m (1 revision)
This article provides information on design considerations and design criteria for permeable pavement.
Permeable pavement is subject to potentials and constraints as noted below.
Permeable pavements should be at least a minimum horizontal distance of 100 feet from any municipal water supply well [Diane will send reference on MN well head reqs] and 50 feet from septic systems. Underground utility lines are best located away from permeable pavement bases. However, if they need to penetrate the base, consideration should be given to waterproofing (depending on the utility) or possible encasement using low-strength flowable concrete fill. Setbacks can be reduced at the discretion of the local authority for designs that use underdrains and/or liners.
Permeable pavement can be used for roads for low low-volume roads. These can be residential collector roads or other applications with similar traffic loads. Additional technical guidance and approval from MnDOT should be sought regarding road applications in the public right of way. Permeable pavement can be used in heavily trafficked/high wheel loads. These applications will likely require specialist pavement engineering and may require using open-graded bases stabilized with asphalt or cement.
Base/subbase thickness is determined for support traffic using structural design methods and for water storage using hydrologic sizing and/or dynamic modeling over time. The thicker of the two resulting designs is used.
The structural design process for supporting vehicles varies according to the type of pavement selected. The pervious concrete industry is in the process of developing ASTM test methods for characterizing compressive or flexural strengths of pervious concrete. These tests are needed to model fatigue under loads. As an interim step, fatigue equations published by the American Concrete Pavement Association (ACPA 2010) assume such inputs to be comparable in nature (but not magnitude) to those used for conventional concrete pavements. The ACPA design method should be consulted for further information as well as pervious concrete industry software. General guidelines for pervious concrete surface thickness are published by the National Ready Mix Concrete Association and the Portland Cement Association (Leming 2007).
Porous asphalt (Hansen 2008) and permeable interlocking pavements (Smith 2010) use flexible pavement design methods adopted from the 1993 AASHTO Guide for Design of Pavement Structures (AASHTO 1993). Base/subbase thicknesses can be calculated using PICP industry software. In addition, MnDOT design methods, approved mechanistic principles, and manufacturer’s specific recommendations should be consulted.
There has been little research on or full-scale testing of the structural behavior of open-graded bases used under permeable pavements to better characterizing relationships between loads and deformation. Therefore, conservative values (i.e., AASHTO layer coefficients) should be assumed for open-graded base and subbase aggregates in permeable pavement design.
Regardless of type of permeable pavement, structural design methods consider the following in determining surface and base thicknesses to support vehicular traffic:
Soil stability under traffic should be carefully reviewed for each application by a qualified geotechnical or civil engineer and lowest anticipated soil strength or stiffness values used for design. Structural design for vehicular applications should generally be on soil subgrades with a CBR (96-hour soaked per ASTM D 1883 or AASHTO T 193) of 4%, or a minimum R-value = 9 per ASTM D 2844 or AASHTO T-190, or a minimum Mr of 6,500 psi (45 MPa) per AASHTO T-307.
Soil compaction required to achieve these soil strengths will reduce the infiltration rate of the soil. Therefore, the permeability or infiltration rate of soil should be assessed at the density required to achieve one of these values. If soils under vehicular traffic have lower strengths than those noted above, or are expansive when wet, there are several options: use underdrains, thicken the base/subbase layer(s), stabilize the base layers with cement or asphalt, and or stabilize the soil with lime or cement. These options are typically used in combination. Pedestrian applications can be placed on lower strength soils than those noted.
Permeable pavement will likely be designed to store more than 1.1 inch of stormwater from the site. In addition, a mix of adjacent pervious and impervious surfaces contributing runoff will require that the base storage volume to hold water depths over 1.1 inches, especially if volume and peak discharge reductions are design objectives. The soil subgrade infiltration rate typically will be less than the flow rate through the pavement, so some reservoir storage will usually be required. When working with HSG A or B soils, designers should initially assume that there is no outflow through underdrains. Equation 1 can be used to determine the depth of the reservoir layer, assuming runoff fully infiltrates into the underlying soil:
The maximum allowable depth of the reservoir layer is constrained by the maximum allowable drain time, which is calculated using Equation 2.
The following design assumptions apply to Equations 1 and 2:
If the depth of the reservoir layer is too great (i.e. dp exceeds dp-max), then the design method typically changes to account for underdrains. The storage volume in the pavements must account for the underlying infiltration rate and outflow through the underdrain. In this case, the design storm should be routed through the pavement to accurately determine the required reservoir depth.
If the depth of the base/subbase for the full infiltration system is excessive, (i.e. dp exceeds dmax) because, as an example, the design subgrade soil infiltration rate is not adequate to remove the water from the design storm within the designated period of time, then the design should include underdrains. The following procedure is for sizing the base/subbase for partial infiltration designs (i.e. contains underdrains). The same symbols apply, but with the following additions:
Equation 3 can be used to approximate the outflow rate from underdrain(s). This equation is based on Darcy’s Law, which summarizes several properties that groundwater exhibits while flowing in aquifers. Although the hydraulic conductivity (measure of the ease with which water can move through pore spaces of a material) of the aggregate subbase is very high (~17,000 ft/day or 8,500 in./hr), the discharge rate through underdrains is limited by the cross sectional area of the pipe. As the storage volume above/around the underdrain(s) decreases, i.e., the hydraulic head or water pressure decreases; the base/subbase and in turn the underdrain(s) will drain increasingly slower. To account for this change in flow conditions within the subbase and underdrain(s) over time, a very conservative coefficient of permeability (k) of 100 ft/day 8.33 ft/hr) per pipe can be used to approximate the average underdrain outflow rate.
Once the outflow rate through each underdrain has been approximated, Equation X.4 can be used to determine the depth of the base/subbase needed to store the design storm. To estimate the number of underdrain pipes (n), take the dimension of the parking lot in the direction the pipes are to be placed and divide by the desired spacing between pipes – round down to the nearest whole number.
Tfil = T when the underdrains are at the bottom of the subbase. Tfill = ½T (approximation) when the underdrains are raised.
With full infiltration systems, the maximum allowable drain time (dmax) needs to be calculated to make sure the stored water within the base/subbase does not take too long to infiltrate into the soil subgrade. However, for partial infiltration systems, there is a second method of storage water discharge, namely the underdrains. The depth and number of underdrains are variables that can be adjusted (unlike the infiltration rate into the soil subgrade) so that the actual drain time equals or is less than the maximum allowable drain time. If the discharge of the underdrains is added to Equation X1, then:
Rearranging the previous creates Equation X.5, with the only remaining variable (T1) being on the left hand side.
The elevation of the pipe above the soil subgrade is then calculated using Equation X.6.
Permeable pavement can also be designed to augment detention storage needed for channel protection and/or flood control. The designer can model various approaches by factoring in storage within the base/subbase, expected infiltration and any outlet structures used as part of the design. For example, some applications in the Chicago, Illinois area are designed to detain and partially infiltrate the 100-year storm for managing combined sewer overflows.
Once runoff passes through the surface of the permeable pavement system, designers should calculate outflow pathways to handle subsurface flows. Subsurface flows can be regulated using underdrains, the volume of storage in the reservoir layer, the bed slope of the reservoir layer, and/or a control structure at the outlet.
Permeable pavement nutrient reduction capabilities can be enhanced in partial infiltration designs that detaining water in the base/subbase for over 24 hours for de-nitrification. Any infiltration will reduce the water volumes and mass outflow of nutrients and suspended solids. PICP can use specially coated aggregates in the joints and bedding and all systems can use them in the base to reduce nutrients. Coated aggregates (sometimes called ‘engineered aggregates’) have an effective life of seven to ten years.
A filter layer made of sand or fine aggregate placed under or sandwiched within permeable pavement bases are occasionally used to as a means to reduce nutrients. This layer can be enhanced with iron filings for phosphorous reduction (Erickson 2010). Their effectiveness, initial and maintenance costs should be weighed against other design options for nutrient reductions. Sand filters will incur additional construction expense and this can be reduced by placing sand filters under the subbase at the down slope end of a permeable pavement. The disadvantage of sand filters is that they will eventually require removal and restoration. Therefore concentrating their location in the down slope areas of the site can help reduce future maintenance costs and site disruptions. Simply storing water for and slowly releasing it over 48 hours can be less expensive alternative for nitrogen reduction as long as there is a carbon source in the base/subbase.
A second approach useful for nutrient reduction can occur on sloping sites by creating intermittent berms in the soil subgrade. These enable settlement of suspended solids and encourage de-nitrification. A third alternative is using a “treatment train” approach where a permeable pavement initially filters runoff and the remaining water outflows to bioswales or rain gardens adjacent to the pavement for additional processing and nutrient reduction. There may be additional BMPs used to remove nutrients as the water moves through the watershed.
A minimum of one test per ASTM D3385 or D5093 must be taken per 7,000 sf of planned permeable pavement surface area. In most cases, a single soil test is sufficient for small-scale applications. At least one soil boring must be taken to confirm the underlying soil properties at the depth where infiltration is designed to occur (i.e., to ensure that the depth to water table, depth to bedrock, or karst is defined). Soil infiltration testing should be conducted within any confining layers that are found within 5 feet of the bottom of a proposed permeable pavement system.
Permeable pavement designs should include methods to convey larger storms (e.g., 2-year, 10-year) to the storm drain system. The following is a list of methods that can be used to accomplish this:
The reservoir below the permeable pavement surface should be composed of clean, washed crushed stone aggregate and thickness sized for both the storm event to be stored and the structural requirements of the expected traffic loading. The recommended minimum void ratio should be 40% per ASTM C29. Reservoir base layers for pervious concrete are typically washed ASTM No. 57 stone and those for porous asphalt are ASTM No. 2, 3, or 5. PICP uses ASTM No. 2,3 or 4 stone.
If exposed to vehicular loads, all crushed stone should be MnDOT Class A or B coarse aggregate, minimum 80% crushed, typically granite, basalt, gneiss, trap rock, diabase, gabbro or similar material. The maximum Los Angeles Rattler Loss should be 35% per AASHTO T-96 and no greater loss than 10% per AASHTO T-104 Magnesium Sulfate Soundness Test on the non-igneous portions and as modified by the MnDOT Laboratory Manual (MNDOT 2005). Limestone aggregates not meeting these requirements are not recommended in vehicular applications. Class C and D aggregates may be used in areas subject only to pedestrian traffic.
Underdrains install quickly when placed on or in the soil subgrade, surrounded by stone base or subbase materials. The outflow portion at the end is not perforated and is raised to a designed height allow for some water detention prior to outflow. Placement at this elevation also protects the pipe with aggregate during base compaction. An underdrain(s) can also be installed and capped at a downstream structure as an option for future use if maintenance observations indicate a reduction in the soil permeability.
Some key actions help ensure the long-term performance of permeable pavement. The most frequently cited maintenance problem is surface clogging caused by organic matter and sediment, which can be reduced by the following measures:
Permeable pavement material specifications vary according to the specific pavement product selected. Table X.5 describes general material specifications for the components installed beneath the permeable pavement. Note that the size of stone materials used in the reservoir and filter layers may differ depending whether the system is pervious concrete, porous asphalt or permeable interlocking pavers. A general comparison of different permeable pavements is provided in Table X.6. Designers should consult industry association and manufacturer’s technical specifications for specific criteria and guidance.
A detailed geotechnical investigation may be required for any kind of stormwater design in karst terrain. Permeable pavements are not recommended at sites with known karst features as they can cause the formation of sinkholes. In addition, permeable pavements are not recommended in active karst areas and require a minimum vertical separation of 50 ft between the bottom of the pavement base and the top of the karst formation.
Plowed snow piles should be located in adjacent grassy areas so that sediments and pollutants in snowmelt are partially treated before they reach all permeable pavements. Sand is not recommended for winter traction over permeable pavements. If sand is applied, it must be removed with vacuum cleaning in the spring. Traction can be accomplished on PICP using jointing stone materials, some of which will find its way into the joints by springtime. Deicing material use on permeable pavement is not recommended to minimize build up in soil and outflows. A significant winter advantage or permeable pavements is that they require less deicing materials than their impervious counterparts.