• Reddit
• Digg
• Facebook
• MySpace
• del.icio.us
• StumbleUpon
• Technorati
• Furl
• Newsvine
• Google Bookmarks

Additional Article Content

• Figures One, Two, and Three
• Figure Four
• Figure Five
• Figure Six
• Table One
• References

You may also be interested in...

• Development of a Scoring System for Construction-Site BMP Assessment

Figure 1. Accumulation of sand near an inlet

In this article we describe a new approach developed at the University of Minnesota that includes controlled testing as a key component in a four-level assessment program. In order of increasing effort and cost, the four levels are

1. Visual inspection
2. Infiltration capacity testing
3. Synthetic runoff testing
4. Monitoring

Visual inspection involves examination of the vegetation and soil and is used to quickly determine if a rain garden is malfunctioning and in need of maintenance or replacement. Infiltration capacity testing involves the use of infiltrometers to make local measurements of surface infiltration throughout an LID practice. In synthetic runoff testing, a fire hydrant or water truck is used to fill the basin with water, and the overall drain time of the rain garden is determined. Monitoring is the measurement of runoff quantity and quality from natural storms; it can be used to assess the function of infiltration practices within a given watershed. Monitoring is especially useful for watershed-scale studies to assess overall pollutant loads to receiving waters and the impact of LID practices on these loads.

Level One: Visual Inspection
Visual inspection involves a comprehensive evaluation of the vegetation and the soil in the rain garden and is commonly used by many organizations. We have simply standardized this assessment approach for the state of Minnesota to consider in meeting the “evaluate the effectiveness” requirements of National Pollutant Discharge Elimination System permits. The visual inspection may be simple or comprehensive depending on the site conditions and the purpose of the assessment. A comprehensive visual inspection requires some knowledge of both vegetation and soils. Simpler observations (visiting a site after a storm event to check for standing water) are valuable and require less effort, but limited information is obtained.

The first step in a visual inspection involves examination of the rain garden for obvious hydraulic problems. Ponded water should be present for no more than 48 hours for events that fill the rain garden to the outlet. A drainage period that exceeds 48 hours, or a shorter period with smaller storms, indicates that the rain garden probably needs rehabilitation.

Sediment accumulation in the basin could reduce infiltration rates. Clogged inlet structures (Figure 1) will cause stormwater to bypass the best management practice (BMP). Clogged outlet structures could lead to local flooding problems.

The next step of a detailed visual inspection is the assessment of the vegetation. Several issues to consider when assessing the vegetation include the age of the rain garden, time of the growing season, species present and their growth requirements, and condition of the site. Vegetation plays a critical role in maintaining infiltration capacity and good plant health often signifies proper function. Furthermore, most plants used in rain gardens are not adapted to flooding and cannot survive submergence for long periods of time (Shaw and Schmidt 2003). The species present should be observed over time using the original vegetation design plans and photographic records. For example, the emergence of wetland plant species (Figure 2) suggests prolonged saturation or ponding due to poor infiltration/drainage (Richardson and Vepraskas 2001). A decline in the health of the vegetation (e.g., percent cover) or a dramatic shift in species present is an indication that the rain garden may not be functioning as designed.

The last step of the visual assessment of rain gardens is the inspection of the soil. Examining the entire soil profile of the infiltration practice is important for the detection of restrictive soil layers, which will control the rate at which water moves through the soil profile. The infiltration characteristics of rain garden soil are directly related to the hydraulic conductivity and the porosity of the soil, which, in turn are affected by the texture and the bulk density of the soil (Hillel 1998). The textural class of a soil, shown in Figure 3, allows for the rough estimation of both porosity and hydraulic conductivity (Saxton and Roth 2005). Soil color is easy to determine and provides an indirect measure of soil characteristics, such as water drainage, aeration, and organic matter content (Foth 1990). For example, soils that are gray in color or contain mottles (i.e., small areas of gray, red, yellow, brown, or black that differ in color from the bulk soil) are indicative of hydric soils (Richardson and Vepraskas 2001) associated with prolonged water saturation, suggesting that stormwater runoff is not infiltrating/draining properly.

Case Study. Twelve rain garden sites in Minnesota were included in the development and evaluation of the four-level methodology during the 2006 field season. These sites were selected based on the following criteria: 1) permission and participation from the owner/operator of the site, 2) availability of site information (e.g., site plans, planting diagrams), and 3) proximity to the inspectors’ home location. A summary of the rain garden characteristics and the levels of assessment used at each site are provided in Table 1. The sizes of the rain gardens ranged from 28 to 1,350 square meters. The smallest rain garden was located in a residential area receiving stormwater runoff from the street via a curb cut inlet. Several other rain gardens received runoff from parking lot areas or from a combination of stormwater runoff sources.

Application of Visual Inspection. Four of the sites had obvious problems with infiltration observed during visual inspection. These included ponded water, wetland plants, and a lack of plants on compacted soil. These rain gardens needed to be rehabilitated and were not evaluated further. Of the remaining eight, the three oldest rain gardens were installed in the fall of 2003 and were online (i.e., receiving runoff) in the spring of 2004.

At the University of Minnesota–St. Paul campus rain garden, a lower-permeability silt loam soil layer was placed during construction under the sandy loam topsoil. This decision was made onsite by individuals who were qualified to move soil but not qualified to estimate the permeability of soil. Two additional sites had underlying native soil of finer texture than the overlying topsoil. These two sites were designed with underdrains to compensate for the restrictive layer. The soil profile at the Cottage Grove site consisted of 40 inches of sand overlying gravel. The poor retention of water and nutrients by sand are a potential cause of the failing plants observed during inspection of the vegetation.

Level Two: Infiltration Capacity Tests
The ability to infiltrate water under saturated (flooded) conditions is indicated by the saturated hydraulic conductivity (). Several field devices for determining the  of surface soils were evaluated by Nestingen et al. (2008a). The novel Modified Philip-Dunne (MPD) infiltrometer developed in our laboratory (Nestingen et al. 2008b) is preferred due to the minimal volume of water necessary, ease of use in the field, low cost of the device, and transportability of the equipment.

The MPD infiltrometer (Figure 4) is a relatively simple device consisting of a thin-walled (2-millimeter-thick) aluminum cylinder with a height of 45 centimeters and an inner diameter of 10 centimeters. A transparent piezometer tube is attached to the outside of the device alongside a measurement tape for making visual water-level readings. The device is pounded into the soil to a depth of 5 centimeters after any mulch or detached/decaying plant material has been brushed aside, and then it is filled with water to a height of 43 centimeters. The water level over time is then recorded. MPD infiltrometer measurements should be made at a number of locations throughout each basin, and multiple MPD measurements can be made simultaneously.

Soil moisture content before and after the infiltration test is also needed to determine . A capacitance probe, which measures the dielectric constant of the soil, can be used to indirectly estimate the initial and final soil moisture content of the top 6 centimeters of soil in the vicinity of the infiltrometer. A soil-specific calibration, using several gravimetric soil moisture measurements, is necessary at each rain garden. Bulk density measurements, required to convert gravimetric water content to volumetric water content, should be made using the core method (Klute 1986).

The radius of the tube, the change in volumetric moisture content, and the water level versus time data are used in a curve-fitting routine to compute an estimated near the surface and to compute the wetting front suction for the saturated-unsaturated interface at each test location (Nestingen et al. 2008b). Low  values indicate areas with soil pore clogging that require maintenance to restore the infiltration capacity of the system. The measurements can also be used to estimate drainage time if the near-surface soil is the restrictive layer. Infiltration capacity testing can be performed at most rain gardens, and does not rely upon a large supply of water (such as the 4,000 or more gallons needed for synthetic runoff tests) or proper locations to quantify inflow and outflow (such as for monitoring).

Application of Capacity Testing. Infiltration capacity tests were applied to the eight functional rain gardens remaining after visual inspection. Six MPD infiltrometers were used simultaneously in our tests, allowing infiltration capacity testing of up to 40 locations in a rain garden to be completed within eight hours. The calculated  values for each measurement location were entered into ArcView to provide a map showing the spatial variability in  for each rain garden, as shown in Figure 5. The median and the arithmetic mean were then calculated for each site. For comparison with synthetic runoff tests, the drain time was calculated based on the median or mean  value,  from the infiltrometer results; the surface area in square meters, A, of the rain garden at full capacity; and the volume of the basin in square meters, V:

The range of  values in each rain garden was substantial (two to four orders of magnitude), considering that all rain gardens were built with engineered soils. The median values were between 1.6 x 10^-3 and 2.0 x 10^-2 for the eight sites. Most of these values are typically associated with sand or gravel (Saxton and Rawls 2005), with the exception of one site with a  which is typically associated with sandy loam. Our sites, however, had a sand loam to silt loam at the soil surface. The large values are likely caused by the mulch and macropores from decaying roots in the surface of the rain gardens.

Level Three: Synthetic Runoff Tests
Synthetic runoff tests provide the drainage time that can be expected when the basin encounters a design storm. They involve the use of a fire hydrant or water truck to fill the basin to capacity, as shown in Figure 6. The time required for the water to drain is recorded and the drain time is determined. Pollutants can be added to the water to determine the pollutant-removal efficiency of the infiltration practice if a means of collecting outflow is available. The synthetic runoff test provides a quick, easy, and cost-effective way to determine the infiltration capacity and pollutant retention of a rain garden, green roof, or other infiltration practice, as long as sufficient water is available.

Once permission to use a fire hydrant is obtained or arrangements to use a tanker truck are made, about four hours is required to test a rain garden. First, the basin is filled with water and the water supply is turned off. Then, water level versus time is recorded using a staff gauge and a stopwatch until the basin is fully drained. The observed drain time is then compared to the design drain time (when known).

Although synthetic runoff testing is relatively quick and easy to perform, there are some limitations to its applicability. For example, some rain gardens, such as the Thompson Lake or the University of Minnesota–Duluth rain gardens in this study, have basins that are too large to fill with available supplies of water. A large tanker truck contains only 27 square meters (6,000 gallons) of water, and most fire hydrants may be used with permission to supply between 2 and 4 cubic feet per second for up to 20 minutes, or 80 to 160 square meters of water. Furthermore, unlike infiltration capacity testing, synthetic runoff testing cannot be used to identify specific locations of poor infiltration within a rain garden because a basinwide overall infiltration rate is determined. Thus, if a rain garden drains too slowly, infiltration capacity testing can be used to identify areas within the rain garden that require maintenance.

Application of Synthetic Runoff Tests. Three rain gardens—Cottage Grove, the University of Minnesota–St. Paul, and the Ramsey Washington Metro Watershed District—were subjected to synthetic runoff tests. A water truck was used for Cottage Grove, and fire hydrants with fire hoses to convey the water to the rain garden were used for the other two sites. Sandbags were used to hold the fire hoses, and pads to dissipate the water were placed temporarily in each rain garden. Complete drainage occurred in 0.14, 2.13, and 3.13 hours, respectively, or well below the required 48-hour drainage period. This indicates that the rain gardens were functioning properly with respect to infiltration. One problem with the Cottage Grove rain garden, however, is that the soil has very low water-holding capacity, making it poorly suited for development of a healthy stand of vegetation.

Garden drain times can also be compared with those calculated from the near-surface  values of earlier MPD tests. The median values of obtained from the MPD tests were found to provide the best estimate of drain time based on the same initial volume used for each synthetic runoff test. The exception would be the University of Minnesota–St. Paul rain garden, which had a lower predicted drain time than the corresponding synthetic runoff test drain time. The difference between the estimated and measured drain times for the University of Minnesota–St. Paul rain garden is likely due to the influence of the restrictive soil layer on drainage when the rain garden is filled to capacity.

Level Four: Monitoring
Monitoring is the most comprehensive assessment method and relies on natural rainfall and runoff. Flow measurement, sample collection, and sample analysis are required to determine the volume of water and mass pollutants entering and exiting the system. Although both monitoring and synthetic runoff testing can provide information on infiltration and pollutant-removal performance, monitoring can also provide data on runoff and pollutant loading for the local watershed. Nevertheless, monitoring results will typically have a larger associated uncertainty because of the lack of control over storm flows and pollutant concentration (Weiss et al. 2007).

Monitoring, however, is still valuable for rain gardens as a composite assessment tool. If there is a small watershed with a number of rain gardens and other LID practices, careful monitoring of “replicate” watersheds before and after installation will provide a powerful measure of their composite impact in both volume and pollutant-load reduction. It is at the composite assessment level that monitoring is advocated, because the effort and expense can be justified in the demonstration of both reduced volume and reduced loads.

Conclusions
The four-level assessment approach provides several new tools for the assessment toolbox in addition to monitoring: visual inspection (level one), infiltration capacity testing (level two), and synthetic runoff testing (level three). All three assessment approaches provide useful information regarding the overall function of rain gardens and can also be used on other LID practices. Visual inspection of the vegetation and soils gives a preliminary indication of the ability to uptake and infiltrate stormwater runoff. Infiltration capacity testing provides information on the spatial variability in saturated hydraulic conductivity. The combination of these two assessment methods is particularly useful for assisting in the development of maintenance tasks and schedules. Infiltration capacity testing can also be used to ensure that the construction of the infiltration practice was done properly and allows for the identification of locations that may have been compacted during construction. Although infiltration capacity testing has numerous benefits, use of  provides only a rough estimate of the time required for the infiltration practice to drain. The synthetic runoff test will give a more precise measure of the drainage time, if sufficient water is available.

A multilevel assessment approach allows for the identification of problems, potential causes, and possible solutions. Therefore, a three-level assessment (levels one, two, and/or three) will thoroughly evaluate the ability of an infiltration practice to infiltrate and treat stormwater. Nevertheless, when there are a large number of LID practices to evaluate and a multilevel assessment of each practice is not feasible, assessment by visual inspection (level one) should be done annually to identify problematic infiltration practices requiring further assessment and rehabilitation. Monitoring (level four) is recommended as a comprehensive assessment tool for a watershed with a number of LID practices, including rain gardens. A before-and-after installation comparison will provide strong evidence of the effectiveness of these practices in reducing the volume of runoff and reducing the pollutant load associated with the runoff.

The four levels are described more fully with case studies in the online manual Assessment of Stormwater Best Management Practices at http://wrc.umn.edu/outreach/stormwater/bmpassessment/assessmentmanual/index.html.