The South Carolina Department of Transportation operates section shed and maintenance yard facilities throughout the State. The U.S. Geological Survey conducted a cooperative investigation with the South Carolina Department of Transportation to characterize water-quality constituents that are transported in stormwater from representative maintenance yard and section shed facilities in South Carolina. At a section shed in Ballentine, S.C., stormwater discharges to a retention pond outfall (Ballentine). At the Conway maintenance yard, stormwater in the southernmost section discharges to a pipe outfall (Conway1), and stormwater in the remaining area discharges to a grass-lined ditch (Conway2). At the North Charleston maintenance yard, stormwater discharges from the yard to Turkey Creek through a combination of pipes, ditches, and overland flow; therefore, samples were collected from the main channel of Turkey Creek at the upstream (North Charleston1) and downstream (North Charleston2) limits of the North Charleston maintenance yard facility.

The storms sampled during this study had a wide range of rainfall amounts, durations, and intensities at each of the facilities and, therefore, were considered to be reasonably representative of the potential for contaminant transport. At all facilities, stormwater discharge was significantly correlated to rainfall amount and intensity. Event-mean unit-area stormwater discharge increased with increasing impervious surface at the Conway and North Charleston maintenance yards. The Ballentine facility with 79 percent impervious surface had a mean unit-area discharge similar to that of the North Charleston maintenance yard (62 percent impervious surface). That similarity may be attributed, in part, to the effects of the retention pond on the stormwater runoff at the Ballentine facility and to the greater rainfall intensities and amounts at the North Charleston facility.

Stormwater samples from the facilities were analyzed for multiple constituents and characteristics. Concentrations of sediment and concentrations of nutrients and fecal indicator bacteria, which are commonly transported with the sediment in stormwater, were measured. Total and dissolved concentrations of six trace metals were determined in the samples. Stormwater samples also were analyzed for organic compounds including 10 herbicides, 18 organochlorine pesticides, 7 Aroclor or polychlorinated biphenyl congeners, 44 volatile organic compounds, and 16 polycyclic aromatic hydrocarbons.

Stormwater often transports large quantities of sediment and sediment-bound contaminants, including nutrients and fecal indicator bacteria. Median event-mean concentrations of suspended sediment in stormwater at these facilities ranged from 54 milligrams per liter in Turkey Creek at North Charleston2 to 147 milligrams per liter in stormwater discharging from the Ballentine retention pond outfall. In general, event-mean concentrations of total nitrogen consisted mainly of total Kjeldahl nitrogen (organic nitrogen plus ammonia) rather than nitrate plus nitrite in stormwater, and the median event-mean concentrations of total nitrogen ranged from 1.59 milligrams per liter at the Conway1 pipe outfall to 2.00 milligrams per liter at the Ballentine retention pond outfall. Median event-mean concentrations of total phosphorus in stormwater ranged from 0.15 milligram per liter at the Conway1 outfall to 0.42 milligram per liter in Turkey Creek at North Charleston1.

Escherichia coli and enterococcus concentrations often varied by 3 to 4 orders of magnitude in grab samples collected during the “first flush” of stormwater discharging to the sampled outfalls of Turkey Creek. Additionally, enterococcus concentrations consistently were greater than the corresponding Escherichia coli concentrations in stormwater. Specifically, median "first-flush" Escherichia coli concentrations ranged from 30 colonies per 100 milliliters at the Conway1 outfall to 4,359 colonies per 100 milliliters in Turkey Creek at North Charleston2, whereas enterococcus concentrations ranged from 512 colonies per 100 milliliters at the Conway1 outfall to 6,329 colonies per 100 milliliters in Turkey Creek at North Charleston2. In comparison to the proposed South Carolina Department of Health and Environmental Control primary and secondary body contact criterion of 349 colonies per 100 milliliter, stormwater had Escherichia coli concentrations that were greater than the criterion in 4 of the 9 storms at Ballentine retention pond outfall, 1 of the 8 storms at the Conway1 pipe outfall, 5 of the 7 storms at the Conway2 grass-lined ditch outfall, 2 of the 8 storms at North Charleston1 on Turkey Creek, and 8 of the 8 storms at North Charleston2 on Turkey Creek.

Of the six trace metals measured in stormwater, only copper and zinc had event-mean concentrations greater than the hardness-dependent South Carolina Department of Health and Environmental Control aquatic life criteria maximum concentrations. Measured dissolved copper event-mean concentrations in stormwater were greater than the criterion in 5 of the samples at the Ballentine facility, 1 of the samples at Conway1, 2 of the samples at Conway2, and 1 of the samples at North Charleston2. Measured dissolved zinc event-mean concentrations in stormwater were greater than the criterion in 3 of the samples at the Ballentine facility, 1 of the samples at Conway1, 2 of the samples at Conway2, and 0 of the samples at North Charleston2. At North Charleston1 upstream from the North Charleston maintenance yard, the measured dissolved trace-metal concentrations were all less than the criterion maximum concentrations.

Among the three facilities, Conway1 outfall had the greatest range in event-mean yields in stormwater for total phosphorus, total nitrogen, total suspended solids, and suspended sediment, and both Conway outfalls tended to have median event-mean yields greater than those of the Ballentine and North Charleston yard facilities. "First-flush” yields of Escherichia coli in stormwater were not statistically different among the three facilities.

Median event-mean yields of suspended sediment, total nitrogen, total phosphorus, total copper, and total zinc in stormwater demonstrated a strong linear relation to impervious surface at the three facilities. However, median "first-flush" fecal indicator bacterial yields did not have a linear relation to impervious surface.