home > reports > crwr online report 2005-13

Characterization of Stormwater Runoff from a Bridge Deck and Approach Highway, Effects on Receiving Water Quality in Austin, Texas

by Joseph F. Malina, Jr., Ph.D., P.E., DEE, Blaine W. Linkous, M.S.E., Tejashri S. Joshi, M.S.E., and Michael E. Barrett, Ph.D., PE

ABSTRACT

Nonpoint source pollution represents one of the largest environmental problems currently facing water quality professionals. A fraction of this pollution is conveyed to receiving waters by stormwater drainage from highways. Some highway runoff is treated by structural or non-structural systems (best management practices [BMPs]) or is diverted to municipal treatment systems depending on locale. However, much highway runoff and almost all bridge deck runoff enter receiving streams without treatment. Highway runoff may contain suspended solids, metals, oil and grease, fecal coliform, and oxygen demanding organics. Highway runoff characteristics have been reported in some detail over the years; however limited data on the characteristics of runoff from bridge decks are available. The objectives of this study are: 1) characterization of bridge deck and approach highway stormwater runoff in Austin, Texas 2) a statistical comparison of the water quality characteristics of stormwater runoff from the two sources, and 3) an assessment of the impacts of the runoff on the quality of the receiving water. Flow-weighted composite and grab samples of runoff were collected from a bridge and approach highway. The average daily traffic count was 58,000 vehicles per day. The sampling period extended over a period of more than one year. ISCOŽ automatic flow monitoring and sampling equipment was installed to record runoff flow and collect samples at the two sites. The samples were analyzed for a suite of runoff constituents including: total and volatile suspended solids, (TSS/VSS), total and dissolved metals, phosphorus, nitrogen species, chemical oxygen demand (COD), and coliform organisms. A total of 15 storm events at the bridge site and 16 storm events at the adjacent approach highway were sampled. The receiving water is Barton Creek, an ephemeral stream with peak flows exceeding 30,000 ft3/s. Water quality and flow data for Barton Creek at Loop 360 were obtained from the United States Geological Survey (USGS) and compared with the composite runoff samples from the bridge deck. The initial runoff data from the two sites were compared with runoff data observed at other highway sites in Austin, TX as well as national sites to determine the relative quality of the samples collected in this study. The observed data for the two sites were representative of local conditions, but generally the concentrations of constituents were lower than the concentrations reported for samples collected nationally. The bridge and approach highway runoff data were subjected to paired-event hypothesis testing to establish any significant differences in concentrations observed for the bridge site and approach highway sites. The results of the paired events testing indicate that the concentrations observed at the bridge site were significantly lower or statistically equal to the concentrations observed at the approach highway site. Therefore highway runoff data may be used as a conservative surrogate for bridge runoff for total copper, dissolved copper, total lead, COD, total phosphorus, total Kjeldahl nitrogen (TKN), TSS, and VSS in the absence of bridge runoff data. Highway runoff data for total zinc, dissolved zinc, nitrate-nitrogen (NO3-N), dissolved phosphorus, fecal coliform, and oil and grease may be used as a more accurate proxy for bridge deck runoff. There was no instance in which the concentration observed at the bridge site was significantly higher than that at the approach highway. At the average storm flow, concentrations for total metals (copper, lead, and zinc) and volatile suspended solids from the Loop 360 bridge were higher than the average concentrations in Barton Creek at the bridge site. The average concentrations of all other constituents for which Barton Creek data are available were lower in the bridge runoff than in the Creek. The loading from bridge deck runoff was calculated to estimate pollutant contributions from the Loop 360 bridge to Barton Creek. The results indicate that the impact of storm water runoff from the bridge deck was small. Loading of typical storm water constituents was much greater in the creek upstream of the bridge than the load contributed by the bridge deck runoff. The difference was several orders of magnitude in most cases. For example, the total suspended solids load upstream of the bridge is 7 x 106 kg/yr (1.54 x 107 lb/yr) compared to the load of 257 kg/yr (567 lb/yr) contributed by the runoff from the bridge. The greatest increase in annual loading from the bridge deck among all the storm water constituents analyzed was 0.056% for total zinc. Therefore, the storm water runoff from the Loop 360 bridge has very little impact on the water quality of Barton Creek.

The highlighted links that follow are connected to Adobe pdf files of the corresponding material. To view them you must have the Adobe Acrobat Reader 3.0.

Download the Adobe Reader

View the Report...(1263 KB)

These materials may be used for study, research, and education, but please credit the authors and the Center for Research in Water Resources, The University of Texas at Austin. All commercial rights reserved. Copyright 2002 Center for Research in Water Resources.