because of the efficient phosphorus removal and because of the low volume of wastewater involved. The simulation results for small-volume systems are not quantitative, however, because wastewater-infiltration rates are much lower than those of the higher-volume system that was used to calibrate the model and to create the plume-length response curves. The response curves for phosphorus plume lengths, as defined by the maximum extent of the 0.015 milligram-per-liter concentration contour, is clearly defined in the model simulations, although the relation between simulated plume length and protective setback distance is subject to interpretation. Phosphorus does move beyond the point at which the simulated 0.015 milligram-per-liter concentration contour has stopped, so that a determination of protective setback distance must include a consideration of whether that continued flux, or some other flux amount, is appropriate. Also, simulations indicate that phosphorus plumes do not reach their full extent until 50 to 200 years after disposal cessation, depending on concentration of phosphorus disposed. No phosphorus plume has been monitored for that long after cessation, so there is no way to verify the long-term simulation results.