Simulation of the effects of water withdrawals, wastewater-return flows, and land-use change on streamflow in the Blackstone River basin, Massachusetts and Rhode Island

Abstract

Streamflow in many parts of the Blackstone River Basin in south-central Massachusetts and northern Rhode Island is altered by water-supply withdrawals, wastewater-return flows, and land-use change associated with a growing population. Simulations from a previously developed and calibrated Hydrological Simulation Program—FORTRAN (HSPF) precipitation-runoff model for the basin were used to evaluate the effects of water withdrawals, wastewater-return flows, and land-use change on streamflow. Most of the simulations were done for recent (1996–2001) conditions and potential buildout conditions in the future when all available land is developed to provide a long-range assessment of the effects of possible future human activities on water resources in the basin. The effects of land-use change were evaluated by comparing the results of long-term (1960–2004) simulations with (1) undeveloped land use, (2) 1995–1999 land use, and (3) potential buildout land use at selected sites across the basin. Flow-duration curves for these land-use scenarios were similar, indicating that land-use change, as represented in the HSPF model, had little effect on flow in the major tributary streams and rivers in the basin. However, land-use change—particularly increased effective impervious area—could potentially have greater effects on the hydrology, water quality, and aquatic habitat of the smaller streams in the basin. The effects of water withdrawals and wastewater-return flows were evaluated by comparing the results of long-term simulations with (1) no withdrawals and return flows, (2) actual (measured) 1996–2001 withdrawals and wastewater-return flows, and (3) potential withdrawals and wastewater-return flows at buildout. Overall, the results indicated that water use had a much larger effect on streamflow than did land use, and that the location and magnitude of wastewater-return flows were important for lessening the effects of withdrawals on streamflow in the Blackstone River Basin. Ratios of long-term (1960–2004) simulated flows with 1996–2001 water use (representing the net effect of withdrawals and wastewater-return flows) to long-term simulated flows with no water use indicated that, for many reaches, 1996–2001 water use did not deplete flows at the 90-percent flow duration substantially compared to flows unaffected by water use. Flows generally were more severely depleted in the reaches that include surface-water supplies for the larger cities in the basin (Kettle and Tatnuck Brooks, Worcester, Mass. water supply; Quinsigamond River, Shrewsbury, Mass. water supply; Crookfall Brook, Woonsocket, R.I. water supply; and Abbott Run, Pawtucket, R.I. water supply). These reaches did not have substantial wastewater-return flows that could offset the effects of the withdrawals. In contrast, wastewater-return flows from the Upper Blackstone Wastewater Treatment Facility in Millbury, Mass. increased flows at the 90-percent flow duration in the main stem of the Blackstone River compared to no-water-use conditions. Under the assumptions used to develop the buildout scenario, nearly all of the new water withdrawals were returned to the Blackstone River Basin at municipal wastewater-treatment plants or on-site septic systems. Consequently, buildout generally had small effects on simulated low flows in the Blackstone River and most of the major tributary streams compared to flows with 1996–2001 water use. To evaluate the effects of water use on flows in the rivers and major tributary streams in the Rhode Island part of the basin in greater detail, the magnitudes of water withdrawals and wastewater-return flows in relation to simulated streamflow were calculated as unique ratios for individual HSPF subbasins, total contributing areas to HSPF subbasins, and total contributing areas to the major tributary streams. For recent conditions (1996–2001 withdrawals and 1995–1999 land use), ratios of average summer (June through September) withdrawals to the long-term (1960–2004) medians of average summer streamflow simulated in the absence of water use ranged from 0.039 to 2.5 with a median value of 0.11 for total contributing areas to HSPF subbasins. The largest ratios of withdrawal rates to streamflow were for Crookfall Brook and Abbott Run, the subbasins with major withdrawals for municipal water supply. The smallest ratios were for the rural subbasins in the Branch River drainage area in the southwestern part of the basin. For recent conditions, ratios of average summer wastewater-return flows to average summer streamflows ranged from 0.0 to 0.20 with a median value of 0.029 for total contributing areas to HSPF subbasins. The largest ratios of wastewater-return flows to streamflows were for the subbasins that contained return flows from municipal wastewater-treatment plants and the subbasins along the Blackstone River because of high wastewater-return-flow rates from upstream facilities. Under the assumptions used to develop the buildout analysis, withdrawal and return-flow ratios were estimated to increase for most of the HSPF subbasins in the Rhode Island part of the basin. Ratios more than doubled for some subbasins, but the large increases mainly were for subbasins that had low ratios in 1996–2001. The HSPF model also was used to estimate the effects of water-conservation measures on low flows in rivers and major tributary streams in the Rhode Island part of the basin, the contribution of wastewater-return flows to streamflow in the Blackstone River, and the effects of changes to two local water supplies in Rhode Island. Water-conservation measures were evaluated by reducing 1996–2001 withdrawals by 20 percent. Simulations with 20-percent reductions in withdrawal rates indicated that conservation measures would result in appreciable increases in low flows in the subbasins with the highest withdrawal rates in the Rhode Island part of the Blackstone River Basin, whereas the effects on streamflow would be much less pronounced in subbasins with lower withdrawal rates. The contribution of wastewater-return flows to streamflow in the Blackstone River was evaluated by comparing simulated flows with and without municipal wastewater-return flows. Under typical summer low-flow conditions, treated wastewater was a major component of streamflow (35 to 50 percent) in the Blackstone River, and the percentage of treated wastewater was larger during the driest periods. The simulations conducted to evaluate changes to local water supplies (effects of potential withdrawals from an inactive well adjacent to Slatersville Reservoir in North Smithfield on flows in the Branch River, and the effects of connecting the town of North Smithfield to the water-supply system for the city of Woonsocket, Rhode Island) indicated that each of these activities would alter low flows only slightly in the associated stream reaches.