P8 Street Sweeping Modeling

Caution: This page is in development

Several methods of crediting street sweeping exist. One crediting method is using simulation modeling to estimate phosphorus removal in leu of more direct monitoring methods. This memorandum provides a summary of a series of P8 (Program for Predicting Pollution Particle Passage through Pits, Puddles, and Ponds) stormwater models used to estimate phosphorus load reduction resulting from street sweeping in four municipalities in the Twin Cities Metro Area of Minnesota.

A challenge of using simulation models is that the algorithms representing pollutant generation and transport in storm runoff, as well as the impact of street sweeping, are themselves uncertain. In addition, results depend heavily on parameter selection within the model. To provide a basis for comparing P8 results to real world data, the models were configured to represent 2019 street sweeping operations for routes with focused data collection efforts conducted by University of Minnesota (U of M). The expectation is that the model results are likely to be biased (low or high) relative to the phosphorus reductions measured by U of M. The team developing the street sweeping credit assessed whether the model can be used within the crediting framework, with an appropriate bias adjustment factor. This page describes the P8 modeling effort.

Overview of P8 model

The P8 Urban Catchment Model is a physically based continuous stormwater quantity/quality model. P8 is capable of predicting the generation and transport of stormwater runoff and associated pollutants (using dissolved fractions and sediment association with four particle size classes) from urban watersheds. P8 simulates pollutant transport and removal in a variety of BMPs including street sweeping. P8 provides loading estimates based on data collected as part of the EPA National Urban Runoff Pollutant (NURP) program. Key inputs include meteorology (hourly rainfall and daily temperature), watershed conditions, and particulate and pollutant loading information.

Watershed runoff is predicted using the curve number method. The user provides various characteristics including watershed area, pervious composite curve number, indirectly connected impervious area fraction, directly connected impervious fraction, impervious depression storage, and two factors that influence initial abstraction in addition to depression storage. Some additional scaling factors are provided as well.

Pollutant mass in runoff is governed by buildup/washoff equations for impervious surfaces and concentration-runoff intensity curves for pervious surfaces. P8 requires the user to specify characteristics for the five modeled particle size fractions (“particle parameters”) and pollutant particle composition associated with each (“water quality components”). These parameters establish how particles are generated by watersheds, how particles are removed by BMPs, and the amount of pollutant associated with each particle class. P8 contains a default particle file as well as two particle files based on findings from the EPA NURP study. Minnesota guidance for using P8 (MPCA, 2019) recommends using the NURP 50th percentile particle parameters and water quality components files, which were used for this modeling project. Particle fractions and associated sediment characteristics are shown in Table 1. Note that the particle classes are focused on sediment that is readily transported in stormwater runoff; courser sands tend not to be mobilized under most storm event conditions.

Street sweeping P8 particle fractions and sediment assumptions
Link to this table

Particle Fraction	Size (µm)	Percent of TSS	USDA Size class
P0%	NA1	0%	NA
P10%	2-8	20%	Clay, Very fine silt
P30%	8-20	20%	Fine silt, Medium silt
P50%	20-60	20%	Medium silt, Course silt
P80%	60-200	40%	Very fine sand, Fine sand

¹ used for dissolved pollutants