Effects of runoff controls on the quantity and quality of urban runoff at two locations in Austin, Texas

Abstract

Rapid urban development in the Austin metropolitan area, Texas, is causing concern about increasing peak discharges from storm runoff and the degradation of the quality of water in receiving streams, lakes, and aquifers. In an attempt to decrease peak discharges and improve water quality, runoff controls are being required in some watersheds. This report summarizes the precipita- tion, streamflow, and water-quality data collected from September 1982 to September 1984 upstream and downstream from runoff controls at two locations, and presents the effects of these runoff controls on streamflow and the quality of runoff water. The two controls are a detention and filtering pond near Barton Creek Square Shopping Center, a large shopping center southwest of down- town Austin, and a grass-swale control in the Al ta Vista Planned Unit Develop- ment, a mul tipie-family housing area. At Barton Creek Square Shopping Center, rainfall for the storms analyzed ranged from 0.14 to 2.88 inches. The rainfall rate for the September 7, 1983, storm exceeded the 100-year return period for the 5and 10-minute duration and was equal to the 50-year return period for the 15-minute duration. Peak dis- charge at the inflow station to the detention pond was closely related to the maximum rainfall during a 5-minute period and occurred about 10 minutes later. The maximum inflow at this station was 185 cubic feet per second and appeared to be the limit of the storm sewer system. For smalland moderate-sized storms, the runoff is contained in the detention pond and passes through a fil- ter system. Runoff from large storms overflows into the drop outlet. For storms contained in the pond, peak discharges at the outflow station generally were less than 3.1 cubic feet per second. As time passed, the outflow peak discharges tended to decrease as a result of reduced permeability of the filter. Cleaning the filter appeared to increase the peak flows but did not restore them to the previous level. The runoff-rainfall ratio averaged 0.85 at the inflow station and 0.36 at the outflow station. A water budget shows unexplained losses to average 20 percent. At the Barton Creek Square Shopping Center, discharge-weighted densities of fecal-col iform and fecal-streptococci bacteria and discharge-weighted con- centrations of biochemical oxygen demand, chemical oxygen demand, total organic carbon, suspended solids, total ammonia plus organic nitrogen, and total phos- phorus generally were larger in the inflow than in the outflow. Discharge- weighted concentrations of dissolved lead, dissolved iron, and dissolved zinc generally were small in both the inflow and outflow; however, the larger discharge-weighted concentrations of these constituents generally were found in the inflow. Discharge-weighted concentrations of volatile dissolved solids were smaller in the inflow than in the outflow for 10 of the 22 storms analyzed. Discharge-weighted concentrations of total nitrite plus nitrate nitrogen and dissolved solids generally were much smaller in the inflow than in the outflow. It is likely that organic and ammonia nitrogen trapped in the pond from previous storms and in the inflow water as it flows through the pond is being oxidized to nitrite and nitrate nitrogen. Similiarly, dissolved solids retained in the filter or on the bed of the pond from previous storms are being leached to the outflow. Measured peak concentrations or densities of most constituents in the inflow were significantly larger than those in the outflow for most constitu- ents. An exception was noted for concentrations of total nitrite plus nitrate which were larger in the outflow than the inflow as indicated by discrete sample analysis for six storms. Loads of most constituents and total numbers of bacteria were significantly larger in the inflow than in the outflow. The total numbers of bacteria were reduced by approximately 80 percent. Average removal efficiencies for suspended solids, biochemical oxygen demand, total phosphorus, total organic carbon, chemical oxygen demand, and dissolved zinc ranged between 60 and 80 percent. The average loads of dissolved solids were approximately 13 percent larger in the outflow than the inflow. Average loads of total nitrite plus nitrate nitro- gen were approximately 110 percent larger in the outflow than in the inflow. The increase in loads of these constituents is due to material being leached from the bed of the pond or from the filter system. At Al ta Vista, rainfall for the storms analyzed ranged from 0.25 to 2.00 inches. The maximum rainfall intensity was 0.30 inch for a 5-minute interval. The runoff-rainfall ratio averaged 0.42 and appeared to be evenly distributed about the mean ratio line. The peak discharge at the inflow station to the grass-covered swale area was 0.93 cubic foot per second. Inaccuracies of dis- charge at the outflow station and variations in the ungaged drainage area with the size of the storm prevented a hydrologic analysis of the basin above this station. Discharge-weighted concentrations of total phosphorus were larger in the outflow than in the inflow for each of the 19 storms analyzed. Discharge- weighted concentrations of dissolved solids, volatile dissolved solids, bio- chemical oxygen demand, chemical oxygen demand, and total organic carbon were larger in the outflow than in the inflow for at least 12 of the 19 storms ana- lyzed. Discharge-weighted densities of fecal streptococci were decreased between the inflow and outflow, with discharge-weighted densities of fecal streptococci being less in the outflow for 15 of the 19 storms analyzed. Because of the relatively small variations in concentrations and densities of consti- tuents between the inflow and outflow sites, and because of the errors in discharge at the outflow gage, it is not feasible to determine the effct of the grass-covered swales on discharge-weighted concentrations and densities of water-quality constituents. Discrete concentrations or densities of most constituents were not de- creased. Peak concentrations of dissolved solids in the outflow exceeded peak concentrations in the inflow for all five of the storms analyzed with discrete -2samples. Peak concentrations of suspended solids, total ammonia plus organic nitrogen, total nitrite plus nitrate nitrogen, total nitrogen, and dissolved iron were larger in the outflow than in the inflow for four of the five storms analyzed. Load-removal efficiencies of water-quality constituents could not be determined because of inaccuracies in measuring discharge at the outflow site, INTRODUCTION The development of urban areas alters the quantity and quality of runoff that enters streams, lakes, reservoirs, and aquifers. Rapid urban development in the Austin metropolitan area, Texas, is causing concern about the impairment of the quality of water in streams, Lake Austin, Town Lake, and the Edwards aquifer. Lake Austin and Town Lake are water-supply reservoirs for the city of Austin and many nearby metropolitan areas. The Edwards aquifer south of the Colorado River near Austin is the source of water supply for many incorporated areas and urban developments in the vicinity of Austin, and discharges to Barton Springs, a popular recreation area. The city of Austin requires that runoff controls be provided for develop- ments in certain watersheds to reduce peak flows, or to minimize the impair- ment of the quality of water in streams and aquifers, or both. Local data are not available to determine the effectiveness of runoff controls in improving the quality of stormwater runoff, or the effect of these controls on reducing peak discharges. In order to help design future runoff controls, data are needed by city planners and developers to determine the effectiveness of present control structures. In 1982, the U.S. Geological Survey, in cooperation with the city of Austin, began a study to determine the effectiveness of runoff controls on the quantity and quality of urban runoff at two locations in Austin. These locations represent two types of runoff control structures that are commonly used in the Austin metropolitan area. Purpose and Scope This report describes the results of a study to determine the quantity and quality of runoff from a shopping center and a multiple-family residential development, and to determine the effectiveness of the runoff controls (deten- tion pond and grass swale) on the storm runoff at the two locations. One site, Barton Creek Square Shopping Center (BCSSC), is a large shopping center located southwest of downtown Austin (fig. 1). The other site, Alta Vista Planned Unit Development (Alta Vista), is a multiple-family residential neighborhood located in northwest Austin (fig. 1). The data-collection period began in September 1982 and concluded in September 1984. Approach To meet the study objectives, rain gages and streamflow and water-quality stations were established and operated at the principal points of inflow and outflow from the runoff control structures at each location. Rainfall and streamflow data were analyzed to show the relation between rainfall and runoff, the change in peak flow between the inflow and outflow of the runoff controls, and the water budget. Water-quality data were analyzed by comparing the discharge-weighted and peak concentrations and loads of selected constituents computed at the inflow station with values computed at outflow stations. -3-