Bridge Stormwater Runoff Analysis and Treatment Options (2014)

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.

36 Stormwater Treatment Controls for Bridges This chapter describes the treatment BMPs that the practi- tioner can consider when treatment of deck runoff is required by the regulatory agency. Installation of treatment BMPs should be reserved for situations where the DOT is required to provide them as part of an MS4 NPDES permit designa- tion, or pursuant to a Section 401 Water Quality Certifica- tion, and/or based upon an assessment that considers other special receiving water conditions such as TMDLs, presence of endangered species, the protection of outstanding natu- ral resource waters (ONRW) or domestic water supply res- ervoirs. Implementation of treatment controls for bridges should be done in concert with other applicable minimum source and operational control measures discussed in Chap- ter 4. The treatment controls available include BMPs located at the abutment and within the deck area itself. There are many comprehensive references (including NCHRP publica- tions) on BMP selection and design for highways. However, the information provided in this chapter focuses on provid- ing the practitioner with the details needed to understand and evaluate the cost efficiency of the treatment BMPs that are included in the BMP evaluation Tool. These BMPs can be considered for new construction and retrofit situations. Vari- ous types and options for deck drainage conveyance are also discussed as an important consideration in bridge design, cost, long-term maintenance, and as an influencing factor on the stormwater treatment approach. The treatment BMPs discussed in Section 5.3 of this chapter generally require con- veyance of the deck runoff to the abutment for treatment, with several exceptions—one being PFC overlay, and the others involving potentially promising concept technologies. Another relevant and potentially cost effective option involv- ing treatment BMP installation at offsite terrestrial roadway locations is discussed in Section 5.4. Operational and structural spill controls are also discussed in this chapter. Historical evidence has shown that probabil- ity of spilling a hazardous chemical over a sensitive receiv- ing water is remote and is best handled by first-responders to contain the pollution. Despite this fact, the construction of capital improvements to contain a spill may be warranted in certain instances and the guidelines associated to apply such an approach are discussed in Section 5.4 of this chapter. 5.1 Tool Overview The spreadsheet Tool accompanying this guide was devel- oped around five treatment BMPs that are suitable for use at the abutment and one that can be used on the bridge deck: Use at the abutment: • Swales • Dry detention basin • Bioretention • Media filter Use on the bridge deck: • PFC Other BMPs the practitioner can consider are also discussed in this section, but not modeled by the spreadsheet tool, due to lack of cost and performance data. However, they are included in this guide as an option for the practitioner to evaluate. The selection of a BMP will largely be based on the physi- cal constraints in the abutment area. The BMPs modeled by the tool were selected because they have relatively flexible sit- ing criteria, can operate under a variety of conditions, and have been shown to be effective for constituents of concern in highway runoff. Siting and design guidance are provided in the sections that follow. The tool can be used to develop performance and whole life cost data for the selected BMP. The tool is also useful for comparing the BMPs and selecting the one with the highest performance at the lowest life cycle cost. Treating bridge deck runoff at the abutment will require conveying runoff to the abutment. The practitioner can select C H A P T E R 5

37 a conveyance system cost for their bridge in a separate spread- sheet tool (included as a separate file from the BMP evalua- tion tool). Guidance for developing the BMP cost using the tool is provided in Appendix E: BMP Evaluation Tool Model- ing Methodology. The conveyance system may consist of a system of deck drains connected to pipes that convey runoff to the abutment(s). Alternatively, runoff conveyance may be provided by the bridge deck, as long as the spread criteria are maintained for the traveled way. Section 5.2 discusses an alternative design approach of offsetting the deck drains from the bridge rail wall to allow the water quality design flow to pass the deck drain, with excess flow entering the drain. This may be an economical way for some bridge installations to convey runoff to the abutment without a closed conduit pip- ing system by using the planned deck shoulder area for con- veyance of the water quality flow. The tool accompanying this guide can be used to estimate the influent concentration and load for many of the con- stituents of concern, as well as the effluent quality from the selected BMPs. The tool computes influent and effluent qual- ity for the constituents listed in Table 5-1. For those constituents not listed, proxies may be used to estimate effluent concentrations. For example, total metal removal may be similar for specific metals not computed by the tool. If the influent concentration of the metal is known, the effluent quality may be estimated from an average of reductions observed for the listed metals as computed by the tool. A similar approach can be used for other dissolved and particulate constituents. 5.2 Bridge Deck Conveyance Systems The purpose of bridge deck conveyance systems in the context of this study is to transfer deck runoff to the abut- ment for treatment. This can be accomplished using a wider deck area that includes a shoulder to convey runoff to the abutment, or one with a comparatively narrower deck that incorporates a piping system serving bridge deck drains. This section will discuss the technical merits of each approach, the challenges of each approach, as well as operation and main- tenance considerations. The maximum spread allowed in the shoulder of the bridge is termed the design spread and generally coincides with the width of the shoulder without encroachment into the traffic lane. Bridge decks can be dif- ficult to drain and meet the standards of the DOT for spread as well as maximum depth of transverse sheet flow because of relatively low cross slopes, bridge railing walls or parapets that collect debris, bridge deck drains and associated piping that are comparatively small, and the resulting potential for clogging of inlets and drainage systems. These issues directly impact the project economics and practicality of various options for runoff treatment as well as the ability to achieve the lowest whole life cost. FHWA (1993) describes how the requirements in the design of deck drainage systems differ from roadway drain- age systems: • Total or near total interception may be a desirable upstream of expansion joints. • Deck drainage systems are susceptible to clogging. • Inlet spacing is often predetermined by bent spacing or piers. • Inlet sizes are sometimes constrained by structural con- siderations. There are a variety of design concepts involving bridge deck inlets and scuppers. Bridge scuppers are defined within this report as openings in the rail or parapet wall to let runoff discharge over the side of the bridge. Deck drains connect to a pipe that discharges at the bridge soffit, through a column or pier wall, or through the bridge superstructure to the abut- ment. Runoff collected in the deck drain enters the pipe and is conveyed to the desired outlet location. Pipe sizing in such systems is usually dictated by hydraulic limitations created by the interception capacity of deck drains. Because of the impact on capital costs and increased difficulty involved with maintenance, direct discharge is preferred over pipe collec- tion systems. Figure 5-1 shows a typical deck drain. Bridge deck inlets are generally fabricated from ductile cast-iron or welded-steel chambers. By contrast, roadway drains are much larger pre-cast, cast-in-place, or masonry structures. Iron is rarely a pollutant of concern for bridge runoff. However, many states require all their metal drainage hardware to be galvanized. Galvanizing is the most popular finish, but it is expensive and can contribute zinc to runoff. Exposed galvanized components of the drainage system should be avoided. Painting of deck inlets is less expensive than galva- nizing. In most locations, painted deck inlets will perform as well as galvanized boxes (Copas and Pennock 1979). For severe duty conditions, epoxy coating can be considered. Pip- ing systems for bridges are typically steel conduits that must Total Suspended Solids (mg/L) Total Zinc (ug/L) Total Lead (ug/L) Total Copper (ug/L) Total Nitrogen (mg/L) Total Phosphorus (mg/L) Nitrate [NO3] (mg/L) Total Kjeldahl Nitrogen [TKN] (mg/L) Dissolved Phosphorus (mg/L) Fecal Coliform (col/100mL) Escherichia Coli [E. Coli ] (col/100mL) Table 5-1. Constituent concentration and load calculations.

38 withstand vibrations and deflections. Fiberglass and PVC conduits are sometimes specified since they avoid contribut- ing iron or zinc to runoff within the bridge conveyance sys- tem and are more flexible than steel and, as such, can better withstand the displacement and associated stresses within a bridge superstructure. They also have the advantage of being inert to oil, gas, salt, ice melting chemicals, and low pH run- off. However, exposed fiberglass and PVC piping should be painted to limit UV exposure. The design of deck drain- age systems must consider pollutant-generating potential, the aesthetics, and maintenance requirements of the conveyance system materials and finishing. The Guide Tool (a separate spreadsheet from the BMP Tool) computes the cost of the conveyance system based on bridge deck area, for three designated levels of drainage system complexity. A description of the three cost levels is provided in Chapter 6. Costs are not calculated in the Tool for a sys- tem that uses a larger deck area or a combination of deck area and scuppers. If appropriate, additional bridge deck area cost would need to be computed separately by the practitioner. 5.2.1 Offset Deck Drain and Raised Scuppers Offset deck drains and raised scuppers are new and rela- tively untested approaches that may effectively collect and convey runoff from small to mid-size bridge projects that are subject to treatment standards of an NPDES permit or Sec- tion 401 Water Quality Certification. Offset deck drains are located at a strategically determined horizontal offset from the bridge side railing. Raised scuppers are vertically raised from the flow line of the bridge deck. Both design approaches allow the standard water quality flow rate to bypass the deck drain or scupper system. Bypass will occur for flow rates at or below the water quality flow rate and will be collected prior to reach- ing the deck joint in an inlet system, from which point runoff Figure 5-1. Deck drain.

39 is piped to a suitably sized treatment BMP. The advantage to this approach is that the length of conveyance pipe can be kept to a minimum along portions of the bridge directly over water (which will typically represent the majority of the structure). The practitioner will be required to determine spacing, offset, and other dimensional details based upon the bridge length, longitudinal gradient, cross slope, and the local water quality and flood flow rates. Since the amount of bypass will accu- mulate along the length of the bridge deck, the offset of deck drains must be designed to increase proportionately along the direction of flow to accommodate an increasing spread. Simi- larly, the height of raised scuppers above the deck must also be designed to increase along the direction of flow. For this reason, implementation of this approach on large or unusu- ally long bridges with flat decks may be impractical. Scuppers or deck drains must also allow for clear travel during flood conditions based on local DOT dry lane standards. Sample relationships between drain spacing, flood encroachment, and longitudinal deck slope are provided for a 1,000-foot example bridge with a 2% cross slope and a water quality flow rate of 0.5 inch per hour. The flooded condition is assumed to pro- duce flow rates at 10 times the water quality storm. Table 5-2 indicates that in many circumstances, scuppers raised within a range of 0.03–0.15 feet (0.36–1.8 inches), or deck drains offset 1.5–7.5 feet from the side rail would be effective in allowing water quality flow rates to bypass a direct discharge to the receiving water, although the limitations on flatter decks (0.5% to 1.0%) is very apparent. A comparison of required head versus allowable head also shows that typi- cal deck drain sizes may have difficulty in capturing flood surcharge within the limit of an 8-foot shoulder at or near the abutment. In these instances, where reduced spacing is not practical, increasing the number of deck drains, provid- ing local inlet depression, and/or using raised scuppers may be more appropriate. 5.3 Treatment Controls Treatment BMPs for runoff from bridge decks can be classi- fied into two general categories: treatment on the bridge deck itself and treatment at the abutment. BMPs located at the abut- ments will require a conveyance system to transfer the deck run- off to the abutment area and the BMP. BMPs on the bridge deck do not require a conveyance system, but will require mainte- nance of the BMP on the bridge structure. The BMPs for bridge runoff treatment addressed in this guide are as follows: Treatment at the abutment: • Swales • Dry detention basin • Bioretention • Sand filter Treatment on the bridge deck or bridge structure: • Bridge scupper treatment • PFC • Floating pile wetland 5.3.1 Treatment at the Abutment Treatment BMPs at the bridge abutment will require suf- ficient right-of-way to construct the BMP and provide main- tenance access (which can be very limited or not available) and a suitable discharge location. Flow from the bridge deck also must be conveyed to the abutment either along the bridge deck or through a pipe system to the abutment area. Prior to selection of abutment treatment BMPs, the additional costs for pipe conveyance systems or additional deck area for runoff conveyance to the abutment should be evaluated. Construction of treatment devices within the floodplain will have regulatory and operational considerations. Improvements that modify the extent or elevation of the floodplain must be submitted for approval by the local floodplain administrator to FEMA. Area under bridges may also be subject to require- ments of the Rivers and Harbors Act. Further, if the BMP is located in the floodplain, it should be designed to ensure that it is not damaged, and will not release pollutants during periods of inundation through re-suspension of accumulated sediment or scour. In general, ensuring that BMPs are above the bank full event is a good minimum standard. Engineering hydraulic analysis of BMP performance during flood events is recommended in the interest of preventing adverse water qual- ity impacts and ensuring a reasonable service life for the BMP. Opportunities to treat runoff from an existing at-grade section of the roadway near the bridge (that is currently untreated), which can be considered in lieu of treating runoff from the bridge, should be assessed. If acceptable to the regula- tory agency, treating off-bridge highway runoff could poten- tially be a more cost-effective alternative with greater benefits to the receiving water. The BMPs described in this guide for treatment at the abut- ment represent the primary, non-proprietary BMPs typically used by DOTs for concentrated flows and can operate pas- sively with extended maintenance intervals. All are proven BMPs that have had substantial study to assess pollutant removal effectiveness and whole life costs. Other BMPs may also be used at the designer’s discretion, but the selected BMPs are those available in the BMP Evaluation Tool (see Chap- ter 6) to allow the designer to evaluate the BMP selection based on performance and whole life cost. 5.3.1.1 Swales Swales are vegetated stormwater conveyances that treat runoff by filtration, shallow sedimentation, and infiltration

Bridge 1/2 Width (ft) Longitudinal Slope (%) Deck Drain/Scupper Spacing (ft) Water Quality Flow (cfs) Surcharge (Bypass) Flood Flow at Each Inlet (cfs) Water Quality Event Flood Event Flow Depth (ft) Flow Spread (ft) Assumed Effective Deck Drain Size Required Head for Surcharge Flow (ft) Allowable Head for Dry Lane Near Abutment (ft) 20 0.5 30 0.01 – 0.23 0.07 0.03 – 0.10 1.50 – 5.00 1’ x 1’ 0.056 0.060 20 1 50 0.01 – 0.23 0.11 0.03 – 0.04 1.50 – 2.00 1’ x 1’ 0.059 0.120 20 3 100 0.02 – 0.23 0.23 0.03 – 0.04 1.50 – 2.00 1’ x 1’ 0.063 0.120 40 0.5 10 0.01 – 0.46 0.05 0.03 – 0.13 1.50 – 6.50 1’ x 1’ (2 each w/depression) 0.020 0.030 40 1 30 0.01 – 0.46 0.14 0.03 – 0.11 1.50 – 5.50 1’ x 1’ (2 each) 0.050 0.050 40 3 100 0.05 – 0.46 0.46 0.03 – 0.04 1.50 – 2.00 1’ x 1’ 0.082 0.120 60 0.5 15 0.01 – 0.69 0.10 0.03 – 0.15 1.50 – 7.50 1’ x 1’ (4 each w/depression) 0.010 0.010 60 1 30 0.02 – 0.69 0.21 0.03 – 0.13 1.50 – 6.50 1’ x 1’ (2 each w/depression) 0.020 0.030 60 3 30 0.02 – 0.69 0.21 0.03 – 0.11 1.50 – 5.50 1’ x 1’ (2 each) 0.046 0.050 Table 5-2. Representative spacing of offset deck drains and raised scuppers.

41 (Figure 5-2). Additional minor removal mechanisms include biochemical processes in the underlying plant- ing media such as adsorption and microbial transforma- tions of dissolved pollutants. Swales provide removal of suspended solids, oil and grease, and metals, in addition to reducing stormwater peak flow. Swales provide limited volume reduction and removal of nutrients and bacteria. The practitioner should take care to ensure that if a swale is used in the floodplain, it would not be scoured during rare events. Primary swale features include: • Dense vegetation layer • Topsoil layer • Optional taller vegetation (height can exceed the design flow depth) • Optional stone or media storage reservoir Typical swale design considerations include: • Slopes – Width and side slope should be chosen such that flow depths in the vegetated swale do not exceed a recom- mended depth of 4 inches. Ideally flows should be at least 2 inches less than grass height – Recommended longitudinal slope of the vegetated swale is between 1% and 2.5% • Design Flow Rate – Design flow velocity should not exceed 1 ft/s to keep the vegetation upright and promote sedimentation 5.3.1.2 Dry Detention Basin Dry detention basins are storage BMPs intended to primar- ily provide peak flow reduction and sedimentation treatment (Figure 5-3). Dry detention basins do not have a permanent pool; they are typically designed to detain stormwater for an extended period for peak flow control (e.g., 36 to 48 hours from full condition) and then drain completely between storm events. The side slopes, bottom, and optional forebay of dry detention basins are typically vegetated. Dry detention basins provide efficient removal of sediments, oil and grease, and particulate-bound pollutants and, where soil conditions allow, can provide substantial volume reduction benefits with infil- tration. Dry detention basins have limited ability to remove dissolved pollutants such as metals, nutrients, and bacteria. Figure 5-2. Vegetated swale schematic. Gravel storage reservoir Primary Overflow Inflow Water quality depth Riser structure Freeboard Emergency Overflow Figure 5-3. Dry detention basin schematic.

42 Primary dry detention basin features include: • Optional sedimentation forebay • Main basin • Optional low flow channel—a narrow, shallow gravel- filled trench that runs the length of the basin to drain dry weather flows • Typical dry detention basin design considerations include: – Space allocation: Consider side slope, maximum depth, and forebay requirements to determine space needed – Outlet design: The outlet should preferably be designed to release the bottom 50% of the detention volume (half- full to empty) over 24 to 32 hours, and the top half (full to half-full) in 12 to 16 hours. – Maintenance access: The basin should be large enough to allow for equipment access via a graded access ramp. – Vegetation: The bottom and slopes of the dry detention basin should be vegetated. 5.3.1.3 Bioretention Bioretention systems (a.k.a. rain gardens) are vegetated shallow depressions filled with an engineered media used to temporarily store stormwater prior to infiltration, evapo- transpiration, or discharge via an underdrain or surface out- let structure (Figure 5-4). By filtering stormwater through an engineered soil mix, bioretention systems can be designed to target a variety of pollutants. Removal of contaminants occurs primarily through filtration, shallow sedimentation, sorp- tion, and infiltration. Additional removal mechanisms include biochemical processes in the underlying engineered planting media such as adsorption and microbial transformations of dissolved pollutants. Bioretention systems remove suspended solids, metals, oil and grease, nutrients, and bacteria, while also reducing volume and peak flow. Primary bioretention features include: • Stones near the inlet for energy dissipation • Shallow mulch layer at the surface • Medium thickness soil layer below the mulch • Optional stone storage layer below the engineered soil layer • Optional underdrain (needed when infiltration rates are low or infiltration is not desired). Upturned elbow to promote infiltration and nitrification • Overflow outlet Typical bioretention design considerations include: • Drawdown Time: Drawdown time of planting media should be less than a few hours • Ponding Depth: Recommended maximum ponding depth is 12 inches • Planting Media: – Recommended minimum planting media depth is 2 feet (3 feet preferred) – Recommended planting media composition: 60 to 70% sand, 15 to 25% compost, and 10 to 20% clean topsoil; organic content 8 to 12%; pH 5.5 to 7.5 Compost for bioretention systems will need to be selected based on local conditions to minimize leaching of nitrogen and phosphorus to groundwater and/or receiving waters. 5.3.1.4 Sand Filter Sand filters treat stormwater runoff via sedimentation, entrap ment, and straining of solids (Figure 5-5). As stormwater passes through the sand filter bed, pollutants are trapped in the small pore spaces between sand grains or are adsorbed to the media surface. Sand filters efficiently remove sediments, oil and grease, metals, and bacteria, as well as reduce peak flow. Sand Figure 5-4. Bioretention schematic.

43 filters provide limited nutrient removal and volume reduction benefits. They typically include a constructed sand bed that receives runoff that spreads over the surface. The sand bed can be contained with a concrete structure; however, when suffi- cient right-of-way is available, design within an earthen contain- ment is preferable. The treatment pathway is vertical (downward through the sand). Ponding on the surface occurs if inflows exceed the rate of percolation through the bed. A system of con- nected underdrain pipes under the sand bed collect and route flows that have percolated through the sand bed to the outlet. Primary sand filter features include: • Sedimentation forebay • Sand filter bed • Optional underdrain in stone trench below the sand filter bed Typical sand filter design considerations include: • Forebay: Recommended forebay size is 20–25% of total volume if no other pretreatment is provided • Media filter bed: Recommended minimum media filter bed depth is 24 inches (36 inches or more preferred) • Slope: Longitudinal slope along length of filter bed should not exceed 2% • Ponding depth: Recommended maximum ponding depth above filter bed is 3 feet • Underdrain: Underdrain should have a minimum diameter of 3 inches and 0.5% minimum slope 5.3.2 Bridge Scupper Treatment Concept NCHRP Report 767: Measuring and Removing Dissolved Metals from Storm Water in Highly Urbanized Areas described a concept level design for treatment of bridge deck runoff specifi- cally targeting metals removal. The concept design is based on the use of granular ferric oxide (GFO) material within a modi- fied scupper to both filter runoff and sorb dissolved metals. The media is housed in a modified bridge deck drain to serve a designated portion of the bridge deck area. Figure 5-6 shows a section of the concept bridge deck drain. The deck drain is designed for interception of the “water quality” design flow only, and must be followed downgrade by a deck drain designed to intercept the remaining drainage flow to maintain spread criteria. The preliminary design procedure for the vault or inlet scup- per is based on laboratory column tests. The media depth is fixed at a minimum value (10 inches), to achieve a minimum contact time with the media consistent with that obtained during the column testing (3 minutes). Thicker media depths may be used, but the required head should be computed using Darcy’s Law. The hydraulic conductivity of the media (K) was determined to be 0.094 in./s. Caution should be used in developing designs with head requirements that are relatively large since the effective solids loading rate of the media will be higher, resulting in shorter runs between maintenance intervals due to possible media occlusion. Hydraulic sizing computations should be based on a media- loading rate of 2 gpm/ft2, a media thickness of 10 inches, and a required head of 8 inches. The required amount of adsorptive media assuming a copper influent concentration of 10 µg/L, a target discharge copper concentration of 3 µg/L, and 30 in./yr of rainfall over a one-acre drainage is 72 kg/yr. The unit weight of GFO was measured as 40 lbs/ft3. The unit cost of GFO is about $15/lb. This cost is likely to decrease for GFO purchased in bulk quantity, since $15 represents the cost of the material obtained for the laboratory trials. The crushed concrete/sand filter layer shown in Figure 5-6 is used to reduce the solids loading to the GFO media and extend the media life. Therefore, the anticipated change in permeability for the media over the operating life of the filter should be relatively small. The crushed concrete/sand layer and geotextile must be removed and replaced whenever the head requirements for the design flow become unacceptably high (exceed the top of the grate). Pre-treatment of flow is recommended to maximize the maintenance interval of the top layer and geotextile fabric. Oversizing the vault will also increase the maintenance interval, but the cost of the GFO media likely makes this option less desirable than a more effective pre-treatment system. Figure 5-5. Sand filter schematic.

44 The custom bridge deck drain would be cast integrally with the bridge deck for concrete box girder and slab designs. Other design configurations would be required for steel bridges. Sol- ids loading on bridge decks is generally consistent with at- grade roadways. Due to the relatively confined dimensions of the bridge deck inlet, pretreatment for solids removal on a bridge deck would be prudent. To reduce solids loading to the bridge scupper, a PFC overlay is recommended. As discussed in this chapter, PFC overlays have been shown to be effective in reducing TSS in highway runoff. The specific design will be based on local conditions, but the overlay should be discontin- ued at a distance from the bridge railing coincident with the edge of the scupper inlet to allow the flow within the overlay to collect in an effective “gutter” area. It is also recommended that a second scupper inlet be provided in the event that the treatment inlet becomes blocked with solids. 5.3.3 PFC PFC is a layer of porous asphalt placed at thicknesses of 1–2 inches on top of conventional impermeable pavement, either PCC or AC. Older mix designs are termed OGFC. PFC is a type of porous pavement, but does not encourage infiltration and reduce runoff volume like full depth porous pavements used in parking lots. Instead, PFC layers remove rainfall from the highway surface and allow it to flow through the porous layer to the side of the road. By removing water from the road surface, PFC improves safety by reducing splash- ing and hydroplaning (NCHRP 2009). PFC could be considered as treatment BMP to avoid the cost of a bridge conveyance system. The PFC layer has been dem- onstrated to improve water quality (Barrett et al. 2006; Pagotto et al. 2000) as well as provide ancillary benefits such as reduced tire noise and improved visibility and stopping distance dur- ing rain events (McDaniel et al. 2010). Performance of the PFC overlay, in areas where a freeze thaw cycle occurs appears satis- factory, as reported by McDaniel et al. (2010), but some dura- bility questions remain (Cooley et al. 2009). Placement of PFC on a bridge deck as a BMP would afford a DOT a logical pilot test site to understand site-specific design life and operational requirements while satisfying runoff treatment objectives. Compared to other practices for treating highway storm- water, PFC has many advantages. The quality of water dis- charged from PFC into the environment is of comparable quality to a sand filter (Eck et al., 2012). However, unlike a sand filter or other conventional practices such as deten- tion basins, bioretention, or filter strips, PFC incorporates stormwater treatment into the roadway surface and does not require additional right-of-way. Maintenance is also not required beyond the periodic milling and re-surfacing that occurs due to structural considerations. The milled PFC is commonly recycled into new conventional pavement, thus preventing any particulate matter retained by the pavement from entering the environment. As a pavement, PFC is more Figure 5-6. Inlet deck drain with media.

45 expensive than conventional asphalt due to better aggregate quality, but when it is installed to reduce noise and improve wet weather drivability the water quality improvements are essentially free. The good quality of PFC runoff combined with the negligible land and maintenance requirements makes PFC a compelling choice for stormwater treatment in the high-speed highway environment. 5.3.4 Floating Pile Wetland A floating pile wetland (Figure 5-7) is a management option for consideration in bridge projects that cross a perennial stream. Floating pile wetlands are a relatively new approach for managing runoff from DOT projects. There has been no known transportation pilot project application Figure 5-7. Floating pile wetland (from WSDOT).

46 of floating pile wetlands. The concept of applying floating wetlands within a transportation project was originally sug- gested by the WSDOT for the SR 520 bridge replacement and HOV project. However, floating wetlands have been con- structed in other applications. For example, over 2,000 square feet of floating wetlands were constructed in the Baltimore Inner Harbor. In this application, the wetlands function to remove nutrients and other pollutants from the water, while providing oxygen and supporting beneficial micro-organism growth (Watershed Partner of Baltimore, Inc. 2014). For more information, the practitioner can refer to: http://www.healthy harborbaltimore.org/whats-happening-now/floating-wetlands Floating pile wetlands appear to present applicability for bridge projects that rely on the use of piers in the structural design, but are limited spatially to provide treatment at the abut- ment area. Like any engineered wetland, floating pile wetlands can be an effective approach at managing soluble pollutants (i.e., soluble metals) generated on the bridge deck. Provision for a sedimentation chamber for pre-treatment within deck is recommended wherever feasible. Flow is intended to discharge through a down-drain system attached to the pier and into a wetland vegetation cell that is constructed within a surround- ing concrete pile. The wetland is intended to operate above the seasonal high water mark of the surrounding river, with a constant ponding depth of several feet. Outflow from the wetland is achieved through a weir built into the side of the pile structure. Much of the original design practice for constructed wet- lands comes from municipal wastewater treatment, as opposed to urban runoff. In municipal wastewater application, ideal residence time is approximately 6 to 7 days (U.S. EPA Office of Research and Development 1988). Due to the inherently limited available storage volume that can be constructed within a typical bridge pier pile, traditional practices will likely require adaptation methods backed by research to determine an achievable residence time. Inadequately short residence times will sacrifice pollutant uptake, and inordinately long residence times can lead to stagnant anaerobic conditions. In addition to evaluating achievable residence and pollut- ant reduction, research should evaluate the influence of fac- tors such as bed slope, plant type, temperature, and organic loading on performance. For more information regarding municipal design procedures for constructed wetlands, the practitioner can refer to http://water.epa.gov/type/wetlands/ upload/design.pdf. The practitioner should also consider some potentially significant limitations associated with floating pile wetlands including adequacy of year round water supply and increased pier scour. A properly functioning wetland system is assumed to require only nominal levels of inspection and maintenance; the degree of difficulty would vary substantially based upon the type and height of bridge as well as the size of the river. 5.3.5 Offsite Mitigation There are a several reasons why offsite mitigation of the impacts of bridge runoff on receiving water quality is preferred. These include the cost and technical feasibility of retrofitting existing or constructing treatment controls for planned bridges; the fact that a significant portion of the contribution of pollut- ants from bridges to receiving waters actually occurs during dry weather through re-suspension; the lack of available space at the bridge abutment areas to construct treatment facilities; and the difficulty of providing routine maintenance for facili- ties installed on or near the bridge structure. Treatment of runoff from an adjacent terrestrial section of highway should result in higher pollutant load reduction as compared to treatment of the bridge deck runoff. Consequently, if treat- ment BMPs are required for bridge deck runoff, this guide recommends constructing the treatment device on a compa- rable section of untreated highway as the most effective and economical option. Selection of offsite mitigation options can be complicated by a number of factors, such as legal restrictions on the use of highway money for projects not part of the road system, the lack of available space for construction of treatment facilities, and the need to collaborate with the public/local officials to obtain project approval. Consequently, it is important to pri- oritize the potential offsite opportunities to reduce the proj- ect cost and speed project delivery. The following ranking of offsite mitigation options is suggested: 1. Untreated runoff from DOT facilities in the watershed that discharge to the receiving water. 2. Small highly impervious catchments within the watershed of concern outside of the highway system. 3. Larger watersheds with less impervious cover outside the highway system within the same watershed. 4. DOT facilities outside the watershed. Clearly, the highest priority for offsite mitigation would be other locations within the highway system (with similar AADT) that discharge untreated runoff to the receiving water of interest. Working within the highway system provides the DOT the ultimate flexibility in determining treatment facil- ity siting, design, and maintenance. The ability to make these decisions unilaterally will substantially speed project delivery and allow the DOT to construct facilities that comply with the DOT’s specifications. Retrofit of roadways for stormwater treatment can occur rapidly and without the need for addi- tional right-of-way (ROW) by using PFC, where appropriate, based on climate and terrain. Retrofitting a small, highly impervious catchment within the watershed, but off the highway system is the next best option. A catchment with characteristics similar to a bridge deck is pre-

47 ferred. These characteristics include a relatively small catch- ment area with high impervious cover. The higher the level of impervious cover, the greater the pollutant load, and the size of the required facility and the number of stakeholders can be minimized. Off-system retrofit for stormwater treatment is more difficult to implement. This approach typically requires a memorandum of understanding (MOU) with either local regulators and/or business and neighborhood interests. An MOU can be extremely time intensive to develop, substan- tially increasing the time for project delivery, and could result in conditions that require the DOT to deviate from standard practices. Retrofitting a larger watershed off the highway system is a less attractive option. In general, a larger watershed will have a much higher pollutant load than the bridge itself. Consequently, the DOT may wish to share the cost of the facility as well as the maintenance requirements. Develop- ing an agreement with other entities to accomplish an equi- table cost sharing can be very challenging, particularly if there is not an immediate need for mitigation by the other stakeholders. A final option, when constraints preclude the retrofit of a catchment within the watershed of interest, is to propose a retrofit project in another part of the highway system. This may be difficult from a regulatory perspective since pollut- ant trading between watersheds for stormwater NPDES com- pliance is not an established practice. The advantage for the DOT, of course, is that it provides the ultimate in flexibility for developing a retrofit plan. 5.4 Spill Controls Spill control requirements for bridges can be viewed in the context of the probability of a spill and the risk to the receiving water. While hazardous material spills within bridge environ- ments are of special concern due to their close proximity to receiving waters and the associated potential for severe water quality impacts, data have shown that spills have rarely occurred on or near bridges (see Section 5.4.1). Nonetheless, a single spill event could cause catastrophic environmental effects depend- ing on the size and sensitivity of the receiving water, requiring intensive response efforts, and subsequent litigious conse- quences. Therefore the probability, potential risk, and impact of spills should be assessed for the bridge water body crossing to determine whether spill controls should be considered. Spill control BMPs can be implemented when deemed necessary to contain accidental spills of hazardous materials. The following sections provide information on bridge spill frequency, costs, characteristics, and recommendations for spill control criteria and structural BMPs. Two case studies are presented where spill controls were implemented as part of the bridge design. 5.4.1 Bridge Spill Frequency The U.S.DOT database (U.S.DOT 2013) on hazardous material incidents was analyzed for the period 2003–2012 to determine the frequency of spills associated with discharge to waterways. Over the 10-year period, there were approxi- mately 140,500 reports of incidents from highways, with 97% of these incidents resulting in spillage. Incidents are classified by transportation phase which includes loading and unload- ing, in-transit storage (e.g., in a terminal or warehouse), and in transit. Loading or unloading accounted for a large major- ity (78%) of the total incidents. For the purposes of the bridge spill frequency evaluation, only in-transit incidents resulting in spillage were evaluated. Thus, of the total reports of incidents resulting in spillage, there were 23,095 (17%) designated as “in transit.” Of these in-transit spill incidents, there were only 329 reports of spills with discharges to storm drains or waterways (less than one/ year/state). Only nine spills were identified as being associ- ated with a bridge located over a waterway. Consequently, these events are extremely rare (less than 0.01% of all reported spills for the analyzed period of record). 5.4.2 Bridge Spill Costs The nine spills associated with bridges over waterways resulted in a total of approximately $2.2 million dollars of damage including: • $78,000 from material loss, • $680,000 from carrier damage, • $450,000 from property damage, • $440,000 of response costs, and • $510,000 of remediation costs. In comparison, $116 million in damages was spent in the 10-year period for all 329 in-transit spills with discharges to storm drains or waterways. Therefore, for the 10-year period, spills associated with bridges accounted for less than 2% of these damage costs. Overall, in-transit spills with discharges to storm drains or waterways account for 0.3% of the total damages from hazard- ous material spill incidents ($761 million of damages for the 10-year period). 5.4.3 Bridge Spill Characteristics The descriptions of the nine spills associated with bridges over waterways vary. Three of the spills occurred after the vehicle made impact with a bridge, while the other incidents occurred in the vicinity of a bridge and were not caused by any characteristic specific to bridge crossings. The spill descriptions

48 do not indicate that special characteristics of bridges are con- sistently the cause of the spills associated with bridge cross- ings. Further study is needed to better understand if bridge characteristics, in comparison to general roadway character- istics, affect the probability of spill occurrence. 5.4.4 Recommended Spill Control Criteria In general, the data on spills presented here do not sup- port special measures to prevent damage to waterways due to bridge spills. Bridge spills represent a small percentage of the in-transit spills associated with discharges to storm drains or waterways and an even smaller subset of highway hazardous material incidents. Therefore, it is recommended that spill prevention measures be taken only when the bridge crosses a water body for which there is zero tolerance for contamina- tion, such as a drinking water reservoir. Other water bodies of special concern for which spill prevention measures should be considered include ONRW. Examples of ONRW water bodies where even a small risk of hazardous material spill may not be tolerable include those that support high-value fisheries and wildlife habitat, and those heavily used for recreation. 5.4.5 Recommended Structural Spill Control BMPs Structural spill control BMPs for the containment of hazard- ous materials must be able to contain and prevent subsequent transport as their primary function. Recommended spill con- trol storage and routing methods are discussed in this section. 5.4.5.1 Spill Storage Methods Various types of spill storage methods can be used either on the bridge or downstream of the bridge deck to contain spills. Recommended hazardous spill containment measures include but are not limited to • Detention basins • Capacity of bridge–incorporated storage within the super- structure (e.g., stability pontoons on floating bridges) • Tanks and vaults • Capacity of the collection and conveyance system (e.g., pipe storage) Excavated detention basins that provide storage and con- trolled release are the most common form of hazardous spill containment measures and can be constructed near bridge abutments when adequate open space exists and conveyance from the bridge deck to this area is feasible. In situations where the slope and hydraulic gradient are limiting or making offsite conveyance infeasible, such as for floating bridges, storage can be incorporated into the bridge structure itself. An example of bridge-incorporated storage is provided in Section 5.4.6, Case Study 2 where supplemental pontoons have been used as temporary storage facilities that can be pumped out by responders in the event of a spill. A variety of tanks, vaults, and conveyance storage exist for the purposes of spill containment in different sizes and materials. An advantage to these closed storage facilities, as opposed to those open to the environment like detention basins and pontoons, is that they can be placed below ground and can reduce the potential contact of spilled contaminants with the atmosphere, rainwater, or soils. A disadvantage of closed storage facilities is that they typically cost more per unit volume than detention basins and can be more expen- sive to maintain (e.g., may require confined space entry). Due to reduced potential for transport and dispersion of hazard- ous materials to the environment, closed systems are generally recommended for hazardous spill control. 5.4.5.2 Spill Conveyance and Routing Methods Spill conveyance and the method of routing the spilled material to the storage BMP are significant design consid- erations for successful spill containment. Routing for spill- dedicated detention basins, or other storage BMPs such as tanks or vaults, can be either in-line or off-line, where in-line represents a conveyance system with one or more storage BMPs along the stormwater flow route and off-line rep- resents a conveyance system with an isolated storage BMP for hazardous spill containment. The isolation point for the off-line system is typically near the bridge abutment where the piped runoff containing the bridge spill material can be redirected. In-line and off-line systems have various pros and cons. In- line systems typically require less infrastructure and design because flows do not have to be routed off the main convey- ance route and isolated. However, in-line systems pose a higher risk to contamination to receiving waters if shutoff valves are not quickly closed in the event of a spill. Thus, off-line sys- tems could potentially avoid contamination to both water and soil by containing spills separate from stormwater systems. Cleanup efforts are also likely to be less intensive and costly with off-line systems due to reduced spread and transport of hazardous materials. Due to less environmental risks and costs of spill cleanup, off-line systems are generally recommended for hazardous spill control. 5.4.6 Spill Control Case Studies The following case studies give examples of hazardous spill mitigation technologies used within bridge-specific environments.

49 Case Study 1: ODOT MAH-80 Project In 2009, the Ohio Department of Transportation (ODOT) completed construction of MAH-80, an extensive $87 mil- lion project to widen and reconstruct portions of I-80 and dual 2,500-foot bridges. The bridges span Meander Creek Reservoir, which supplies drinking water to nearby towns including Youngstown and Niles. The MAH-80 project captured industry attention by including ODOT’s first spill containment system, designed to prevent spills on I-80 from entering the Meander Creek Reservoir. Key components of the MAH-80 spill containment system include the following: • A bridge profile that crests midway over the reservoir span • A crowned bridge deck that sheds runoff to 10-12 foot shoulders sized to store and convey runoff to approach inlets without encroaching on driving lanes • Networks of inlets, piping, and roadside ditches and swales • Two containment basins at low points on opposite sides of the reservoir • Two control chambers equipped with shutoff valves that prevent hazardous materials from entering the reservoir Under typical conditions, stormwater runoff is collected and routed to basins sized to contain the 100-year event and then discharged from the basins to the Meander Creek Res- ervoir. In the event of a spill, the containment system allows emergency responders a maximum response time of 30 min- utes to close the two shutoff valves located at each respec- tive basin. Closing the shutoff valves allows for the spill to be contained within each basin before entering the reservoir that can then be pumped out and disposed of in accordance with local and federal regulations. This project, which included 12.5 acres of wetland habitat creation to mitigate environmental impacts in addition to the spill containment system, was selected as the co-recipient of the 2010 Outstanding New Major Bridge Award in the 2010 Associ- ation for Bridge Construction and Design (ABCD) Northeast- ern Ohio Chapter’s Outstanding Bridge Awards competition. Case Study 2: Washington State Route 520 Bridge Replacement Study The WSDOT State Route 520 (SR 520) bridge replace- ment study developed water quality protection measures for the replacement of SR 520 Evergreen Point Bridge, a float- ing bridge spanning Lake Washington. Water quality protec- tion measures were developed using All Known, Available and Reasonable Technology (AKART) for the handling and treatment of stormwater runoff and spills affecting receiving water quality for bridge applications (CH2M HILL 2010). The AKART study resulted in identification of the follow- ing nonstructural and structural BMPs to protect receiving waters: • High-efficiency sweeping • Large, modified catch basins with scheduled cleaning • Separate, enclosed spill-containment lagoons within sup- plemental stability pontoons The proposed six-lane bridge, targeted to open to drivers in 2014 (U.S. DOT FHWA 2013), will use main pontoons for roadway support and additional lateral pontoons for the purpose of stability, stormwater dilution and spill contain- ment. These lateral pontoons are deemed supplemental sta- bility pontoons (SSPs), and the drainage system of the bridge directs all stormwater runoff to containment lagoons within the SSPs. Once routed to the containment lagoons, floatable materials can then be pumped out by responders. Periodic removal of surface pollutants would also be part of regular maintenance. Under normal operations (when not used for containment purposes), the lagoons provide for dilution of remaining non-surface-removable pollutants and general stormwater treatment prior to discharge. Dilution of storm- water is achieved within the SSPs by providing an internal mixing zone prior to transport to the receiving waters through subsurface openings. Although dilution does not reduce the pollutant load to the receiving waters, it reduces the potential for acute toxic effects to aquatic organisms (WSDOT 2011). The western and eastern approaches for the SR 520 Bridge include various water quality and quantity features. The west- ern approach to the bridge includes several proposed LID improvements and spill containment features. For all improve- ment options, the spill containment features are consistently in the form of an underground vault. The underground vault is designed with the purpose of capturing effluent liquids from fire suppression activities and hazardous spill storage (WSDOT 2011).