UPDATES: March 2013

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


Quantifying Nutrient Removal through Targeted Intensive Street Sweeping

March 2013 (volume 8 - issue 3)

Contributed by Paula Kalinosky (graduate research assistant) and Lawrence A. Baker, Bioproducts and Biosystems Engineering University of Minnesota; Sarah Hobbie, Department of Ecology, Evolution, and Behavior, University of Minnesota; and Ross Bintner, City of Edina, Minnesota.

Funded by the U.S. Environmental Protection Agency through the Minnesota Pollution Control Agency.

Urban watersheds face unique concerns, not the least of which is nutrient management, and it is quite possible we have been underestimating a key piece in the puzzle of how to improve urban water quality. Coarse organic debris (leaves, grass clippings) that finds its way onto streets can leach nutrients into stormwater runoff, and eventually make its way into storm sewers, unless removed by sweeping. Once in storm sewers, this material can accumulate in catch basins and pipes, or be transported into streams, lakes, and rivers, releasing nutrients along the way as it decomposes. The first step in managing this source of nutrients to stormwater is to quantify the magnitude of the problem.

In 2010, the City of Prior Lake partnered with the University of Minnesota in a street sweeping study to address nutrient management. The main objectives of the study were to measure the total amount of sediment and associated nutrients removed by street sweeping and to quantify the influence of overhead tree canopy on the character and quantity of sediments found on the street. Over a two-year period beginning in August, 2010, street sweeping operations were conducted in nine study routes classified as having “high”, “medium”, or “low” tree canopy cover. Sweeping operations were conducted over the entire snow-free period allowing us to capture seasonal influences on loading patterns. The study design, which also includes a comparison of different street sweeping frequencies, is summarized in Figure 1.

Figure 1: Example of low, medium, and high canopy sweeping routes, Prior Lake, MN.

Over the course of the two-year study, we tracked the total amount of material collected by the sweeper and sampled sweeper waste for 392 sweeping operations. Coarse organic material (leaves, grass clippings, seeds, flowers, etc.) was separated from finer, soil-like material by first sieving the sample (2mm sieve) and then floating the material retained on the sieve in a water bath to separate coarse organics (which floated) from pebbles and adhered soil particles (which sank). Chemical analysis (total phosphorus, TP, total nitrogen, TN, total carbon, C, % moisture, and % organic matter, %OM) was carried out on each fraction. Water used in the float separation step was also analyzed for dissolved nutrients.

Coarse organic material made up 15% of the total dry weight of swept material collected during the study, but 36% of the TP and 71% of the TN. As expected, the influence of coarse organics on material loads followed a seasonal pattern (Figure 2). During the fall (Sept – Oct), coarse organics accounted for 36% of solids, 60% of TP, and 80% of TN.

Figure 2: Coarse organic material as a proportion of total sweeping load (monthly average values, high canopy sweeping routes), Prior Lake, MN.

The dry solids and nutrient loads generally increase with tree canopy cover and sweeping frequency (Table 1). Material loading patterns indicate a build-up rate of roughly 10 days, such that sweeping at higher frequencies increased the total amount of material recovered, but at diminishing returns on a per sweeping event-basis. Total solids collected increases with tree canopy at any given sweeping frequency. The one inconsistency in this pattern, low and medium canopy routes swept four times per month, is not seen when looking at the coarse organic material only (not shown).

Table 1: Dry material and phosphorus removed during the snow-free period as a function of canopy cover and sweeping frequency (annual average), Prior Lake, MN.

Frequency Low Canopy Med Canopy High Canopy
Annual Dry Material Load (lb/curb-mile)
Once per month 1825 2179 4122
Twice per month 2894 4227 5815
4 per month 5141 7292 7205
Annual Phosphorus Load (lb/curb-mile)
Once per month 1.4 1.5 2.6
Twice per month 1.9 3.5 5.0
4 per month 3.0 5.6 6.3

To determine whether sweeping loads could be estimated from measures of canopy cover, we used remotely sensed tree canopy cover along with road polygon data provided by the City of Prior Lake to quantify the percent tree canopy covering the street. Tree cover was determined by the University of Vermont Spatial Analysis Laboratory using object-based image analysis that combines satellite imagery and LiDAR data to develop fine-scale land cover maps. A plot of P removal (lb P/curb mile) vs. % canopy cover reveals a linear relationships between P removal and % canopy (figure 3).

Figure 3: Correlation between phosphorus loads and overhead tree canopy, sweeping routes grouped by sweeping frequency, Prior Lake, MN.

We also tracked the cost-efficiency of sweeping for nutrient removal throughout the study. Our calculations accounts for driver wages and benefits, vehicle fuel, and vehicle wear and tear (provided by the City of Prior Lake). Sweeping was most cost effective in the spring and fall (Figure 4) when targeted sweeping operations achieved costs as low as $18/lb P removed, and least efficient during mid-summer and mid-winter, when costs were often several hundred dollars per lb P removed. On the whole, targeted sweeping appears to be a cost-effective strategy for nutrient reduction when compared to treatment ponds, where costs are generally higher.

Figure 4: Average cost of phosphorus removal by month with different sweeping frequencies (high canopy sweeping routes), Prior Lake, MN.

We are currently developing a guidance manual and a workshop series on street sweeping best practices. These workshops will introduce tools and methods for predicting sediment and nutrient loads to street surfaces in urban areas based on overhead tree canopy; estimating the amount of material that can be removed and the cost of removal in targeted sweeping operations; and designing sweeping programs to meet nutrient reduction goals.

Our next steps include (1) extrapolating findings to other cities, based on canopy mapping; (2) developing an economic optimization, to find the “best” frequency and timing of sweeping operations throughout the year for varying canopy levels; and (3) estimating the impact on water quality of downstream lakes, after accounting for additional P removal in stormwater control measures downstream from swept streets. Preliminary analysis suggests that sweeping will substantially reduce P loads to lakes in watersheds with tree-shaded streets. Ultimately, we envision a paired watershed or paired lake study to examine the effect of sweeping over a period of several years.


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:

  • Do street leaves contribute to your maintenance problems by clogging catch basins or stormwater ponds?
  • Would you have the capacity to increase sweeping during the late spring and fall if you knew that you could reduce P loads to nearby lakes substantially (enough to make a visible difference)?
  • Would our upcoming spreadsheet tool "Quantifying nutrient removal by street sweeping" be of interest to you as a planning tool?

I would be curious as to how pH affects the nutrient availability if at all. Do acidic leaves such as pines needles produce similar results to others leaves such as Elm? Can a universal model or planning tool that is developed based on treat types in a specific region be applied to other geographic areas with varying tree types?

Thanks for your interest in the project. I passed along your pH question to our ecologist. Regarding the regional dependence of the model, it is true that our experiment was conducted in a region dominated by deciduous trees (although certainly a mix of species is present) and our metrics may not translate well to predictions of coarse organic loads in regions where tree canopy coverage is dominated by ever-green species or other significantly different types of tree canopy. I imagine it would be possible to develop effective export coefficients that would correct for differences in regional canopy composition for load projections based on percent tree canopy cover. Perhaps this could be developed through a state-wide or national database of street sweeping records.

It is also true that the phosphorus content of leaves and litter varies among tree species. In looking at nutrient analysis results, I have found the the average phosphorus content of our coarse organic load was around 1.7 mg/g dry coarse organics, which is similar to a mean or median value for leaf tissue among common urban deciduous species. But from resources I've look at, conifer needles tend to have a lower phosphorus content than deciduous broadleaf - and then there is your question about pH. So again, our metrics should give a reasonable estimate of nutrient loads in similar climatic regions with mixed species, but may have to be adjusted for significantly different canopy composition.

If practitioners have the resources to do some testing, nutrient concentrations could be determined by fractionation and nutrient analysis for a subset of sweeping operations. The measured concentrations could then be applied to the larger set of records to track nutrient recovery. Our guidance manual will include sections on both estimating nutrient loads for planning sweeping operations and tracking nutrient recovery through sweeping.

That's a really interesting question. Our estimates apply in areas where the canopy is mainly deciduous, as most street tree species are. While leaf and litter phosphorus do vary among tree species, our values are based on a mixed canopy and probably offer a reasonable rough estimate where the canopy is mainly deciduous species. That said, we are currently conducting further research to compare the common street tree species planted in the Twin Cities in terms of their leaf and litter nitrogen and phosphorus. We'll know more by the end of the summer about whether there are big differences among species or not.