Evaluation of Bridge Deck Runoff and Its Effects on In-Stream Health

Abstract

Bridges typically drain through scuppers directly to streams or estuaries below, potentially resulting in degradation of surface water quality. The North Carolina General Assembly enacted Session Law 2008-107 in July 2008, which required the N.C. Department of Transportation to study the effects of stormwater runoff from bridges over waterways. The resulting study characterized 15 bridge decks across three ecoregions of N.C. for runoff quality and quantity. Monitoring sites were located across the Piedmont, Coastal Plain, and Mountain ecoregions of NC and had differences in wearing surface, annual average daily traffic, rural vs. urban watershed, bridge deck surface area, and stream drainage areas. The goals included characterizing bridge deck pollutants of concern and determining the effects of bridge deck runoff on in-stream health. Runoff water quality and quantity were measured at 15 bridges, instream water quality and quantity at 4 sites, streambed sediment quality at 30 sites, bioassay tests at 13 sites, and biosurvey tests at 15 sites. Median EMCs for TSS, TN, TP, total petroleum hydrocarbons (TPH), total copper, total lead, and total zinc were 39 mg/L, 0.97 mg/L, 0.17 mg/L, 3.1 mg/L, 9.6 μg/L, 5.3 μg/L, and 66 μg/L, respectively. Twenty-two parameters of concern, those with maximum observed concentrations above the strictest available state or federal threshold, were identified, including: pH, TSS, TN, TP, five total heavy metals, five dissolved heavy metals, and seven semi-volatiles. The best predictor of pollutant concentration was the land-use type of the watershed (urban vs. rural). Pollutant concentrations and loads observed from bridge decks were similar to or less than those from other highway and urban runoff studies. Pollutant loads from the bridge decks for all analytes studied were less than 0.25% of the pollutant load contributed by the whole watershed, since the ratio of bridge deck area to watershed area was always small ( Ceriodaphnia dubia bioassay tests. All samples were tested at 100% concentration, and bridge deck runoff samples also were tested at dilutions of 50%, 25%, 12.5%, and 6.25%. Of 25 bridge deck runoff samples, 3 exhibited toxicity at 100% concentration due to significantly reduced reproduction in the test organism. Potential reasons for this toxicity included elevated conductivity in one sample and low hardness and pH in another sample. No toxicity was observed at lower dilutions of the bridge deck runoff, indicating the potential for toxicity would be attenuated at instream concentrations. Of 20 instream stormflow and baseflow samples, no toxicity was observed. Instream mixing and large catchment area relative to bridge deck footprints (as much as 10 6 larger) resulted in substantial dilution of bridge deck runoff at these sites. Biosurveys (benthic macroinvertebrate sampling) were conducted upstream and downstream of 12 bridges. Samples were obtained using Qual 5 methodology, and bioclassification was determined using the EPT and N.C. biotic index metrics. One study site had a biotic index change large enough to suggest a decline in water quality downstream of the bridge deck. However, bioclassification rating was never different between upstream and downstream sampling areas. When the data were evaluated en masse, runoff concentrations and pollutant loads were similar to those from other urban and transportation runoff studies. Attempts to tie stormwater runoff to reduced in-stream health (sediment quality, benthic health, aquatic toxicity) were unsuccessful; this suggests that bridge deck runoff does not have widespread effects on receiving water quality.