By Robert A. Phillips, John C. Clausen, John Alexopoulos, Bruce L. Morton, Stan Zaremba, and Mel Cote

Construction sites are commonplace in urbanized watersheds and can be a significant source of nonpoint-source pollution. Pollutants associated with construction-site runoff include sediment, nutrients, and metals. Most communities require developers to utilize erosion and sediment controls (ESCs) to reduce sediment loss during construction activities. ESCs are often inadequate due to poor installation and maintenance, lack of education, and an unwillingness to emphasize ESCs by developers (Paterson, 1994). New methods of retaining sediment on-site, reducing transport of pollutants to downstream waters, and preserving preexisting hydrologic conditions must be investigated to properly maintain water quality.

Traditional subdivision development

Previous studies have focused on hydrologic and sediment exports from construction sites where ESC practices were either not implemented or poorly maintained (Daniel et al., 1979; Madison et al., 1979). Impacts of construction on nutrient and metal concentrations in stormwater and mass export have not been well documented. This study, called the Jordan Cove Urban Watershed Project, compares the effectiveness of construction best management practices (BMPs) in a low-impact residential development to traditional construction practices using weekly storm flow, peak flow, pollutant concentration, and mass export.

This project is one of the Section 319 National Monitoring Program Projects funded by the United States Environmental Protection Agency through the states. The Jordan Cove Urban Watershed Project is a 10-year study aimed at reducing pollutants in stormwater runoff and keeping as much water on-site as possible. There are three phases to the study: calibration, construction, and postconstruction. This article reports on the construction impacts. In the future, we will report on the comparison of BMPs to traditional developments after construction completion.

Study Methods

BMP subdivision

The development is located in Waterford, CT, near Long Island Sound. Three watersheds are being monitored. The control watershed is 13.9 ac. and has 43 lots. The traditional watershed is 4 ac. with 17 lots. The BMP watershed is 6.9 ac. with 12 lots.

The control watershed was developed in 1989 and chosen as a control for climate variations year to year. This site was selected because we believed that there would be no significant watershed changes during the life of the study.

The traditional subdivision uses curb and gutter stormwater collection and a typical 26-ft.-wide asphalt road. Landscaping and turf management practices are similar to those of other new subdivisions and include open, spacious lawns; fertilization; lawn watering; and frequent grass trimming. Roof runoff is directed to lawn areas or onto driveways. ESCs used during construction, such as silt fencing, hay bales, and covered catch basins, were typical of other construction sites. Impervious surface coverage, consisting of asphalt and rooftops, increased from 4% to 35% of the total watershed area during construction.

Lots in the BMP subdivision, such as the one shown at top, incorporate vegetation and have comparatively little impervious surface.

Grass swales are used in place of curbs and gutters

The BMP subdivision incorporates several pollution prevention measures as part of its design. A 20-ft.-wide concrete-paver road was constructed in place of a typical asphalt road. Grass swales were used in place of curbs and gutters to direct concentrated surface flows while allowing infiltration and sedimentation to occur. A bioretention cul-de-sac that allows for detention and infiltration of runoff was constructed in lieu of a conventional paved area.

During construction, stockpiles were located away from runoff-producing areas and impervious surfaces to reduce the transport of sediment and pollutants off-site. Stockpiles and exposed soils were revegetated to prevent erosion and sediment loss. An earthen berm was constructed to retain and infiltrate stormwater on-site. All 12 basements were excavated at one time, providing onsite mini-detention basins. Poststorm maintenance was performed to ensure the efficacy of BMPs. At the time of this analysis, 4% of the BMP watershed was impervious.

The study design was developed as a paired watershed approach (Clausen and Spooner, 1993). This method uses two different time periods (calibration and treatment) and at least two watersheds (control and treatment). During calibration, no changes occur in the watersheds. During treatment, only the treatment watershed is modified. This technique eliminates the effects of climate differences from year to year on water quality. For example, if wet weather occurred during the calibration or baseline period but drier weather occurred during the treatment phase, then water-quality improvement would be expected due to less stormwater runoff.

Stormwater flow was continuously monitored from the three watersheds using a combination rectangular/v-notch weir in a stormwater pipe at the control site, a Palmer-Bowlus flume at the treatment site, and an H-flume at the BMP watershed. Different flow devices were used to suit the individual site characteristics. The v-notch weir was used at a pipe outlet, the Palmer-Bowlus flume within a pipe, and the H-flume for overland flow.

Composite stormwater samples were collected on a storm-event, flow-weighted basis and analyzed for nitrate nitrite–nitrogen (NO₃-N), ammonia-nitrogen (NH₃-N), total Kjeldahl nitrogen (TKN), total phosphorus (TP), total suspended solids (TSS), copper, lead, and zinc.

Results and Discussion

Storm Flow

Storm-flow volume decreased more than 100-fold during construction in the BMP watershed compared to the calibration period, but it increased 100-fold during construction in the traditional watershed compared to the calibration period (Figure 1). This storm-flow decrease in the BMP watershed can be attributed to BMPs used during construction. A 3-ft. earthen berm, constructed upstream of the BMP monitoring station, pooled water and obstructed flow to the station for several months during construction. Additionally, basements were excavated on all lots in a three-month period, resulting in "detention basins" that retained and infiltrated stormwater on-site. Fill used to raise the elevation of the area allowed for more water storage than existed before construction.

The increase in storm flow from the traditional watershed can be attributed to the construction of an asphalt road with a curb and gutter storm-sewer system on a watershed that was open field.

A 300% decrease in weekly peak storm flow was observed from the BMP watershed (Figure 2). In contrast, the traditional watershed experienced an 80% increase in weekly peak storm flow. Water retention measures at the BMP site lengthened the time of concentration of runoff, therefore lowering peak flows during the treatment period. Increases in impervious surface area coupled with the installation of a stormwater collection system on the traditional site shortened the time of concentration of runoff and contributed to increased peak flow from the site during construction.

Sediment

TSS concentrations increased 98% in storm flow from the BMP watershed (Figure 3). This increase can be primarily attributed to two BMP failures that affected 25% of the data collected. First, soil stabilization following construction of the swales was insufficient. Second, poststorm maintenance was not implemented following large storms. A more recent decline of TSS concentrations in storm flow from the BMP watershed can be attributed to the establishment of vegetation. Removal of 5-25% of larger suspended solids under ideal conditions in grassed swales has been suggested (Urbonas, 1994).

Surprisingly, increased TSS concentrations in storm flow from the traditional site during construction were not observed. This finding indicates that ESC implemented by the developer was effective in reducing erosion and retaining sediment on-site.

TSS export increased 97% from the BMP watershed. Higher TSS concentrations observed during the treatment period, as compared to during the calibration period, produced a significant increase in sediment export, even though flow decreased from the BMP watershed.

The traditional watershed experienced a 99% increase in TSS export due to the increase in storm flow. In a Wisconsin residential construction study, flow was the single most important variable explaining sediment load (Daniel et al., 1979).

Nitrogen

Concentrations of NO₃-N and NH₃-N in storm flow increased 50% and 77%, respectively, during construction of the BMP subdivision. Concentrations of NO₃-N and NH₃-N in storm flow did not significantly change during construction of the traditional subdivision. Concentrations of TKN in storm flow from the BMP watershed increased 74% during construction, perhaps due to imported soil that was high (9%) in organic matter.

Starter fertilizers applied at the time of grass seeding affected the concentration of nitrogen in BMP watershed runoff. Madison et al. reported that fertilization activities significantly increased nitrogen concentrations during construction of two subdivisions in Wisconsin. TSS, which increased in storm flow from the BMP site, has also been shown to carry a significant amount of nitrogen.

The mass export of NH₃-N in storm flow from the BMP watershed increased 72% during construction. There was no change in the mass export of NO₃-N and TKN in storm flow during construction of the BMP watershed.

The traditional watershed experienced increases in NO₃-N, NH₃-N, and TKN mass export in storm flow of 99%, 90%, and 98%, respectively, largely due to increased storm flow during construction.

Phosphorus

Concentrations of TP in storm flow increased 99% during BMP subdivision construction but did not increase at the traditional site (Figure 5). Fertilizer applied to the cul-de-sac and swales increased phosphorus concentrations. The increase in TP concentrations can also be attributed to the increase of sediment. The ESC practices implemented by the developer were adequate to prevent excessive sediment transport, and thus phosphorus did not leave the site.

A 98% increase in TP mass export from the BMP watershed was observed during construction. The increase in TP mass export coincides with the increase in TP concentrations in storm flow. In addition, a 99% increase in TP export in storm flow from the traditional watershed was observed during construction due to increased storm flow.

Metals

Concentrations of Cu (Figure 6) and Pb in storm flow from the BMP watershed increased 80% and 90%, respectively, during construction. No significant changes in Zn concentrations were observed. No significant changes in Cu, Pb, or Zn concentrations were detected in storm flow from the traditional watershed during construction.

No significant change in the mass export of metals was observed during construction in the BMP watershed. The lack of change in mass export, despite a significant increase in Cu and Pb concentrations in storm flow from the BMP watershed, was likely due to the reduction in storm flow observed during construction. In contrast, there were significant increases in the mass export of metals from the traditional watershed during construction. The mass export of Cu and Zn increased 99%. The mass export of Pb increased 98%. The increase in runoff volume, due to impervious surfaces, is responsible for the increase in the mass export of metals in storm flow from the traditional watershed.

Lessons Learned

Construction using traditional practices resulted in increased runoff and peak flow while construction in a low-impact subdivision resulted in reduced runoff and peak flow. Impervious surfaces associated with road construction and curb and gutter storm sewers were largely responsible for the flow increases in the traditional subdivision, whereas innovative design and the ability to retain and infiltrate stormwater on-site were keys to the reduction of runoff and peak flow in low-impact development.

Increases in sediment, nutrient, and metal concentrations and mass export in storm flow during low-impact development can be attributed to insufficient use of BMPs.

Single events, such as overfertilization and/or poor ESC maintenance before a large storm, are important and might substantially increase the detriment to water quality and receiving water bodies during construction.

Construction using traditional ESC practices might not increase sediment, nutrient, and metal concentrations in storm flow due to the creation of an impervious road that can prevent erosion. Increased mass exports of pollutants from traditional construction can result due to increased storm flow.

Generally water-quality and -quantity impacts in low-impact versus traditional subdivision construction can be viewed as a trade-off. Low-impact development construction can reduce storm flow; however, it can increase pollutant concentrations in storm flow if BMPs are not performed properly. Traditional subdivision construction might increase flow and thus pollutant export from the site, even if there is no observable increase in concentrations in stormwater.

References

Clausen, J.C. and J. Spooner. Paired Watershed Study Design. USEPA Pub. No. 841-F-93-009. US Environmental Protection Agency, Washington, DC. 1993.

Daniel, T.C., P.E. McGuire, D. Stoffel, and B. Miller. “Sediment and Nutrient Yield from Residential Construction Sites.” Journal of Environmental Quality, 8(3):304-8. 1979.

Madison, F.W., J.L. Arts, S.J. Berkowitz, E.E. Salmon, and B.B. Hagman. Washington County Project: Development and Implementation of a Sediment Control Ordinance or Other Regulatory Mechanism: Institutional Arrangements Necessary for Implementation of Control Methodology on Urban and Rural Lands. USEPA Pub. No. 905/9-80-003. USEPA, Washington, DC. 1979.

Paterson, R.G. “Construction Practices: The Good, the Bad, and the Ugly.” Watershed Protection Techniques, 1(3):95-99. 1994.

Urbonas, B. “Assessment of Stormwater BMPs and Their Technology.” Water Science and Technology, 29(1-2):347-53. 1994.

Robert A. Phillips is with Roy F. Weston Inc. in Glastonbury, CT; John C. Clausen and John Alexopoulos are with the University of Connecticut in Storrs; Bruce L. Morton is with Aqua Solutions LLC in East Hartford; Stan Zaremba is with the Connecticut Department of Environmental Protection in Hartford; and Mel Cote is with USEPA in Boston, MA.

SW - January/February 2003

Home + About + Subscribe + News + Calendar + Glossary
Talk + Images + Advertise + Contact Us + Search + Register + Services

Erosion Control Magazine | MSW Management Magazine
Grading & Excavation Contractor | ForesterPress | Forester Media

© 2000 - 2003 FORESTER MEDIA, INC.