Evaluating Urban Air Deposition on an Industrial Facility in Seattle
To better understand the contribution of metals in urban air deposition into stormwater, a study was conducted at Terminal 18, a 200-acre container terminal located on Harbor Island in Seattle, WA. Although the National Pollutant Discharge Elimination System (NPDES) Industrial Stormwater General Permit (ISGP) does not make allowances in consideration of contribution from air deposition into stormwater, understanding the contribution of copper and zinc may assist in the selection of appropriate operational source control and stormwater treatment best management practices (BMPs).
Passive air deposition samplers were placed at two locations at the terminal. One was placed at the south end on the roof of the main entry gate—which is close to significant truck traffic, including an elevated freeway. The other was located at the north end of the terminal, on the roof a three-story administrative building. A flux could be calculated by use of the sample metal concentration, deployment duration, and funnel size. The flux is the mass of a metal that is deposited over a defined area for a defined period of time (micrograms per square meter per day [µg/m2-day]).
The study began in June 2014, and samples were collected every two to four weeks. Seventeen samples were collected and analyzed for copper, zinc, total suspended solids (TSS), and particulate size. The key findings include:
- The average copper air deposition flux rate at the terminal was 31 µg/m2-day, and the average zinc air deposition flux rate at the terminal was 174 µg/m2-day.
- Weather patterns did not affect the sample results.
- The average zinc flux rate at the southern sample location (Gate 1), 180 µg/m2-day, was slightly greater than the average zinc flux rate at the northern sample location (the administration building), 168 µg/m2-day.
- The average copper flux rate at the southern sample location, 29 µg/m2-day, was similar to the average copper flux rate at the northern sample location, 31 µg/m2-day.
- The proximity to the higher traffic areas and the elevated freeways was not as significant of an impact as anticipated.
- Based on the calculated flux data, an estimate of stormwater concentrations for copper or zinc over a given period of time that would result solely from air deposition indicated that regulatory benchmarks for the terminal (14 micrograms per liter [µg/L] and 117 µg/L for copper and zinc, respectively) will be exceeded for copper (24 µg/L) and zinc (137 µg/L) after a two-day period of no rain, followed by a minimally qualifying rain event (i.e., 0.1 inch of rainfall occurring after a two-day period of dry weather).
Although these calculations assume that all particulates are discharged into the stormwater and do not factor into the deposition within stormwater conveyance structures (e.g., catch basins), they do indicate that facilities that do not implement operational source control BMPs (e.g., sweeping, catch basin cleaning) would likely exceed the NPDES ISGP benchmarks without any industrial contributions of their own. This study shows the importance of these BMPs and also highlights the difficulties for industrial facilities to meet benchmarks in an urban area.
King County performed an air deposition study in 2013 to characterize the quantity and nature of nonpoint-source contamination that was being deposited in the form of dust onto the surfaces of various land-use types throughout the county. The Lower Duwamish Waterway Source Control: Bulk Atmospheric Deposition Study Final-Draft Report was published by King County Department of Natural Resources and Parks, Water and Land Resource Division, in December 2013. This study relied upon passive sampling methods to quantify contributions from both wet and dry deposition; that is, a sampling apparatus collected dry particulate matter that settled out of the atmosphere, and wet particulate matter that settled out due to rain. After a finite amount of time, or when the sampling jar was full of water, the sample was collected for analysis. After performing approximately one year of sampling, the results were analyzed and net fluxes (in µg/m2-day) were calculated. Flux is a flow rate—in this case, air deposition in units of µg per day per unit area in meters. These fluxes represent the contaminant mass loading that can be expected in the vicinity of each of the sampling locations. Sampling locations were chosen to be representative of broader geographic areas so that the results could be extrapolated to nearby properties and areas with similar land-use characteristics.
The three locations included in the study that have the most similarity to Terminal 18 are: the Duwamish Valley Station, the South Park Station, and the Beacon Hill station. These locations are shown in Figure 1, and sampling results are presented in Table 1. The Duwamish Valley Station represents industrial and urban land uses, while the South Park Station represents a mix of suburban, industrial, and residential land uses. These two sites are positioned relatively centrally in the Lower Duwamish Waterway corridor, but are spaced approximately half a mile apart and on opposite sides of the Duwamish Waterway. The Beacon Hill Station is located at the top of a hill near Terminal 18, and represents urban residential land use.
In 2010, The Boeing Company conducted air deposition sampling to quantify the amounts of copper and zinc being deposited on its Plant 2 site from nonpoint-source atmospheric deposition. The Boeing Plant 2 sampling location is shown in Figure 1. The Boeing Plant 2 air deposition data are included in the average flux calculation presented in Table 1.
Potential Loading to Terminal 18
The King County air deposition report quantified average fluxes for each of the stations included in the 2013 study in units of µg/m2-day. If these fluxes are multiplied by the surface area of an area of interest or stormwater basin, the flux can be converted into an amount of copper or zinc deposited per day over a particular area. Because Terminal 18 is an industrial site in the Duwamish River Valley of Seattle, the three stations within the valley (South Park, Duwamish Valley, and Boeing Plant 2, which comprise industrial and mixed-use or industrial land uses) were considered to be most appropriate (see Figure 2 for Terminal 18 location). The average wet and dry deposition fluxes for these three sites for each of the two metals of interest (total copper and zinc) were averaged together to determine an estimated flux for Terminal 18. This average flux for copper is 21 µg/m2-day and the average flux for zinc is 117 µg/m2-day.
Next, the area of each stormwater drainage basin at Terminal 18 was calculated in units of square meters (m2). Stormwater drainage basins were determined using the drainage basin boundaries for the terminal. For those drainage basins that drain stormwater from offsite facilities, the terminal facility boundary was used to delineate the extent of the drainage basin. It was assumed that each basin was flat and that deposition to the top surfaces of containers present onsite would enter the stormwater system. Finally, the total surface area of the site was calculated by summing the area of each of the drainage basins. Results of this calculation, and an estimated 1-acre subbasin, are presented in Table 2.
1 This represents the mass (concentration) of metal particulate that is expected to fall on the surface of the site following each sweeping event.
2 This represents the concentration that would be present in stormwater across the site due to air deposition after a minimally qualifying rain event (0.1 inch of rainfalloccurring after a three-day period of dry weather), assuming that sweeping happened immediately prior to the two-day dry period.
3 This represents the concentration that would be present in stormwater across the site due to air deposition after a minimally qualifying rain event (0.1 inch of rainfall occurring after a week-long period of dry weather), assuming that the rain event was preceded by a week of dry weather and that sweeping happened immediately
prior to the week of dry weather.
The potential amount of copper and zinc that could be entrained in stormwater was calculated based on the air deposition fluxes developed. This was calculated using the following assumptions:
- Total mass of copper and zinc from air deposition present on the surface of the terminal would be flushed into the stormwater conveyance system during the modeled storm event, and would be uniformly mixed. There are no allowances for the metals to be trapped in the catch basins.
- Rainfall during a 24-hour period was 0.1 inch, the minimum for a qualifying storm event for 2008 NPDES permit sampling. Greater amounts of rainfall would serve to dilute the measured concentrations of copper and zinc. Therefore, the 0.1-inch-modeled storm is a worst-case scenario for the three-day or seven-day deposition period.
- Copper and zinc accumulate onsite between rainfall events, which occur either three days or one week apart. These durations were selected to recognize both the frequency of sweeping occurring onsite and the frequency of rainfall events during the wet season.
Based on these assumptions, concentrations of total copper are expected to vary between 25 µg/L (if three days elapse between rainfall events) and 58 µg/L (if one week elapses between rainfall events). Similarly, total zinc concentrations are expected to vary between 138 and 322 µg/L for three-day and one-week scenarios, respectively. These concentrations would exceed NPDES requirements for total copper and total zinc, which are 14 µg/L and 117 µg/L, respectively.
These calculations indicate that dry and wet air deposition of copper and zinc potentially result in substantial contributions to stormwater. It was determined that the copper and zinc fluxes in the vicinity of Terminal 18 could be better defined if site-specific air deposition data were collected. Based on this, an air deposition Sampling and Analysis Plan (SAP) was prepared to collect site-specific samples to further evaluate the actual pollutant loading occurring at the terminal.
Air Deposition Sampling and Analysis Plan at Terminal 18
The air deposition SAP was modeled after the air deposition study that was recently conducted by King County. Collecting air deposition data can be challenging because of variability in sampling equipment, variable weather patterns—including wind and humidity—and other factors. Consistency in methodology between the Terminal 18 air deposition SAP and the broader air deposition study conducted throughout the region allowed comparison of these results collected under this SAP to the broader dataset. This helped characterize whether typical air deposition at Terminal 18 is greater than or less than the observed air deposition measured at other industrial sites in the Duwamish River Valley.
Collection of Field Samples
Sampling Locations. Sampling locations were selected based on a variety of factors, including the following:
- Accessibility. Potential sampling locations on rooftops and other stationary structures were prioritized, as these sampling locations would be out of the lanes of traffic and were expected to be consistently accessible for sampling.
- Representativeness. Sampling locations at the interior of the site were expected to be representative of onsite and nonpoint-source offsite air deposition contributions. Potential sampling locations at the western edge of the site were expected to be representative of onsite and nonpoint-source offsite air deposition contributions, and may also be representative of offsite air deposition from operations at neighboring facilities. As such, it was desirable to have two sampling locations: one from a structure at the interior of the site, and one from a structure near the site boundary.
After evaluating the potential sampling locations, Floyd|Snider determined that the optimal sampling locations are on the top of the roof of the administration building and on the roof of the Gate 1 Entry Building (Figure 3). By sampling at two locations at Terminal 18, the representativeness of the data collected can be compared to the other data collected at the terminal and to the air deposition data collected elsewhere in the Duwamish River Valley.
In contrast to the King County air deposition study, the proposed list of analytes that were analyzed in Floyd|Snider’s evaluation differed. The Floyd|Snider SAP included total copper, total zinc, TSS, and particulate size because of the NPDES permit stormwater sampling requirements. In the 2013 King County air deposition study, copper and zinc were the only two analytes that were deposited in a significant amount; other analytes (e.g., mercury, polycyclic aromatic hydrocarbons, polychlorinated biphenyls) had much lower flux rates.
Sampling Devices. Samples were collected by a collection funnel that drained directly into the laboratory-provided 2-liter sample bottles. The funnel was supported by a wood-framed structure. One such sampling system was constructed at each of the two sampling locations. A third funnel was used to collect field duplicate samples. See Figure 4 for an example sampling system.
Each sampling system was constructed with the following components:
- A pre-cleaned 6-inch-diameter funnel with a side vent
- A pre-cleaned fluoropolymer sample jar that accepts a screw cap
- Tubing to connect the base of the funnel to the sample jar, as needed
Sampling materials were tested for metals content prior to use in the field. Only sampling materials that did not contain measurable amounts of copper and zinc were employed at the site. Glass, high-density polyethylene, Teflon, or stainless steel equipment were used for sample collection wherever possible.
Sampling Procedures. Samples were collected consecutively and approximately monthly throughout the one-year duration of the sampling. Shorter collection periods of one to two weeks occurred during the wet season (October 1 through May 31) in order to prevent sample jar overflow. Once a minimum volume of water was reached, the sample jar was removed for laboratory analysis, the sampling apparatus was cleaned, and a new sampling jar was positioned to begin collection of the next sample. During the wet season, sampling periods included at least one rain event.
Longer collection periods occurred during the dry season (June 1 through September 30), maximizing atmospheric deposition. There is no minimum rainfall requirement for a successful sample in the dry season, as it may be composed of particulate alone.
At the time of sample retrieval, the amount of rainfall in the jar was quantified by comparing the jar’s mass at the end of the collection period to its mass prior to sample collection. At the time of active sample collection, a known quantity of reverse osmosis (RO) water (up to 400 milliliters [mL]) was used to rinse dry particulates into the sample jar (rinse volume). The rinse volume and funnel area were recorded on field forms. The laboratory determined the deposition volume. During the dry season, when little or no water may have been present, volume was measured after the known volume of RO water was added. During the wet season, volume was measured prior to addition of the RO water.
Sampling Results
The 17 rounds of air deposition samples collected at Terminal 18 between July 2, 2014, and June 15, 2015 consisted of material that had accumulated in the sampling devices at the administration building and Gate 1 Entry Building for periods between 12 and 87 days, with an average of 27 days. The air deposition samples were analyzed for copper, zinc, TSS, and particulate size. The results reported by the laboratory were converted from a concentration (in µg/L) to a flux (in µg/m2-day), or the amount of deposition of metals in a unit area. The flux were used to estimate loading of copper and zinc to a given stormwater conveyance basin. The data are presented in Table 3.
The flux values observed at Terminal 18 during the sampling period were typically greater than the reported flux values for the South Park Station, Duwamish Valley Station, and Boeing Plant 2 locations (see Table 1 and Figures 5 and 6).
The air deposition samples were also analyzed for particulate size so that the data could be compared to the particulate size data from the stormwater samples. A comparison of the datasets indicated that there was a larger concentration of larger particles (250 to greater than 500 micrometers [µm]) in the air deposition samples, but similar concentrations of the smaller sized particles (less than 1 to 62.5 µm) between the air deposition samples and the stormwater samples. Generally, there were no detections of the mid-sized particles (62.5 to 250 µm) in either the air deposition samples or the stormwater samples. In general, the air deposition samples had similar percentages, roughly 25–35%, for particles in the 3.9–62.5-µm range and the particles smaller than 1 µm in size. The samples collected at the administration building had a higher percentage, 22.1%, of the largest particles (greater than 500 µm) than the Gate 1 Entry Building, 9.6%. With the exception of the absence of the mid-size particles (62.5 to 250 µm), the results had significant variability throughout the sampling period.
Conclusion
The data collected at Terminal 18 indicate that copper and zinc are being deposited at concentrations that substantially affect stormwater quality relative to meeting the NPDES ISGP benchmarks. Both the site-specific average copper and zinc flux values were greater than the estimated values from the nearest representative sampling locations (30 µg/m2-day versus 21 µg/m2-day for copper, and 174 µg/m2-day versus 117 µg/m2-day for zinc).
Based on the data, the permit benchmarks for total copper and total zinc—14 µg/L and 117 µg/L, respectively—would likely not be met after only two days of air deposition if there was a two-day period with no rain and no BMPs (i.e., sweeping) employed. Concentrations in stormwater from the site of total copper are expected to vary between 24 µg/L (if 2 days elapse between rainfall events) and 83 µg/L (if 1 week elapses between rainfall events). Similarly, total zinc concentrations are expected to vary between 137 and 480 µg/L for three-day and one-week scenarios, respectively.
Note that the scenarios are considered worst-case. In reality, a portion of the deposited metals is addressed by BMPs or captured in the subsurface stormwater conveyance catch basins. Additionally, a larger storm may result in lower concentrations. The scenarios do clearly illustrate the difficulty in meeting the permit benchmarks. If Terminal 18 were vacated, for example, the permit benchmarks would likely not be met if BMPs such as sweeping were not implemented, solely due to air deposition of copper and zinc.