Guidance for Usage of Permeable Pavement at Airports (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

28 4.1 Overview The design of permeable pavements must take into consideration both hydrologic and structural requirements, as discussed in the following sections. In roadway systems subjected to vehicular loading, structural pavement design is typically more than adequate for the layer thicknesses determined in hydrologic design. However, at their upper end, aircraft weights are significantly greater than those of vehicles, so structural design can play a greater role in the required layer thicknesses. 4.2 Hydrologic Design 4.2.1 Hydrologic Design Overview Permeable pavement systems are designed to meet a variety of stormwater management goals, including reductions in runoff volume and peak discharge rates and improvements in water quality. The hydrologic design process ensures that the proposed permeable pavement cross- section is hydrologically adequate to meet these goals in light of site-specific characteristics and constraints. In the airport context, this is of particular relevance due to safety and wildlife hazard management concerns. To minimize the attraction of wildlife, stormwater facilities at airports are commonly designed to limit open water. Airport drainage design, as directed through the FAA AC 150/5320-5D, aims to safely and efficiently remove water from airport premises, to aid in safe travel on runways and other surfaces, and to discourage waterfowl and other wildlife. The FAA states, “one of the fundamental objectives of stormwater management is to maintain the peak runoff rate from a developing area at or below the predevelopment rate to control flooding, soil erosion, sedimentation, and pollution” (FAA 2013). Therefore, permeable pavement systems at airports should be designed to meet these hydrologic objectives. A permeable pavement system can be modeled as a water balance of stormwater sources and destinations (Figure 12). Water balance variables, as follows, are used to quantify permeable pavement hydrology: • Precipitation – the amount of precipitation falling onto a permeable pavement surface. • Surface run-on – runoff that flows onto a permeable pavement surface from adjacent areas. • Surface runoff – water that flows off the permeable pavement surface and does not infiltrate into the base/subbase layers; runoff can occur if the rainfall intensity exceeds the surface infiltration rate, if the pavement base/subbase becomes saturated, or if the pavement surface geometry allows for a direct runoff flow path. • Surface infiltration – passage of direct precipitation or surface run-on through the permeable pavement surface layer. C h a p t e r 4 Design Considerations

Design Considerations 29 • Subgrade infiltration – water that exits the subbase of a permeable pavement system and enters the soil subgrade. • Underdrain outflow – water exiting via underdrains. In most cases, evaporation is not considered a significant water balance variable in the airport context. However, evaporation is relevant for pavement systems in arid climates and for vegetated grid pavements (ASCE 2015). If the site is determined to be suitable for permeable pavement, as discussed in Chapter 3, the hydrologic design considerations discussed in the following may be used to guide permeable pavement design. The hydrologic design process is consistent among different types of perme- able pavement systems (e.g., porous asphalt, pervious concrete, and permeable interlocking concrete pavers). Hydrologic design of permeable pavement systems relies on: • Existing conditions. • Design storms and rainfall depth. • Run-on from surrounding areas. • Infiltration rates of the soil subgrade. • Outflow configuration. • Base/subbase reservoir thickness and storage capacity. The hydrologic design of permeable pavements for airport-specific applications must ensure that the pavement system complies with FAA regulations for stormwater management facili- ties. These regulations address safety concerns that arise from standing water on the surface of the pavement. To address these concerns, pavement systems must be designed to infiltrate and store stormwater at a rate that prevents water from pooling. As a result, the hydrologic design for airport locations incorporates a specific design storm, dewatering time, and overflow conveyance. Other design requirements and steps for permeable pavement systems at airports follow a more generic hydrologic design process. Pavement systems must temporarily store the design storm volume, which is dependent on the design storm and the size of the surrounding area con- tributing surface run-on. The base/subbase reservoir thickness is designed to provide adequate storage of this volume, considering the infiltration rate of the existing subgrade and system outflow configuration (ASCE 2015). Figure 12. Water balance variables for permeable pavement design.

30 Guidance for Usage of Permeable Pavement at Airports To meet stormwater management goals, permeable pavement systems should be designed to encourage the processes of filtration, detention, and subgrade infiltration to reduce peak runoff rates, improve water quality, and promote groundwater recharge. Systems can be designed to infiltrate all stormwater (full-infiltration or retention), some stormwater (partial-infiltration), or no stormwater (no-infiltration or detention) into the underlying subgrade (ASCE 2015). While no-infiltrating designs do not reduce stormwater volume, benefits still include reduced peak discharge rates and improved water quality. The hydrologic design of permeable pavement systems involves an iterative process influenced by numerous variables (Figure 13). Design begins by characterizing the site to determine any site constraints. The design storm volume is then calculated after selection of the design storm and calculation of surface run-on volumes from surrounding areas. Together with subgrade infiltration rates and an initial outflow configuration design, the thickness of the initial base/subbase reservoir system is determined. The storage and runoff of this initial design are evaluated for the design storm given the materials of the cross-section. The initial design may be adjusted by changing the outflow configuration and base/subbase thicknesses until the pavement system meets design goals and complies with site constraints (ASCE 2015). Soil conditions that should be identified during a field analysis include: • Soil logs at least 3 ft below the bottom of the base/subbase. • Test pits at least 5 ft deep every 7,000 ft2 of paving (minimum two per site). • USCS classification of soil types (using test method ASTM D2487). • Soil infiltration permeability tests (recommended test method ASTM D5093). • Evidence of impermeable soil layers. • Evidence of bedrock/glacial till. • Soil testing for hazardous waste or other contaminants (ASCE 2015). Adapted from Smith (2015) courtesy of ICPI. Figure 13. Hydrologic design steps for permeable pavement systems.

Design Considerations 31 4.2.2 Design Storm and Rainfall Rainfall events, along with watershed characteristics, determine the runoff flows upon which permeable pavement design is based. Different sizes of storms, classified by their return intervals, will have different rainfall characteristics, and therefore runoff flows (Leming et al. 2007). This section discusses how to determine precipitation characteristics of the selected design storm. The required design storm for drainage features at airports is addressed in Chapter 2, Sec- tion 2-2.4 of FAA AC 150/5320-5D, Airport Drainage Design (FAA 2013). The recommended minimum design storms are summarized in Table 5. The FAA recommends a 5-year storm for airfield pavements because the increased cost of a drainage system to accommodate a larger storm may be greater than the damage or inconve- nience caused by the larger storm (FAA 2013). However, AC 150/5320-5D further notes that to reduce the likelihood of flooding a facility essential to operations, and to prevent loss of life, some portions of drainage system designs have been based on storm events of as high as 50 years. In addition to the minimum design storms, AC 150/5320-5D also indicates that the center 50% of runways and taxiways serving those runways and helipad surfaces along the centerline should be free from ponding resulting from a 10-year storm event frequency and intensity (FAA 2013). While Table 5 provides FAA recommended minimums, local requirements may be more stringent. As seen in the case studies discussed in Chapter 3, local regulations required the Culpeper stormwater storage capacity be designed to maintain peak runoff below pre-construction levels for a 10-year storm. The Paine Field pervious concrete project was designed to a 100-year storm event to meet local requirements. However, funding through the FAA is generally limited to that required to meet FAA standards. Eligibility for federal funds for storm drainage in excess of FAA standards will need to be evaluated by the FAA’s Airports Financial Assistance Division. After selecting the appropriate design storm, designers should identify several aspects of pre- cipitation characteristics used in hydrologic design calculations. Intensity–duration–frequency (IDF) curves can be used to find the intensity (inches/hour) of the design storm. The IDF curve provides a summary of a site’s rainfall characteristics by relating storm duration and exceedance probability (frequency) to rainfall intensity (FHWA 2013). IDF curves can be obtained for most states from the National Oceanic and Atmospheric Administration (NOAA) Atlas 14 through NOAA’s Precipitation Frequency Data Server (http://hdsc.nws.noaa.gov/hdsc/pfds/). If needed, the design storm intensity can be converted to a maximum depth of expected rain- fall (inches) by multiplying by the storm duration. This information would be used to calculate the volume of runoff or peak flow to be captured, infiltrated, and/or released by the permeable pavement system from the design storm (ASCE 2015). 4.2.3 Surface Run-on Capture Permeable pavement systems are often designed to accept surface run-on from existing adjacent areas. These areas may be either impervious or pervious as long as the surface run-on is fairly free Application Minimum Storm Event Department of Defense airfields and heliports 2 year FAA facilities 5 year Areas other than airfields 10 year Table 5. Summary of FAA minimum design storms.

32 Guidance for Usage of permeable pavement at airports of sediments and contaminants. Surface run-on from adjacent surfaces should not be allowed if the area is under construction, contains unstable soils, is used for snow storage, or contains mulch or leaf debris from landscaping (ASCE 2015). These conditions may increase the potential for clogging and increase maintenance requirements. Pretreatment of stormwater run-on to the permeable pavement system is rarely required (Virginia Department of Conservation and Recreation 2013). In some cases, a pretreatment gravel filter strip may be required if the sediment load of an impervious area draining to the permeable pavement is high. To avoid potential clogging, the pavement may be designed so that adjacent surface run-on is discharged directly to the reservoir layer (e.g., run-on from roof drains). Peak run-on volumes should be calculated using the NRCS Technical Release 55 (TR-55) method (NRCS 1986). It is not recommended to use the Rational Formula (the most common method used for sizing sewer systems) to estimate run-on to permeable pavements since it is a simplistic approach using reference list runoff coefficients, which can lead to both over- and under-prediction of flows (ASCE 2015). The NRCS Technical Release 55: Urban Hydrology for Small Watersheds peak flow method calculates peak flow as a function of drainage area, potential watershed storage, and the time of concentration (NRCS 1986). This method is appropriate for permeable pavement design because of the following: • It captures the essential elements of permeable pavement system behavior. • It is appropriate for the design of a structure intended to capture and hold some portion of the runoff in small urban watersheds. • It is flexible and easily adapted to a site with several types of surfaces contributing to runoff. • It is implemented by adapting well-known stage storage-discharge principles to the simple geometry of a permeable pavement system. • It can be used to analyze systems intended to function within the constraints of many different regulatory requirements (Leming et al. 2007). The NRCS TR-55 method is outlined in Chapter 2, Section 2-3.2 of FAA AC 150/5320-5D, Airport Drainage Design. Example calculations are also provided. An easy-to-use graphical approach to this method can also be found in the TR-55 publication (NRCS 1986). In addition, there are several TR-55 hydrology programs compatible with Windows operating systems that implement NRCS methods for calculating time of concentration, peak flows, hydrographs, and detention basin storage volumes. 4.2.4 Infiltration Rates In a full- or partial-infiltration design (see Section 4.2.5, System Outflow Configuration), the subgrade infiltration rate of the permeable pavement system will help maintain the effective storage capacity of the permeable pavement system by removing some of the rainfall over time (ASCE 2015). The effect of infiltration on storage capacity and, therefore, excess surface runoff, is a critical element in design (Leming et al. 2007). For infiltration testing, field tests are preferred over laboratory tests because they are more reflective of site conditions. The tests should be completed at the elevation for which natural soil subgrade infiltration is being proposed. Individual test results should not be considered absolute values for infiltration rates but should be interpreted with soil texture and structure (ASCE 2015). Because predicting sediment loading of the soil subgrade is difficult, a conservative infiltra- tion reduction factor of 0.5 (safety factor of 2) should be applied to the average soil infiltration rate measured on site (ASCE 2015). This ensures that compaction of the soil subgrade during

Design Considerations 33 construction and the resultant reduction in the infiltration rate are considered in the design process. A higher safety factor may be appropriate for sites where samples show highly variable infiltration rates (ASCE 2015). Most designs assume that water infiltration into the soil subgrade occurs uniformly across the bottom of the permeable pavement as the base/subbase reservoir becomes saturated (ASCE 2015). The simplest approach uses Darcy’s Law, which assumes a constant rate of infiltration into a saturated subgrade:  Q k h A= where Q = rate of flow (ft3/h), k = coefficient of permeability (ft/h), h = hydraulic gradient, and A = area of flow (ft2). Since the water table is typically some distance below the base/subbase reservoir layer, the hydraulic gradient can be assumed to be 1.0 as the drop in elevation causes downward flow (ASCE 2015). The water flux (flow per unit area) is then equal to the measured infiltration rate and gives the depth of water that will be infiltrated into the subgrade over a specific time. This method provides an appropriate estimation for most designs where permeable pavement systems are modeled as an infiltration basin with a constant rate of infiltration (ASCE 2015). In reality, the infiltration rate into the subgrade will vary (ASCE 2015). As the water depth in the base/subbase reservoir increases, the static pressure also increases, encouraging infiltration. Moisture conditions of the underlying subgrade also affect infiltration rates. Physics-based models based on the Green-Ampt or Richard’s equation provide a more accurate representation of varying infiltration rates and may also be used in permeable pavement design. However, these methods are more complex and require more detailed information about subgrade properties (ASCE 2015). 4.2.5 System Outflow Configuration There are several design options for stormwater discharge for permeable pavement systems. After infiltration through the base/subbase reservoir, stormwater can either infiltrate into the underlying subgrade or be directed to be discharged into a piped drainage system through an underdrain (ASCE 2015). FAA regulations require that drainage systems be capable of draining 85% of the design storm volume within 24 h if they are used on airfield runways and taxiways. The drainage system should be capable of draining 85% of the design storm volume within 10 days for airfield parking aprons and other pavement areas receiving only low-volume, low-speed traffic (FAA 2013). Drainage systems should be designed so that no runoff from the design storm encroaches onto taxiway and runway pavements, including paved shoulders (FAA 2013). Designers should ensure that the total system outflow rate, including overflow, underdrain dis- charge, and subgrade infiltration, meets these dewatering time requirements. The following relationship between required system storage and required dewatering time can be used to calculate the required outflow rate (ASCE 2015): Total system storage volume Total outflow rate volume/time Dewatering time time( ) ( ) ( )= If the infiltration rate of the underlying subgrade is not sufficient to meet dewatering times [as is often the case with C or D soils (from the hydrologic soil groups of the Natural Resources Conservation Service)], the design of a permeable pavement system may need to include an

34 Guidance for Usage of permeable pavement at airports underdrain. A perforated underdrain pipe may also be required for sites where stormwater infiltration into underlying soil is limited or prohibited, such as those with high groundwater elevations, with shallow bedrock, or where natural soils are contaminated or have low perme- ability (ASCE 2015). Underdrains are located in the base/subbase reservoir and are typically perforated polyvinyl chloride (PVC) pipes that are 4 to 6 in. in diameter. The Virginia Department of Environmental Quality Stormwater Design Specification No. 7 provides additional information on underdrain design (Virginia Department of Environmental Quality 2011). Table 6 compares the differences in outflow configurations between full-infiltration, partial-infiltration, and no-infiltration designs ASPHALT PAVERCONCRETE ASPHALT PAVERCONCRETE ASPHALT PAVERCONCRETE ASPHALT PAVERCONCRETE Full-Infiltration Designs • Do not use underdrains. • Infiltrate all stormwater into soil subgrade. • Used in areas with high- permeability native sandy soils. Partial-Infiltration Designs • Use perforated underdrains. • Infiltrate some stormwater into soil subgrade and discharge some via underdrain. • Used in areas with lower- permeability native soils. • May include upturned elbow or other flow restriction device to increase temporary storage and promote infiltration by ponding water in the reservoir. No-Infiltration Designs • Use perforated underdrains. • Discharge all stormwater via underdrain. • Prevent infiltration with impermeable liner (geosynthetic liner or clay barrier). • Used in areas with low- permeability native soils or areas with soil or groundwater contamination. Table 6. Permeable pavement outflow configurations.

Design Considerations 35 (ASCE 2015). Although not shown, an upper underdrain may be needed to make sure the water does not back up into the permeable pavement layer (overflow conveyance), especially in areas susceptible to freezing. The physical configuration/elevations of underdrains affect the outflow rates and storage perfor- mance in permeable pavement systems. The following are the three most common underdrain/ outflow configuration designs with systems using perforated underdrain discharges: 1. Perforated pipe placed at elevation above frequent water level in reservoir storage. The water level in the storage area during certain storm events is below the perforated pipe elevation with no discharge. This might be encountered in certain storm events if the reservoir is designed as an infiltration system and the permeability of the underlying soils is rapid, or the pipe is placed at a higher elevation to encourage a greater depth or ponding/greater head below the pipe to encourage higher rates of infiltration to the existing soils. 2. Water is ponded above the perforated pipe while discharging. Flow is determined by the head of water above the pipe and the size of the pipe and its performance. In this condition, the discharge rate can be manipulated by applying well-documented principles of orifice and pipe hydraulics to the perforated discharge pipe. In most cases, the outflow rate from the underdrain will exceed system inflows. If desired, the underdrain can be fitted with a small orifice to control extended detention rates. 3. The perforated pipe’s capacity is large enough to allow free flow without limiting the discharge out of the reservoir. Instead, discharge is limited by the lateral flow rate through the reservoir aggregate; the pipe and its perforations carry water away as fast as the reservoir delivers it. In this condition, the discharge rate is determined by the storage area’s ponding depth, porosity, and hydraulic conductivity. It can be manipulated by controlling the number of pipes and the reservoir’s hydraulics (ASCE 2015). The underdrain should also be at an elevation to prevent water from freezing in the aggregate storage bed due to the frost depth. The University of New Hampshire Stormwater Center recom- mends a minimum bottom of reservoir depth of 65% of the frost depth locally observed, with the underdrain then placed 4 in. above the bottom (University of New Hampshire Stormwater Center 2014). 4.2.6 Base/Subbase Reservoir Thickness and Storage Capacity The base/subbase depth is related to the required runoff volume that needs to be temporarily stored within the permeable pavement system. Storage in permeable pavement systems includes the void space in base/subbase layers, which is replenished through infiltration of water into the subgrade. It is important to note that the structural design process for permeable pavement will also determine a base/subbase material thickness required to support loads on the pavement. If the thickness requirements for structural and hydrological needs differ, the thicker of the two designs should be selected (ASCE 2015). As shown in Figure 14, permeable pavements designed on a slope will have lower storage capacity than flat permeable pavements with the same base/subbase design. Leming et al. (2007) discuss how to adjust storage capacity calculations for sloped permeable pavement designs. For sloped designs, subsurface terracing, check dams, baffles, or berms may be incorporated into the permeable pavement design to encourage vertical infiltration and reduce potential for lateral flow onto the surface, as seen in Figure 15 (ASCE 2015). Storage can be calculated as a static volume for more conservative assessments or for worst- case scenarios by neglecting infiltration. In the case of permeable interlocking concrete pavers, the total storage capacity of the permeable pavement system should not include the capacity

36 Guidance for Usage of permeable pavement at airports Figure 14. Difference in storage capacity for permeable pavement designs on flat versus sloped surfaces. Figure 15. Sloped permeable pavement design with check dams, baffles, or berms. within the surfacing and the bedding layer (ASCE 2015). Assuming static conditions, the effec- tive depth of storage can be calculated as: dr dp r= η where dr = depth of runoff stored (ft), dp = depth of the reservoir layer (ft), and hr = effective porosity of the reservoir layer.

Design Considerations 37 The net storage of a permeable pavement system is dynamic and should include the amount of water that leaves the system during a storm through infiltration into the subgrade. Storage in permeable pavement is a function of the surface run-on and precipitation rates, available void space, depth of the base/subbase materials, any runoff that has accumulated from previous rainfall, the subgrade infiltration rate, and discharge through underdrain pipes and overflows (ASCE 2015). Storage of permeable pavement systems can be calculated using storage routing and follows a basic water balance equation: storage inflow outflow∆ = − Computational methods and routing techniques, such as dynamic storage-indication routing, are used to address the storage and hydraulic processes in the system (ASCE 2015). These methods, for the most part, can be applied to permeable pavement design, as long as site-specific conditions are modeled. One approach is to model the permeable pavement system as a detention or recharge system with modest infiltration or controlled outlets for discharge. The major point of difference is the presence of material in the storage reservoirs and the various layers of permeable pavement designs. Some modeling tools have incorporated open storage chambers within an aggregate bed (HydroCAD, n.d.) for additional water storage (ASCE 2015). 4.2.7 Overflow Conveyance All permeable pavement systems must be designed to safely convey overflows. Overflows may occur when flow exceeds the design storm capacity or when the rate of rainfall exceeds the rate of infiltration into the subgrade (ASCE 2015). Even though full-infiltration systems are designed to infiltrate the entire runoff volume, overflow conveyance must still be included in the design. Designs must take into consideration the flow path from the point of overflow, including downstream receiving areas and potential impacts. Overflows may be directed to surface con- veyance channels or closed drainage systems (ASCE 2015). The North Carolina Department of Environment and Natural Resources (NCDENR) Stormwater BMP Manual, Section 18: Permeable Pavement includes detailed drawings of a variety of configurations for outlet control and bypass control structures (NCDENR 2007). 4.2.8 Water Quality Considerations Water quality control using permeable pavement depends on two pollutant removal pathways: infiltration and filtration (ASCE 2015). The effectiveness of permeable pavement for water quality improvements depends on the depth of the base/subbase layers, material properties, native soils, and quality of inflow. The base/subbase materials in permeable pavement improve stormwater quality through the removal of heavy metals, oil/grease, total suspended solids, and some nutrients (ASCE 2015). Soluble solids, such as in deicing fluids, will not be treated (ASCE 2015). Filter courses of poor- graded sand may be included in the base/subbase layer for additional water quality treatment, as seen in Figure 16. Permeable pavement systems with a filter course can provide a very high level of filtration prior to infiltration and provide exceptional water quality treatment. Filter course material and depth are based on targeted pollutants for removal. A choker course may be required below the filter course to prevent the filter course material from moving to the reservoir course below (ASCE 2015). There is currently no single recommended method for water quality modeling of permeable pavement given the complexity of the variables (ASCE 2015). Continuous simulation models can be used to estimate pollutant loads based on historical precipitation records and to help assess

38 Guidance for Usage of permeable pavement at airports the performance of permeable pavement within a watershed (ASCE 2015). The Environmental Protection Agency (EPA) Best Management Practice (BMP) Performance Curves provide some guidance on the pollutant removal rates of permeable pavement using these methods (EPA 2010). Permeable pavement design must take into consideration the initial stormwater runoff that will carry the highest concentration of pollutants (called the first flush). In more arid areas, with long periods between rains, a seasonal first flush may need to be considered. Permeable pavement helps to meet one of the common goals of mitigation, which is to capture and treat the first flush of runoff (Leming 2007). Permeable pavements may also be designed to detain water, which can assist in nutrient reduction. This approach is more applicable in low-infiltration rate soils, which can also capture metals. Besides detention, which encourages denitrification, additional nutrient treatment can be realized with discharge to other stormwater management BMPs (ASCE 2015). 4.2.9 Additional Hydrologic Design Considerations 4.2.9.1 Percent Imperviousness Permeable pavement can be used to reduce the percent imperviousness of a site to meet design goals or requirements. The percent imperviousness of permeable pavement systems depends on outflow configurations, base/subbase depth, and subgrade infiltration rate. The NCDENR assigns this percentage based on the hydrologic soil group of the subgrade and whether the system meets water quantity and pollutant removal requirements. 4.2.9.2 Intersection of Permeable Pavement with Conventional Pavement A variety of design approaches and materials are used for the intersection (or tie-in) between permeable pavement and conventional pavement. It is important that the sides of the reservoir Figure 16. Permeable pavement system with filter course for additional water quality treatment.

Design Considerations 39 course be lined with an impermeable liner or barrier to prevent stormwater from entering the sub- base of the conventional pavement. It is recommended that a detail be prepared to specify the construction process at the pavement intersection between two pavement types. For maintenance purposes, a visual delineation between pavement types may be beneficial (ASCE 2015). 4.2.9.3 Drainage Structures Beyond underdrain pipes that may be used for outflow or overflow conveyance, other drain- age structures are necessary. For example, cleanouts will be necessary for underdrain systems. Cleanouts provide surface access points to the underdrains for the purpose of the removal of sediments that might accumulate in the underdrain pipe. Underdrains can be connected to a manhole structure or outlet structure. Section 18: Permeable Pavements of the Stormwater BMP Manual includes different solutions for the outlet control (NCDENR 2012). Large-scale permeable pavement systems require the installation of one or more observation wells and cleanouts. The observation well should be capped and positioned at the lower end of the structure to monitor the time necessary for the base course to fully drain between subsequent storm events (ASCE 2015). 4.3 Structural Design 4.3.1 Structural Design Overview A structural design method specific to permeable pavements has not yet been developed. Instead, the structural design of permeable pavements has been accomplished using methods established for conventional pavements, such as design procedures from AASHTO or, to a lesser extent, the FAA. The Culpeper apron and Richmond taxiway case study projects assessed struc- tural requirements using both AASHTO (AASHTO 1993) and FAA (FAA 1995 and FAA 2009) design methods. Each of these design procedures has the ability to analyze pavement structures containing permeable layers. However, the models behind the procedures are not calibrated with any data on in-place permeable pavements. Although there are other industry guides (e.g., from the Portland Cement Association and the Asphalt Institute [AI]) for the structural design of airport pavements, the primary guidance for airport pavement structural design for commercial and general aviation airports is the FAA’s AC 150/5320-6F, Airport Pavement Design and Evaluation (FAA 2016). This AC is accompanied by the FAA’s FAARFIELD pavement design software. The design methods incorporated in the design procedure include layered elastic analysis (LEA) for flexible pavements and a finite-element method (FEM) for rigid pavements. The FAA design procedure uses the anticipated aircraft traffic mix, along with layer modulus and strength parameters, to characterize pavement response to loading. Structural design for permeable pavements solely carrying roadway vehicles can be performed using design procedures developed by AASHTO, individual state highway agencies, or pavement industry organizations. AASHTO procedures include both the long-used empirical approach and the newer mechanistic–empirical approach: • Guide for Design of Pavement Structures (AASHTO 1993): Uses 18,000-lb equivalent single-axle loads (ESALs) to define traffic loadings, layer structural coefficients (ai) to characterize the strength/stiffness of individual pavement layers, and certain modulus parameters to charac- terize the load-bearing capacity of the subgrade soil. • Mechanistic–Empirical Pavement Design Guide Manual of Practice (AASHTO 2015): Uses an array of vehicle classes and loading characteristics to define loadings, along with modulus, strength, and mixture parameters, to characterize the response of individual pavement layers and the subgrade soil when subjected to loading.

40 Guidance for Usage of permeable pavement at airports The American Concrete Pavement Association (ACPA) has developed combination software (PerviousPave) for conducting structural and hydrological design for highways and streets incor- porating pervious concrete (ACPA, n.d.). PerviousPave is based on the organization’s thickness design methodology for jointed plain concrete pavements, StreetPave, and uses a modified version of Los Angeles County’s hydrological design method. The program is capable of determining (1) the required minimum pervious concrete pavement thickness based on the design traffic, design life, and other structural inputs, and (2) the required subbase/reservoir thickness necessary to satisfy stormwater management requirements based on volume of water to be processed by the pavement within the required maximum detention time. AASHTO design methods can be applied to full-strength pavements for aircraft in some limited circumstances. For nonprimary airports, a sponsor can request the use of state standards that are different from FAA specifications and state highway construction and material specifications for full-strength airfield pavements. 49 USC 47105(c) and 49 USC 47114(d)(5) give the FAA the authority to approve state standards and the use of state highway specifications. The use of this method is limited to the design of pavements at airports with runways 5,000 ft long or less and serving aircraft weighing 60,000 lbs or less. The requirements for using state standards are contained in FAA Order 5100.38D, Airport Improvement Program Handbook (FAA 2014c) and AC 150/5100-13A. An MOS would need to be submitted to and approved by the FAA in accordance with FAA Order 5300.1 (FAA 2014c). Brief descriptions of the AASHTO 1993 and FAA pavement design procedures are provided in the following. The layer properties of permeable pavements for use in these structural design methods are discussed in the following sections. 4.3.1.1 FAA Thickness Design As mentioned previously, FAARFIELD uses LEA for flexible pavement design and FEM for rigid pavement design. For flexible pavements, rutting in the subgrade is the primary analyzed failure mode. The process is based on determining the vertical compressive strain on the subgrade using the layered elastic subprogram, LEAF, and correlating that response to failure through a performance model. FAARFIELD also allows the option of evaluating the horizontal tensile strain at the bottom of the HMA surface layer to investigate the potential for fatigue cracking of the HMA surface, but fatigue is not the primary design criteria. Rigid pavement design in FAARFIELD is based on bottom-up cracking of the slab. The hori- zontal tensile stress at the bottom of the slab is determined, and that response is correlated to failure through a performance model. The following inputs are required for new pavement design using FAARFIELD: • Design period. • Traffic data. • Subgrade support. • Layer types and characteristics. In addition to these inputs, flexural strength is required for rigid pavement designs. The pavement cross-section is input in the “Structure” screen of FAARFIELD, using appro- priate layers for the design being considered. Many of the default layer type characteristics are fixed within the program. Because most of the default materials within FAARFIELD have set moduli, the use of user-defined and “variable” layers is needed to model permeable pavements. The FAA’s design procedure also has minimum layer requirements. However, these are based on the default material types (i.e., HMA, PCC, and aggregates) meeting the specifications in FAA AC 150/5370-10G. Therefore, minimum thicknesses for permeable

Design Considerations 41 pavements may be different. FAARFIELD performance models are also based on observed performance of conventional pavements, which may not be accurate for permeable pave- ment materials. FAA AC 150/5100-13A, Method A, provides guidance for equating FAARFIELD-determined thicknesses based on P-specification materials to thicknesses using state highway materials. In general, additional thickness is added to the FAARFIELD-determined layer thickness. However, these thickness adjustments are based on conventional state highway materials. Greater thicknesses would likely be required for permeable materials. 4.3.1.2 AASHTO 1993 Thickness Design The AASHTO 1993 pavement design method is based on the use of nomographs to determine a structural number (SN) for flexible pavement and concrete thickness (D) for rigid pavement. The SN is an abstract index representing the pavement strength based on soil support, traffic loadings, serviceability, and environment (AASHTO 1993). To facilitate this empirical design procedure, AASHTO developed the computerized design program known as DARWin. The AASHTO general design equation for flexible pavements requires the following inputs: • Expected 18-kip ESALs (W18). • SN. • Standard normal deviate for selected reliability level (ZR). • Overall standard deviation (SO). • Change in 0-to-5 scale present serviceability index (DPSI). • Subgrade resilient modulus (MR; psi). SN is determined using the following equation: SN a D a D m a D m1 1 2 2 2 3 3 3= + + where ai = structural layer coefficient for layer i, Di = thickness of layer i (inches), and mi = layer drainage coefficient. The layer coefficient (ai) is the empirical recognition of the ability of a layer to function as a structural component. In general terms, a higher ai indicates a stronger pavement layer. The drainage coefficient (mi) represents an assessment of how long a layer remains saturated. A higher value represents a layer that drains well, while a lower value indicates a poorly draining layer. The AASHTO general design equation for rigid pavements requires the following inputs: • Expected 18-kip ESALs (W18). • Concrete thickness (D; in.) • Standard normal deviate for selected reliability level (ZR). • Overall standard deviation (SO). • Change in 0-to-5 scale present serviceability index (DPSI). • Terminal serviceability index (pt). • Concrete modulus of rupture (S′c; psi). • Drainage coefficient (cd). • Load transfer coefficient (J). • Concrete elastic modulus (Ec; psi). • Modulus of subgrade support (k; psi/in.) (AASHTO 1993).

42 Guidance for Usage of permeable pavement at airports For vehicular design, state highway agencies and AASHTO have established required values for inputs depending on anticipated vehicular traffic. FAA AC 150/5100-13A, Method B, provides guidance and inputs for the use of the AASHTO 1993 thickness design method. For flexible pavement design, minimum SNs are required based on the aircraft weight and subgrade support. Minimum concrete thicknesses are provided for rigid pavement design. However, these design requirements are based on conventional materials. 4.3.2 Traffic Data As mentioned previously, the AASHTO 1993 pavement design method uses ESALs to character- ize traffic for design. For parking lots and roadways, traffic data can be obtained and conversions made for the axle types to reach a number of design ESALs. FAA pavement design uses the aircraft type, including weight, gear type, tire pressure, and volume, for each aircraft in the traffic mix. The FAARFIELD design software has an extensive aircraft library that contains most of the common aircraft models in use as well as generic gear configurations. Aircraft gear configurations are quite different from vehicle axle configurations, and there is no straightforward method for conversion of aircraft loadings to ESALs to use the AASHTO 1993 design procedure. As part of the Airport Asphalt Pavement Technology Program (AAPTP) report, Guidelines for Use of Highway Specifications for HMA Airport Pavements, two categories are recommended for ESAL and aircraft gross weight correlation (Buncher and Boyer 2009), sum- marized in Table 7. The categories in that study are based on the correlation developed as part of a 2008 AAPTP report, PG Binder Grade Selection for Airfield Pavements (Christensen et al. 2008). 4.3.3 Subgrade Characterization In order to promote as much infiltration as possible, full-infiltration designs for vehicular applications indicate not compacting the subgrade. However, under the heavier loads of aircraft, not compacting the subgrade may lead to additional settlement. FAA pavement design indicates that subgrade compaction is needed to avoid settlement from loadings. Compaction can reduce the infiltration rate of the soil. Therefore, design must take into consideration both compaction requirements and the impact on infiltration. Subgrade strength is often characterized using California Bearing Ratio (CBR) testing. FAA flexible pavement design indicates the use of saturated CBR results because subgrades gen- erally reach saturation over time or from seasonal moisture variability (FAA 2016). Rigid pavements are designed using modulus of subgrade support (k-value, k). However, plate load testing to determine k-value is costly and time consuming. Therefore, correlations to labora- tory CBR testing are generally used. CBR, k-value, or elastic modulus (Esg) can be entered in the FAARFIELD program, and FAARFIELD will automatically calculate the other parameters using the following correlations:   E CBR E k sg sg 1500 20.15 1.284 = = Aircraft Weight (lbs) ESALs (millions) Less than 12,500 Less than 0.3 12,500 to less than 60,000 0.3 to less than 3.0 Source: FAA (2011a). Table 7. Correlation of aircraft weight to ESALs.

Design Considerations 43 4.3.4 Paving Layer Properties 4.3.4.1 Base/Subbase Reservoir Aggregate The reservoir aggregate is an open-graded or uniform aggregate with high permeability properties. The high permeability of the layer results in a lower layer modulus (and structural number) for the layer. NAPA recommends a structural coefficient of 0.10 to 0.14 for the reservoir layer (Hansen 2008). In the recommendations for using the AASHTO 1993 design, the FAA guidance indicates a structural coefficient of 0.14 and 0.11 for the base and subbase, respectively (AASHTO 1993). These values would be representative of a P-209 crushed aggregate base course and P-154 sub- base course, respectively. The recommended values from NAPA for the reservoir layer generally range between these two materials. The AASHTO 1993 correlation between the granular base layer structural coefficient and layer modulus (EBS) is as follows: a EBS0.249 log 0.9772 ( )= − Similarly, the granular subbase layer and modulus (ESB) correlation is as follows: a ESB0.227 log 0.8393 ( )= − These equations and suggested structural coefficients indicate moduli ranging from around 13,000 to 30,000 psi for a reservoir layer. When a default aggregate layer type (such as P-209 or P-154) is selected in FAARFIELD, the modulus is internally calculated, but the initial moduli for P-209 and P-154 layers are 75,000 and 40,000 psi, respectively. Based on structural coefficients and these equa- tions, these layers correlated to moduli of around 31,000 and 15,000 psi, respectively. It appears that the AASHTO equations result in a very conservative correlation compared to FAARFIELD. Regard- less, a user-defined layer is needed in FAARFIELD to characterize the reservoir layer. 4.3.4.2 Stabilized Permeable Base Layers Although FAA pavement design requires a stabilized base layer only for pavements supporting aircraft weighing more than 100,000 lbs, it may be beneficial to provide a stabilized base for all pavements supporting aircraft. Stabilized permeable bases can be either cement-treated permeable bases (CTPBs) or asphalt-treated permeable bases (ATPBs). Neither the FAA’s pavement design procedure nor the associated software currently contain stabilized permeable layers. For the AASHTO 1993 flexible design method, NAPA suggests a structural coefficient of 0.30 to 0.35 for ATPBs. For the rigid pavement design, the effective modulus of subgrade support under the surface slab is determined based on the thickness of the base layer and the underlying subgrade k-value. However, the correlations are based on either unbound aggregate or more typical stabilized material. For design in FAARFIELD, a user-defined layer can be used to incorporate a stabilized per- meable base. However, there does not appear to be clear consensus on a layer modulus. The Vermont Department of Transportation (Vermont DOT) recommends an ATPB modulus of 110,500 psi (Pologruto 2004). The Ohio DOT has developed a temperature-based correlation: M T Tr 0.00005 0.011 0.74812= − + where Mr = resilient modulus (Mpsi), and T = °F; along with statewide ATPB temperatures of 48°F (spring, fall), 75°F (summer), and 33°F (winter) (Masada et al. 2004).

44 Guidance for Usage of permeable pavement at airports For this correlation and these temperatures, the ATPB modulus ranges from 204,000 to 439,000 psi. A modulus of 750,000 psi is recommended for CTPB in reference materials (Masada et al. 2004, Colorado DOT 2016), which is even greater than the cement-treated base modulus (700,000 psi) used in FAARFIELD. 4.3.4.3 Porous Asphalt Surface Conventional dense-graded asphalt mixtures have a structural coefficient of 0.44 in the AASHTO 1993 design method. The NAPA design procedure recommends a structural coefficient of 0.40 to 0.42 for porous asphalt (Hansen 2008). The modulus of porous asphalt is temperature dependent, as with conventional HMA. FAA pavement design assumes a higher design temperature (90°F) than that for roadway design (77°F). The HMA surface modulus in FAARFIELD is 200,000 psi, while a value of 400,000 psi is used for an HMA base. The FAA’s HMA base modulus is comparable to the HMA modulus used in roadway design. Therefore, temperature needs to be considered when selecting the porous asphalt modulus for use in FAARFIELD and would require the use of a user-defined layer. Roadway studies where deflection testing and back calculation of layer properties were con- ducted have indicated that porous asphalt has a modulus of approximately half that of conven- tional HMA controls (Hossain and Scofield 1991, Uju 2010). These studies did indicate high modulus values for porous asphalt, ranging from approximately 500,000 to 1,000,000 psi, which are quite high for the material. These results are from pavements that were in use (i.e., aged) but do suggest that porous asphalt may have a similar initial modulus as that of the FAARFIELD default surface layer. 4.3.4.4 Pervious Concrete Surface Pervious concrete properties depend on many factors related to materials selection and mix design. However, typical values of 28-day compressive strength for pervious concrete range from 400 to 4000 psi (Tennis et al. 2004). Relatively high values of compressive strength of pervious concrete can be achieved, but with the reduction of void content and compromising permeability (ACI 2010). The reported values of pervious concrete flexural strength in ACI 522R-10 range from 300 to 600 psi. The relationship between compressive and flexural strength can be expressed as follows: f fr c2.3 2/3= ′ where fr and f ′c represent the flexural and compressive strength of pervious concrete, respectively, in psi (ACI 2010). Slightly lower flexural strength values of 150 to 550 psi are reported on the Pervious Pavement website (http://www.perviouspavement.org/engineering.html). FAARFIELD currently only allows flexural strengths of as low as 500 psi, providing a range of 500 to 900 psi. Therefore, it may not be possible to model a pervious concrete surface in the current FAARFIELD software unless the mix design has a higher flexural strength. Additionally, the concrete modulus in FAARFIELD is fixed at 4,000,000 psi, and pervious concrete can be expected to have a lower elastic modulus.