Pennsylvania Handbook of Best Management Practices for Developing Areas

Appendix F -- Runoff Capture Design

The development of natural watersheds generally produces two types of changes in hydrology:

1. Peak discharge rates increase, leading to increases in the prevalence of impacts from erosion and flooding
2. Percentage of total precipitation that runs off without infiltrating increases, leading to reduced base flow to streams. The reduced base flow affects groundwater resources, surface-water quality, and habitat.

In most natural watersheds in Pennsylvania, 50 percent or more of the annual rainfall infiltrates. A part of the infiltrated water volume will be returned to the atmosphere through evaporation or plant transpiration. The remainder will percolate to the water table from where it will replenish the groundwater supply or reemerge as base flow (i.e., dry-weather discharges) to streams and wetlands. Development, which increases the proportion of runoff, poses a significant threat to the environmental resources of Pennsylvania’s groundwater, streams, and wetlands. Therefore, attention is increasingly being turned to methods of increasing infiltration and groundwater recharge in developed areas.

The recommended criterion for addressing groundwater recharge is to maintain the annual volume of total runoff at predevelopment levels. This design approach is implemented by selecting a minimum runoff capture volume that must be completely infiltrated on the site during every storm. The runoff capture volume is conveniently stated as a rainfall volume, in inches, over the area of the site. The approach for complying with this design criterion is straightforward. For example, if the specified minimum runoff capture volume for a watershed is 0.75 inches, all rainfall events of less than 0.75 inches of rainfall should be completely retained on the site and infiltrated.

The criterion may include a design storm that shows how the 0.75 inches of rainfall should be distributed in the design calculation. If a runoff capture design storm is used, then a stormwater runoff simulation algorithm can be used to evaluate the performance of the site’s stormwater management system. If the analysis shows that runoff from the site will be negligible during the runoff capture design storm, then the criterion for runoff capture will be satisfied for all storms.

This approach requires the selection of an appropriate runoff capture volume. The first step is to estimate the runoff ratio that characterizes the undeveloped or natural watershed. The runoff ratio can be estimated by constructing a water budget for a representative watershed. As an alternative, an approximate runoff ratio that reasonably characterizes the watershed can be drawn from the literature. The prevalence of low-permeability soil (i.e., NRCS hydrologic soil group D), steep slopes, or thin soil generally will be associated with higher runoff ratios. In Pennsylvania, runoff ratios can be expected to range between 25 and 50 percent of total annual rainfall.

The analysis of storm frequency data available from PaDOT (Aron, et al., 1986) or the NRCS (USDA, 1972, USDA, 1986) can be used to select an appropriately sized runoff capture storm. In the following example, a runoff capture design storm is developed for Pennsylvania rainfall region No. 2 (Aron, et al., 1986). A runoff ratio of 40 percent was assumed (i.e., 60 percent of annual rainfall is retained and infiltrated). In the example, the runoff capture design storm will have the property that 60 percent of the annual rainfall will occur in storms of equal or smaller magnitude. Therefore, controlling runoff from the runoff capture design storm will be the only requirement for preserving the overall water budget for the watershed.

The 24-hour rainfall amounts for storms of various magnitude, up to the 100-year return-frequency event, have been reported (Hershfield, 1961, Aron, et al., 1986, USDA, 1986). When the data are plotted as a log-frequency distribution, the data approximate a straight line (see Chart F-1). By using a linear interpolating function to represent log-frequency versus 24-hour rainfall, the probability function for the rainfall data was constructed (see Chart F-2). The example emphasizes that rainfall events with magnitudes larger than about 2.5 inches can be ignored when considering runoff volume.

The importance of the probability function is that the cumulative quantity of rainfall from a range of storm event sizes can be determined by integrating under the curve within the range. By using this approach, the cumulative rainfall volume of rainfall events of 0.59 inches or less was determined to be approximately 60 percent of the total annual rainfall. Therefore, the runoff capture volume was selected to be 0.60-inch rainfall. The runoff capture design storm was selected as the 24-hour, 0.60-inch rainfall event with a NRCS type II rainfall distribution (USDA, 1972). A storm of this magnitude will occur on average 9 times a year. Table F-1 presents the 60-percent, 75-percent, and 90-percent storm magnitudes computed for the five Pennsylvania rainfall regions. A map showing the locations of the regions is provided at the end of this appendix.

Chart F-1. Exceedance Frequency

Chart F-2. Probability Function

Because the technique is based on frequency data for large storm events, the contribution of small rainfall events to the total rainfall is underestimated. However, it has the virtues that storm frequency data is readily available, and that the results of this analysis will always provide a conservative overestimate of design storm magnitudes. If desired, a more accurate, and less conservative, estimate of design storm magnitude can be developed using daily rainfall records from local weather stations (Schueler, 1996).

By using this information, a runoff capture table can be generated for the runoff capture design storm. Table F-2 offers a convenient way for determining if a site’s stormwater management plan will comply with the minimum runoff-capture criterion. Table F-2 was calculated for a 0.65-inch runoff capture volume. Runoff capture storage is computed using the runoff formula implemented in the NRCS soil-cover complex methodology (USDA, 1972, USDA, 1986). The runoff capture storage is the quantity of water that must be captured and retained onsite to prevent runoff from occurring during the runoff capture design storm.

where:

R_s = Runoff capture storage (inches, over the drainage area)

V = Runoff capture volume (inches, over the drainage area)

S = Potential maximum retention after runoff begins (inches, over the drainage area)

CN = NRCS runoff curve number

The runoff capture storage requirement for areas underlain by soils of the "D" hydrologic soil group can be adjusted to reflect the naturally higher runoff rates associated with undeveloped conditions.

The table is used by dividing the project site into groups of connected surfaces using the following rules:

• Surfaces can only be combined into one group if the least-permeable surface is tributary to the most-permeable surface. For example, a rooftop area that discharges to a filter strip can be combined with the filter strip.
• Surfaces that receive concentrated flow from tributary areas cannot be combined with these areas. In the example above, the rooftop and filter strip areas can be combined only if the rooftop drainage is distributed to the filter strip by a manifold or spreader.

Average runoff curve numbers (CN) for the combined surfaces then can be developed using the NRCS soil cover complex method (USDA, 1972, USDA, 1986). Equivalent CNs for permeable pavements are provided in the permeable pavement BMP. The runoff capture storage, expressed in inches, for each group can be determined directly from the table. The value of routing flow from impervious surfaces to permeable surfaces becomes apparent. The runoff storage requirement can be satisfied by retaining and infiltrating runoff in a variety of BMPs including:

• Infiltration trenches
• Dry wells (i.e., volume below the overflow)
• Storage-infiltration reservoirs in a pavement subbase
• Bioretention facilities

The storage volumes of these facilities can be subtracted directly from the runoff capture storage computed in this procedure.

Compliance with the runoff capture requirement also can be demonstrated using a stormwater runoff simulation algorithm.

Figure F-1. Delineated regions with uniform rainfall (Source: Field Manual of Pennsylvania Department of Transportation, May 1986).