UPDATES: November 2011

UPDATES is a monthly email newsletter on stormwater management, assessment (including monitoring), and maintenance research at St. Anthony Falls Laboratory and the University of Minnesota.

Click HERE to subscribe to UPDATES Newsletter

Or Text SWUPDATES to 22828 to Join. Message and data rates may apply.

For Email Marketing you can trust


Capture and Release of Pollutants by Rain Gardens

November 2011 (volume 6 - issue 10)

Contributed by Joel G. Morgan, John S. Gulliver, and Raymond M. Hozalski (Department of Civil Engineering, University of Minnesota)


Bioretention practices, or rain gardens, are a seemingly ubiquitous stormwater management practice, developed for the infiltration of stormwater runoff. Using the same processes that describe treatment of drinking or wastewater, one can also design rain gardens to remove stormwater pollutants. A clear understanding of the retention of pollutants on the media of rain gardens may lead to strategic implementation for TMDL load reductions.

Toxic metals are a specific stormwater pollutant of concern that contributes to the impairment of ecosystems. These metals, which include cadmium, copper, lead, and zinc, increase lesion rates, limit reproduction by delaying or inhibiting hatching, and cause negative developmental effects on freshwater organisms. The primary source of cadmium, copper, and zinc in stormwater runoff is vehicle deterioration, such as brake pads and tires. We investigated materials commonly used in rain gardens like C-33 sand and compost for their ability to remove dissolved toxic metals from solution.

Phosphorus retention on the media is also an objective of rain gardens, especially for those with under-drains that discharge into surface waters. This UPDATES will report on experiments performed to investigate the ability of rain garden media to remove and retain dissolved toxic metals under controlled conditions designed to mimic representative stormwater qualities and loading rates. This study will also investigate dissolved phosphorus release from the media under similar, controlled conditions. Lastly, a recommendation for one possible rain garden design will be provided to aid with long-term maintenance.

Experiment and Results

Vertical columns were designed from 2 inch diameter clear PVC and packed with different ratios of compost mixed in sand (50, 30, 10, 0% compost, by volume). Intermediate sampling ports were installed after every 2 inches of filter media using ¼ inch perforated tubing. A 1-inch layer of sand served to protect the sampling tube from clogging by small particles (Figure 1).

Figure 1. Diagram of column experiment.

Synthetic stormwater, containing 1 mg/L each of Cd2+, Cu2+, and Zn2+, was pumped from an adjacent 120-Liter influent reservoir up through the columns from bottom to top to ensure saturation of the filter media. The mean velocity through the columns was adjusted daily to a target value of 12 cm/hr using clamps attached to the influent and effluent tube lines because head loss increased as the experiment progressed.

The retention on the media of cadmium, copper, and zinc were examined at three different depths and for different ratios of compost. The retention on the media of cadmium at 6 inches of filter media depth is presented in Figure 2. The experiment also showed that the retention on the media of cadmium, copper, and zinc increases with filter media depth.

Figure 2. Retention of cadmium on the media by 30% compost through 6 inches of filter media. Data from three replicate columns (D, E, F) are presented with a model fit to the data.

The compost fraction within the filter media affects the retention of metals in the columns. Figure 3 shows the a model fit to the breakthrough data for different compost-sand compositions collected at 6 inches of filter media.

Figure 3. Model fits for cadmium (top) and zinc (bottom) at 6 inches of filter media depth indicating the effect of compost fraction on retention.

The result is that the 100% sand column approaches saturation soonest, followed by the 10%, 30%, and lastly, the 50% compost columns. Time to breakthrough increases as compost fraction increases indicating that sorption capacity is related to compost fraction. Copper retention on the media also increased with increasing compost fraction. Copper was not entirely in the soluble form at the stormwater pH, so sorption and precipitation are both retention mechanisms. The precipitated copper was removed via filtration. Results from a post-experimental media analysis indicate 62% of copper was removed in the first 2 inches of filter media whereas only 33% of the cadmium and zinc was found in the first 2 inches, which is consistent with the filtration of copper.

The column study results indicate that phosphorus leaches from the media at close to a constant rate. The columns released average concentrations of 0.29, 0.29 and 0.21 mg/L from the 50, 30, and 10% compost columns, respectively, at 15 cm of depth (Figure 4). Maestre and Pitt (2005) report median runoff concentrations for total and dissolved phosphorus of 0.27 and 0.12 mg/L, respectively, so the concentrations are substantial.

Figure 4. Cumulative P released normalized by the mass of filter media.

It is possible that the leaching rate will remain constant over time because organic compost will continually decompose and release phosphorus. Phosphorus uptake is expected in live plants; however, upon death, phosphorus will be released back into the media. Lastly, refreshing the top mulch or compost layer will add a phosphorus source to the rain garden. The column studies indicate that phosphorus leaches off the compost and leaves the system. If a rain garden outlets the infiltrated water to groundwater, then the leaching of phosphorus may not be a concern (Weiss et al., 2008). If, however, the infiltration practice is built with an under-drain or is located near a lake or river where export of phosphorus to surface waters is possible, then degradation of the water bodies due to phosphorus leaching may occur.


Understanding retention on the media of toxic metals by compost and sand illuminates the fate of toxic metals in rain gardens. Determining the longevity of a rain garden for retention of metals will aid in Total Maximum Daily Load (TMDL) planning and implementation, as well as maintenance. Typical hydrologic and rain garden characteristics can be used to compare the results from the column study to a full-size rain garden and to show the life span of rain gardens that use the same filter media as this experiment.

Using the model, a filter media composed of 30% compost with an influent concentration of 0.1 mg/L will achieve 10% breakthrough in 95 years for cadmium and 76 years for zinc at a rain garden depth of 15 cm. Hydraulic failure due to clogging of the pore space by suspended solids is expected to occur before breakthrough or exhaustion of the filter media by toxic metals.

Based on Figure 4, the 30% compost columns have released approximately 0.06 mg P per g of filter media over the course of the experiments (approximately 1,700 pore volumes, or 25 years of typical runoff to a rain garden). If we assume that the rain garden is 15 cm deep, has a surface area of 500 m2, and the filter media has a bulk density of 1,200 kg/m3 then the yearly load is 216 mg of phosphorus released from the filter media. The column studies indicate that this could be a consistent release rate over time. Net release of phosphorus from rain gardens means that they must either be designed or not used for retention on the media of dissolved phosphorus. The retention on the media of dissolved phosphorus may not be a goal of rain gardens that infiltrate stormwater to the groundwater table.

Rain Garden Recommendation

We need to consider the rain garden's ability, or lack thereof, to remove the whole spectrum of pollutants commonly found in stormwater runoff. Compost amended sand is an excellent filter media for the retention of petroleum based products (LeFevre et al., 2011) and toxic metals on the media, yet leaches phosphorus, a major limiting nutrient in most inland waters. Another filter media will be needed to remove the phosphorus or other pollutants that are not removed by the compost/sand filter. If phosphorus retention on the media is a goal, then rain gardens could be constructed as a two-stage filter. The top layer would be constructed of sand amended with compost of approximately ½ to 1 ft deep, depending on the necessary root depth for plants. Underneath the compost layer would be an iron-enhanced sand layer (Erickson et al., 2007) approximately ½ to 1 ft deep to remove phosphorus that passes through and is leached off the compost layer (Figure 5). A rain garden constructed in this manner would be designed to remove dissolved toxic metals and hydrocarbons in the compost amended sand layer at the top and phosphorus with the iron enhanced sand layer in the middle.

Figure 5. Proposed construction of rain gardens with under-drains.

A sand infiltration basin was constructed in Maplewood, MN using iron filings as an amendment to remove phosphorus from stormwater runoff. At the time of writing, the authors were only aware of planned or partially constructed (Carver County and Maplewood, MN) two-stage rain gardens as described herein.

A previous UPDATES article was published in November 2010 on preliminary results of this research and was presented at the 2010 Minnesota Water Resources Conference and at the 2010 StormCon conference. The StormCon proceedings paper, “When Do We Need to Replace Bioretention Media?” is available upon request. A comprehensive discussion of the experiments and results described herein can be found in the project report delivered to the MPCA and in Joel Morgan’s thesis, both available upon request from the St. Anthony Falls Laboratory.


  • Erickson, A.J., Gulliver, J.S., Weiss, P.T. (2007). “Enhanced sand filtration for storm water phosphorus removal.” Journal of Environmental Engineering, 133:5, 485-497.
  • LeFevre, G., Hozalski, R., Novak, P. (2011) The role of biodegradation in limiting the accumulation of petroleum hydrocarbons in raingarden soils. (In Review Water Research 2011).
  • Maestre, A. and Pitt, R. (2005). The National Stormwater Quality Database, Version 1.1 A Compilation and Analysis of NPDES Stormwater Monitoring Information. U.S. EPA Office of Water. Washington, DC.
  • Weiss, P., LeFevre, G., and Gulliver, J. (2008). Contamination of Soil and Groundwater Due to Stormwater Infiltration Practices. SAFL Project Report No. 515. http://www.pca.state.mn.us/index.php/view-document.html?gid=7732


We want to hear from you!!!

Let us know your thoughts, experiences, and questions by posting a comment. To get you thinking, here are a few questions: 

  • How do you envision long-term maintenance of a tiered or layered bioretention cell?
  • What are potential benefits or negative consequences of this proposed geometry?
  • How could we expand upon this structure to efficiently incorporate a wider variety of pollutant reduction?

Citation: "Stormwater Research at St. Anthony Falls Laboratory." University of Minnesota, St. Anthony Falls Laboratory. Minneapolis, MN. http://stormwater.safl.umn.edu/