Planning for enhanced sweeping
In a prior study we estimated the amount of P in leaf-fall for various species. The resulting database, Twin Cities Household Ecosystem Project/Experiment 267, of 4,000 trees can be used to estimate P inputs.
For example, in a typical residential neighborhood with boulevard trees (20-inch-diameter maples placed at 60-foot intervals), 30% of the P entering streets comes from leaves. Based on these calculations, we wondered: Could street sweeping be part of a solution that increases the effectiveness and efficiency of stormwater programs?
Answering that question led us to collaborate with the City of Prior Lake, Minn., on unique research. Five attributes make the Prior Lake Street Sweeping Study different from other sweeping studies:
- Sweeping was conducted from snowmelt to first snowfall, thereby including autumn leaf-fall.
- P content of coarse organic material (leaves and other vegetative debris) was measured.
- P removal was quantified by measuring P in the swept material, enabling a direct calculation of P removal.
- Sweeping frequencies of seven-, 14-, and 28-day intervals were used.
- Routes varied from 0% to 20% canopy cover, thus allowing us to correlate tree canopy with P removal.
We also evaluated labor, maintenance, and capital costs for each of 400 sweepings conducted over the two-year project period.
We realize that no city is likely to sweep as often as we did in this experiment. The objective of high-frequency sweeping was to provide data for developing a planning tool that would enable public works departments to estimate removal of solids, P, and N based on various combinations of tree canopy percentages and sweeping frequency.
Overall, we found that canopy percentage and sweeping frequency do indeed affect mass of solids, total P, and total N removed by sweeping.
Figure 1 above, for example, shows how these two factors specifically impact P removal. Rates vary by a factor of four, from 1.4 pound/curb mile per year for a route with less than 1% canopy cover that’s swept every 28 days to 6.2 pound/curb mile per year for a route with 19% cover swept every seven days.
We also found that the “coarse organic fraction” of sweepings (the mostly vegetative debris that many earlier researchers discarded) represents a substantial amount of nutrients. On average, the material comprised 20% of solids, 42% of P, and 74% of N on high-canopy routes.
A statistical cross-validation study showed that the model’s predictions were within 6% of measured removal rates, indicating that the Prior Lake research is statistically robust. As a result, a public works department using the Street Sweeping Planning Calculator Tool can assume P removal can be estimated within 5% by the total weight of sweepings.
Next page: Opportunities and barriers
