Pollutant Removal and Peak Flow Mitigation by a Bioretention Cell in Urban Charlotte, N.C.

Bioretention is a stormwater treatment practice that has gained popularity due to its aesthetics, potential to reduce flooding, and early documented improvements to stormwater quality. A bioretention cell in an urban setting was examined in Charlotte, N.C. from 2004 to 2006. Flow-weighted, composite water quality samples were collected for 23 events and analyzed for TKN, NH4-N, NO2-3-N, TP, TSS, BOD-5, Cu, Zn, Fe, and Pb. Grab samples were collected from 19 storms for fecal coliform and 14 events for Escherichia coli (E. coli). There were significant reductions (p<0.05) in the concentrations of TN, TKN, NH4-N, BOD-5, fecal coliform, E. Coli, TSS, Cu, Zn, and Pb. Iron concentrations significantly increased (p<0.05). NO2-3-N concentrations were essentially unchanged. Efficiency ratios for TN, TKN, NH4-N, TP, and TSS were 0.32, 0.44, 0.73, 0.31, and 0.60, respectively. Fecal coliform and E. coli efficiency ratios were 0.69 and 0.71, respectively. Efficiency ratios for Zn, Cu, and Pb were 0.77, 0.54, and 0.31, respectively. Concentrations of Fe increased by 330%. The peak outflow of the bioretention cell for 16 storms with less than 42?mm of rainfall was at least 96.5% less than the peak inflow, with a mean peak flow reduction being 99%. These results indicated that in an urban environment, bioretention systems can reduce concentrations of most target pollutants, including pathogenic bacteria indicator species. Additionally, bioretention can effectively reduce peak runoff from small to midsize storm events.     

Estimating Pollutant Mass Accumulation on Highways during Dry Periods

For determining the accumulated pollutant mass on highways, two years of monitoring data were used from eight highway sites in southern California. Buildup over antecedent dry days was calculated from mass washed off from the following storm and retained pollutant mass. Mass accumulation rates were determined for total suspended solids (TSS), chemical oxygen demand (COD), oil and grease, total Kjeldahl nitrogen, and total phosphorus, and are reported in g/m2-day. A revised buildup model is proposed using an alternative modeling approach to describe buildup during dry days between storms. The result shows that, between 1 and 10 antecedent dry days, the pollutant mass buildup rates are determined to be 0.544?g/m2-day for TSS, 0.114?g/m2-day for COD, and 0.0113?g/m2-day for oil and grease. Buildup rates decline in subsequent periods rates decreased by 79% for TSS, 78% for COD, and 61% less for oil and grease in the following 10–70?day period.     

New Approach to Evaluate Pollutant Removal by Storm-Water Treatment Devices

A methodology was developed to monitor and evaluate the removal of solids and associated constituents by a nutrient separating baffle box (NSBB) storm-water treatment device treating runoff from a 4.3 ha (10.6 acre) residential watershed discharging into the Indian River Lagoon, Florida. The NSBB was monitored over a 359-day time period using autosamplers to quantify water column removal during runoff events, and by quantifying and analyzing solids that accumulated within the NSBB. Flow composited influent and effluent samples were collected to represent water column performance. Event mean concentration (EMC) reduction was moderate (mean: 17%) and variable (range: ?39 to 68%) for suspended solids, and negative for nitrogen, phosphorus, fecal coliforms chromium, and copper. The mass of solids that accumulated in bottom chambers and in a strainer screen was quantified and analyzed for nitrogen, phosphorus, heavy metals, and polycyclic aromatic hydrocarbons. A quantitative evaluative framework was devised to estimate the total pollutant mass removal by NSBB, which consisted of the summation of the separately calculated mass removals for water column, bottom chamber material, and strainer screen material. The water column accounted for only 4% of total solids that accumulated in the NSBB, which was equally divided between bottom chamber and strainer screen. Removal of nitrogen, phosphorus, and metals could be accounted for only by considering mass accumulations. Results suggest that overall assessment of pollutant removal by NSBB must be cognizant of the materials not captured by typical autosamplers: larger size sediment particles, large floating and suspended matter, and the pollutants associated with these materials. Using water column EMCs as the sole measure of performance significantly underestimated loading reduction of storm-water constituents by the NSBB. The monitoring and evaluative methodology applied to the NSBB may be applicable to load reduction evaluations for other storm-water treatment devices with a similar function.     

Accretion and Partitioning of Heavy Metals Associated with Snow Exposed to Urban Traffic and Winter Storm Maintenance Activities. II

Compared to storm water runoff, urban roadway snow exposed to traffic and winter maintenance practices has a much greater capacity to accumulate and retrain heavy metals and other anthropogenic constituents. Heavy metals once released in the environment are not degraded and partition between the dissolved and particulate-bound fractions. Residence time, solid loadings, alkalinity, hardness, and pH influence partitioning. Accretion and partitioning of Pb, Cu, Cd, Zn, Al, Ca, Na, Mg, and Fe from a series of urban highway sites in Cincinnati, Ohio, are compared to temporal accretion trends at a control site removed from the highway environment. Results from partitioning analysis indicate that Pb, Cu, Cd, Zn, Al, Mg, and Fe were all highly particulate bound, while Na and Ca were mainly dissolved for all highway sites. Partition coefficients for most heavy metals in snowmelt ranged from 103 to 106?L/kg. Concentrations for Pb, Cu, Cd, Zn, and cyanide were orders of magnitude higher than at the control site and exceeded storm water runoff concentrations by one to two orders of magnitude. For residuals analyses, the specific surface area generally increased with decreasing particle size while the predominance of total surface area (SA) was associated with the medium to coarser size fractions. Heavy metal mass trends followed similar general trends to that of the SA. Characterization of accretion and partitioning of these metals is a necessary first step toward development of management and treatment strategies designed to address urban snow pollution.     

Storm Flow from First-Flush Precipitation in Stormwater Design

Regulations for mitigating nonpoint source pollution from urban areas often include a requirement for treatment of a first-flush depth of stormwater. When the stormwater treatment technology requires specification of a design flow rate, the first-flush depth must be related to a first-flush flow rate. This paper describes a methodology to determine the relation between accumulated storm depth and corresponding flow to a treatment device for small, urban watersheds. The approach uses the rational method under the assumption that the time-of-concentration is small. Intensity–frequency relations, expressed in the form of return periods and associated intensities, were determined using a partial duration frequency analysis of regionalized precipitation data. The results of the method can be used to determine the flow rate to a treatment device that will meet a specified return period for a first-flush depth. The method is demonstrated using 13 years of 15-min data from seven Massachusetts precipitation stations.     

Investigation of Boundary Shear Stress and Pollutant Detachment from Impervious Surface during Simulated Urban Storm Runoff

Pollutant detachment rates have been determined for four chloride salts during simulated urban storm runoff. Under rainfall and/or overland flow conditions, chloride mass flux was measured and related to boundary shear stress of the test surface. Washoff coefficients, presumed to depend only on pollutant characteristics, were computed based on the slopes of dimensionless mass flux versus dimensionless time plots. Washoff coefficients were found to vary among and between the chloride compounds studied. In general, higher overland flow rates produced lower boundary shear and lower washoff coefficients. The combination of simulated rainfall and overland flow resulted in an increased boundary shear and an increased washoff coefficient. An empirical washoff coefficient based on a load characteristic curve derived from an exponential washoff relationship was also computed from the runoff data and compared with the previous washoff coefficient. A linear correlation between these two washoff coefficients was observed. The magnitude of the latter coefficient under simulated rainfall was consistent with reported values obtained from field data.     

Effects of a Permeable Friction Course on Highway Runoff

This project documents the impact of a porous asphalt overlay on the quality and quantity of highway storm-water runoff. A permeable friction course, also known as open graded friction course, is a layer of porous asphalt approximately 50?mm thick, which is often applied on top of conventional asphalt or concrete highways to enhance safety and reduce noise. Storm-water runoff from a four-lane divided highway in the Austin, Texas area was monitored at two sites before and after the installation of a PFC. Observed concentrations of total suspended solids and pollutants associated with particulate material were much lower in the runoff from the PFC than that derived from the conventional asphalt surface. Concentration reductions were observed for total suspended solids, total lead, total copper, and total zinc at both monitoring locations. In addition to the above-mentioned constituents, concentrations of chemical oxygen demand and total Kjeldahl nitrogen were also lower in the runoff from the PFC at a site collecting paired samples from both pavement types. Concentrations of dissolved constituents were not significantly different and concentrations of polycyclic aromatic hydrocarbons were below the detection limit for both pavement types. The runoff coefficient for the PFC appears to be higher than for conventional pavements.     

Particle Size Distribution in Highway Runoff

Particles in highway runoff contain various sorbed pollutants, and many best management practices (BMPs) are selected for particle removal efficiency, which makes particle size distribution a crucial BMP design parameter. Particles between 2 and 1,000?μm in diameter were quantified for three rainfall events during the 2002–2003 rainy season at three highway sites in west Los Angeles. Rainfall, runoff flow rate, and a large suite of water quality parameters were also measured. An experimental protocol was developed for bottle cleaning, sample storage, and mixing that provided repeatable results. Particle aggregation occurred which required samples to be analyzed in less than 6?h; the concentration of small particles decreased with a corresponding increase in the concentration of larger particles in stored samples. The particle concentration decreased as the storm progressed and the number of large particles decreased more rapidly than the total number of particles. Particles demonstrated a strong first flush. On average, 40% of the particles were discharged in the first 20% of the runoff volume.     

Modeling Flow and Pollutant Removal of Wet Detention Pond Treating Stormwater Runoff

A mathematical model of pollutant removal by wet ponds was developed based on the mass balance principle and the release–storage equation. The release–storage equation can be linear or nonlinear depending on the wet pond shape and the spillway crest features. If the exponent index of the storage relationship (d) is equal to the index of the outflow equation (b), the wet pond hydrological routing model is linear. Otherwise, the model is nonlinear. Substituting the release–storage equation into the continuity equation produces a nonlinear ordinary differential equation (ODE). A Runge–Kutta method was used to solve the resulting nonlinear ODE. When the ratio of indexes d/b = 2/3, the hydrologic wet pond model reduces to a special case that leads to an implicit analytical solution. The pollutant removal and flow routing models were tested with data obtained from an actual wet pond for treating highway runoff. The predicted flow discharges and pollutant concentrations compared well with the observed data.     

Heavy Metal and PAH Concentrations in Highway Runoff Deposits Fractionated on Settling Velocities

The correlation between settling velocity and associated pollutant concentrations is of major importance for best management practice in designing, redesigning, or evaluation of the efficiency of existing pond facilities for retaining unwanted pollutants. The prospect of this note is to state the relationship between the settling velocity of the runoff particles and the corresponding metal and polyaromatic hydrocarbon (PAH) concentration directly instead of dealing with two unknowns—the density and the shape of a single particle fraction in a settling velocity calculations. The measurements show that the highest cadmium, chromium, zinc, and nickel concentrations is associated with the most slowly falling particles and the lowest concentration associated within the faster falling fraction. This tendency is not clear for some of the sediments due to high content of organic matter and clearly not for lead and copper and there is no significant correlation between PAH concentration and settling velocity. The largest mass of metals and PAH within each pond can be found on the particle fraction with a settling velocity of 5.5–2.5 mm/s.     

Water Quality Improvement through Reductions of Pollutant Loads Using Bioretention

As an increasingly adopted storm-water best management practice (BMP) to remedy hydrology and water quality impairment from urban development, bioretention facilities need rigorous investigation to quantify performance benefits and to allow design improvements. This study examines water quality improvements [total arsenic, total cadmium, chloride, total chromium, total and dissolved copper, E. coli, fecal coliform, lead, mercury, nitrogen species, oil and grease, phosphorus, total organic carbon (TOC), total suspended solids, and total zinc] via monitoring for a 15-month period at two bioretention cells in Maryland. Both bioretention cells effectively removed suspended solids, lead, and zinc from runoff through concentration reduction. Runoff volume reduction promotes pollutant mass removal and links BMP water quality benefits with hydrologic performance. From a load perspective (kg/ha?year), all but TOC at one cell showed pollutant reduction. Bioretention effluents exhibited good water quality for all significant pollutants except for nitrate, copper, and phosphorus in one cell, the latter two of which may be attributed to media organic matter dissolution. Copper dissolved/particulate analyses showed that significant changes in copper speciation behavior result from transport through the bioretention media.     

Trace Metal Pollutant Load in Urban Runoff from a Southern California Watershed

In order to implement efficient and effective management strategies for coastal water quality in Southern California, it is important to consider the relative pollutant contributions from urban dry-weather flow (DWF) and wet-weather flow (WWF). This study uses both historical flow coupled with water quality monitoring data and computer modeling to characterize the annual DWF and WWF discharges from an urban catchment in Los Angeles, Calif. The DWF and WWF pollutant loading of the trace metals copper, lead, nickel, and chromium for 6 water years dating from 1991 to 1996 is predicted. The results indicate that DWF contributes a considerable amount of flow and pollutants. Approximately, 9–25% of the total annual Ballona Creek flow volume is DWF. The simulations indicate DWF accounts for 54, 19, 33, and 44% of the average annual load of total chromium, copper, lead, and nickel, respectively. In the dry season, the simulations indicate DWF accounts for 89, 59, 58, and 90% of the load of total chromium, copper, lead, and nickel, respectively. This research suggests DWF controls may be an important part of pollution mitigation plans for urban stormwater drainage systems in Southern California.     

Characterizing Physicochemical Quality of Storm-Water Runoff from an Urban Area in Calgary, Alberta

Understanding storm-water runoff quality is required to develop effective urban storm-water runoff management for regions of semiarid climate. In this study, the quality of storm-water runoff from a semiarid, urban residential catchment, draining through separated storm-water sewers was investigated in 2006 and 2007. Water temperature, conductivity, pH, dissolved oxygen, and turbidity were continuously measured during 16 storm events. Storm-water runoff quality was characterized in terms of event mean values (EMVs), loads, and first flush (FF) loads and their relationships with rainfall characteristics. Discharge of total suspended solids (TSSs) is in general governed by the flow magnitude in storms and no significant relationships exist between the FF loads of TSS and rainfall intensity. The discharge of dissolved solids is independent of the flow magnitude. Strong FF effect for dissolved solids and weak FF effect for TSS were observed. This semiarid region provided no relationship between the EMVs of both TSS and conductivity and the antecedent dry period. This raises doubts on storm-water runoff being more heavily loaded with pollutants after a longer dry period in semiarid regions.     

Documenting Stormwater Quality on Texas Highways and Adjacent Vegetated Roadsides

The primary objective of this study is the documentation of stormwater quality of vegetated roadsides of two Texas highways (State Highway 6 in College Station and Loop 360 in Austin), both of which had high average daily traffic. Three sites each in Austin and College Station were monitored using passive “first flush” stormwater samplers for 16 months. Results from this study indicate that significant removal of sediment and heavy metals occurred over the width of vegetated roadsides, but no apparent nutrient (nitrogen and phosphorus) removal was observed. The results also show that vegetation density has a direct effect on the performance of vegetated roadsides. When roadsides are densely covered with grasses above 90%, significant sediment removal is expected, often within the first 4?m of the edge of pavement. A stepwise regression analysis identifies the antecedent dry period (ADP) as the most significant predictor to pollutant concentration. The pollutant event mean concentration was found to decrease with increasing ADP for all pollutants at the College Station sites, but not the Austin ones.     

Particle Destabilization in Highway Runoff to Optimize Pollutant Removal

Sedimentation column studies and simulations using particle size distribution suggest that low removal efficiencies of smaller particles in highway runoff would be obtained using sedimentation if coagulation-flocculation is not performed. Coagulation-flocculation studies, using metal salts (alum and ferric chloride) and one organic polymer in three molecular weights, were evaluated over the 2004–2005 storm seasons. Only the first flush or approximately the first hour of runoff was coagulated. Efficiencies were quantified with particle size distribution measurements and turbidity. Results with low dosages of metal salts were ineffective and did not improve water quality. High dosages of metal salts using a sweep floc mechanism were effective in dramatically lowering runoff turbidity, but resulted in large quantities of sludge production and required pH control. A cationic organic polymer at low dosages (<10?mg/L) was effective in coagulating highway runoff and reducing particle charge. Extended mixing time was required to achieve low turbidities ( ～ 5 NTU). A combination of organic polymer, followed by small doses of alum (<10?mg/L), reduced mixing time and produced high quality effluent.     

Oil and Grease Measurement in Highway Runoff—Sampling Time and Event Mean Concentrations

An event mean concentration (EMC), usually collected with an automatic, flow-weighted composite sampler, is often used to characterize stormwater pollutants. Automatic samplers are not recommended for collecting oil and grease (O&G) samples due to possible biases associated with interactions with tubing and pumps. To measure the EMC without sampler interferences, a series of grab samples (often over ten samples) must be collected along with the flow measurement to compute the EMC. This paper examines 22 O&G pollutographs from small, impervious highway sites, to determine when a single O&G grab sample most closely approximates a flow-weighted composite sample. Samples collected within the first hour of a storm event overestimated the O&G EMC by 20?mg/L or more, while samples collected toward the end of the event underestimated the EMC. The best time to collect a single grab sample ranged from 1 to 6 h after the beginning of runoff, and was related to site or storm-specific factors. Results obtained from this study also showed that strong correlations (R2 = 0.9) exist between O&G and other organic constituents, such as chemical oxygen demand (COD) and dissolved organic carbon (DOC). Correlations also exist between O&G EMC, antecedent dry days, and total rainfall. Depending upon site and regulatory specific factors, using COD or DOC EMCs in lieu O&G samples may be a better strategy.     

Steady Groundwater Transport of Highway Deicing Agent Constituents from an Infiltration Basin

The highway deicing agent groundwater plume from an infiltration basin in the Plymouth-Carver Aquifer near State Route 25 in southeastern Massachusetts is modeled and measured in order to assess the impact of the basin on groundwater transport processes. The advective transport model superimposes the existing steady models of axisymmetric basin hydraulics of Zlotnik and Ledder in 1992 and the two-dimensional ambient flow of Gelhar and Wilson in 1974. The basin component incorporates a surface source of finite-radius into a Hankel transform model for unconfined aquifers, whereas the ambient component varies linearly in the horizontal and vertical directions. Contaminant streamlines describe the resulting groundwater plume, and highway deicing agent constituents provide useful tracers. A simple vertical dispersion model quantifies the spread of contaminants across the bottom of the plume. Deicing agent constituent data calibrate a 17?m basin radius, a 45?m plume width, a bottom streamline 6–10?m below the water table, and a vertical dispersivity of 34?cm. The latter value is comparable to the amplitude of vertical excursions caused by storm scale fluctuations of hydraulic head, as measured by Ostendorf et al. in 2007. Infiltration basins alter ambient advection by displacing streamlines downward and augment vertical mixing by imposing aperiodic vertical fluctuations on the ambient flow field.     

Utility of Suspended Solid Measurements for Storm-Water Runoff Treatment

In this paper alternative solutions are presented to solve problems associated with the measurement of total suspended solids (TSS) in storm-water runoff. Results revealed that the accuracy of TSS measurement is largely related to sample representativeness, particle size distribution (PSD), sampling pipette position, and sample mixing. In general, when the PSD in the runoff was mostly larger than 75?μm, the most accurate and reproducible results were obtained when samples were collected from a position of mid-depth and midway between the walls of the beaker and the vortex and mixed at speeds in the range of 600–700?rpm. For runoff samples with a PSD smaller than 75?μm, mixing at a higher rpm is not a significant factor. As long as the PSD in the TSS subsample is representative of the original sample, a strong correlation between TSS and suspended solid concentration can be achieved. The results showed that density was largely correlated with the organic content of the particles, and, in general, smaller particles tended to have a lower density. The density results revealed that assuming a single sand size density of 2.6?g/cm3 for storm-water runoff produced a large error in the computation of sediment load and particle settling velocity.     

On Parameter Estimation of Urban Storm-Water Runoff Model

An existing accumulation and wash-off model was applied and calibrated on a standard asphalt parking lot located in the northeastern United States. The field measured data consisted of rainfall, flow, and runoff samples taken from over 26 storm events monitored from 2004 to 2006. The contaminants under consideration include: total suspended solids, total petroleum hydrocarbons-diesel range hydrocarbons (TPH-D), dissolved inorganic nitrogen (DIN) (comprised of nitrate, nitrite, and ammonia), and zinc (Zn). The objective of the study was to provide probability distributions of model parameters for contaminants that have not been documented much (TPH-D, DIN, and Zn). The best fitting parameter values were found on a storm by storm basis. Subsequently, the range and variability of these parameters are provided for modeling purposes and other urban storm-water quality applications. A normal distribution was fitted to the optimized model parameter values to describe their distributions. A simulated annealing algorithm was used as the parameter optimization technique. Several examples are given to illustrate the methodology and the performance of the model. Finally, a Monte Carlo simulation was performed to assess the capability of the model to predict contaminant concentrations at the watershed’s outlet.     

Prediction of Nutrient Concentrations in Urban Storm Water

Excessive quantities of nutrients in urban storm-water runoff can lead to problems such as eutrophication in receiving water bodies. Accurate process based models are difficult to construct due to the vast array of complex phenomena affecting nutrient concentrations. Furthermore, it is often impossible to successfully apply process based models to catchments with limited or no sampling. This has created the need for simple models capable of predicting nutrient concentrations at unmonitored catchments. In this study, simple statistical models were constructed to predict six different types of nutrients present in urban storm-water runoff: ammonia (NH3), nitrogen oxides (NOx), total Kjeldahl nitrogen, total nitrogen, dissolved phosphorus, and total phosphorus. Models were constructed using data from the United States, collected as a part of the Nationwide Urban Stormwater Program more than two decades ago. Comparison between the models revealed that regression models were generally more applicable than the simple estimates of mean concentration from homogeneous subsets, separated based upon land use or the metropolitan area. Regression models were generally more accurate and provided valuable insight into the most important processes influencing nutrient concentrations in urban storm-water runoff.