National Academies Press: OpenBook

Bridge Stormwater Runoff Analysis and Treatment Options (2014)

Chapter: Appendix A - Literature Review

« Previous: References
Page 64
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 64
Page 65
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 65
Page 66
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 66
Page 67
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 67
Page 68
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 68
Page 69
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 69
Page 70
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 70
Page 71
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 71
Page 72
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 72
Page 73
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 73
Page 74
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 74
Page 75
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 75
Page 76
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 76
Page 77
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 77
Page 78
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 78
Page 79
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 79
Page 80
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 80
Page 81
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 81
Page 82
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 82
Page 83
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 83
Page 84
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 84
Page 85
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 85
Page 86
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 86
Page 87
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 87
Page 88
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 88
Page 89
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 89
Page 90
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 90
Page 91
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 91
Page 92
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 92
Page 93
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 93
Page 94
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 94
Page 95
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 95
Page 96
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 96
Page 97
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 97
Page 98
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 98
Page 99
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 99
Page 100
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 100
Page 101
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 101
Page 102
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 102
Page 103
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 103
Page 104
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 104
Page 105
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 105
Page 106
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 106
Page 107
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 107
Page 108
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 108
Page 109
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 109
Page 110
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 110
Page 111
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 111
Page 112
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 112
Page 113
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 113
Page 114
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 114
Page 115
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 115
Page 116
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 116
Page 117
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 117
Page 118
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 118
Page 119
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 119
Page 120
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 120
Page 121
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 121
Page 122
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 122
Page 123
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 123
Page 124
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 124
Page 125
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 125
Page 126
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 126
Page 127
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 127
Page 128
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 128
Page 129
Suggested Citation:"Appendix A - Literature Review." National Academies of Sciences, Engineering, and Medicine. 2014. Bridge Stormwater Runoff Analysis and Treatment Options. Washington, DC: The National Academies Press. doi: 10.17226/22395.
×
Page 129

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.

A-1 Literature Review Project Overview NCHRP Project 25-42 provides guidance for assessing potential water quality impacts and selecting BMPs for storm- water runoff from bridge decks and vehicle approaches. The study focuses on bridge structures that cross a waterway and discharge directly to the receiving water. As an additional resource, the reader may find value in reviewing the report developed as a part of NCHRP Proj- ect 25-40, “Long-Term Performance and Life-Cycle Costs of Stormwater Best Management Practices,” which is currently in process and will develop guidelines for the selection and maintenance of highway related stormwater BMPs based on long-term performance and life-cycle costs. The NCHRP Project 25-40 literature review, survey, and associated inter- views describe what DOTs and others are doing to under- stand maintenance needs and costs of post-construction stormwater BMPs. NCHRP Project 25-40 provides decision- making guidance on a number of key areas for highway BMPs, including: • Defining and predicting long-term performance, service life, and maintenance costs, and selecting appropriate per- formance measures based on the best current information and practice; • determining appropriate inspection schedules and procedures; • determining appropriate maintenance schedules and procedures; • incorporating long-term performance and life cycle costs into BMP selection processes; • ensuring that funding, staffing, and training requirements are understood and considered by all relevant functional areas within the transportation agency for the selection, installation, inspection, and maintenance of BMPs; and • identifying life-cycle data collection and analysis protocols to facilitate future evaluation of long-term BMP performance. DOTs, cities, and counties have installed few structural BMPs to treat bridge decks. The quality of bridge deck runoff is generally comparable to non-bridge deck roadway runoff. Bridge decks represent only a small fraction of the impervious area of the highway system with runoff that reaches receiving waters. Still, agencies are concerned that the direct connec- tion and untreated runoff from bridges may affect receiving waters; this project and individual DOTs are examining the environmental benefits that can be attained with additional structural and non-structural controls, as well as their costs. Literature Review Methodology Generally, the literature review builds on a previous NCHRP research study (2002).1 In addition to summarizing the most pertinent information in NCHRP Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volumes 1 and 2 and Stormwater Runoff from Bridges by the North Carolina Department of Transportation and URS (2010), literature on the topic was examined to accomplish the following: • Define the characteristics of bridge deck runoff and its potential impacts on receiving waters. • Identify runoff management strategies and how they are influenced by the physical constraints of bridge structures in new construction and retrofit scenarios. • Identify appropriate mitigation strategies for bridge deck runoff, including structural controls and source control measures. • Create a BMP selection tool for specific application on bridge decks. • Accurately quantify “whole life” cost/benefit relationships for bridge deck runoff mitigation. A P P E N D I X A 1 NCHRP 25-42 panel meeting, project kick-off, December 4, 2012

A-2 DOT Survey Methodology The research team contacted a range of DOTs, including those known to be active in BMP and highway stormwater investigations to gain insight into current issues and practices relating to the management of stormwater discharge from bridges. Several state DOTs have stormwater research divi- sions that are engaged in original highway runoff assessments and were able to suggest additional studies that were utilized in the literature research effort. The research team performed a targeted survey, consisting of personal interviews, with the nine DOTs listed below. • Florida DOT (FDOT) • Massachusetts DOT (MassDOT) • Louisiana Department of Transportation Development (LADOTD) • Maryland State Highway Administration (MDSHA) • Nebraska Department of Roads (NDOR) • North Carolina Department of Transportation (NCDOT) • South Carolina Department of Transportation (SCDOT) • Texas Department of Transportation (TxDOT) • Washington State Department of Transportation (WSDOT) Interviews and email exchanges occurred between Decem- ber 4, 2012 and January 29, 2013. A wide variety of practi- tioners participated, including bridge designers, hydraulic engineers, hydraulic division chiefs, landscape architects, and water quality program managers. Each DOT was asked the following questions, at minimum: 1. Are you currently treating bridge deck runoff? 2. Why/Why not? 3. If you do treat bridge runoff, what bridge runoff manage- ment strategies/BMPs do you use? a. New construction strategies b. Retrofit strategies c. Source control approaches d. Emerging BMPs for bridges i. Additives to PFC to target removal of specific con- stituents of concern ii. Alternatives to bridge materials (such as zinc coatings) iii. Coatings for pavement to improve runoff sanitary quality iv. Alternatives for mitigation of hazardous materials spills v. Addition and use of smart controllers for maxi- mizing BMP performance under constrained conditions vi. Other e. General design strategies f. How hazardous material spills are handled g. Endangered Species Act (ESA) implications h. Emerging BMPs for bridges 4. What issues and implementation barriers are you facing? 5. What methods and/or tools do you use to identify and select appropriate mitigation strategies for bridge deck runoff? 6. How do you assess cost-benefit of runoff mitigation strategies? 7. Do you have bridge runoff datasets you could share with the NCHRP 25-42 research team? Bridge Deck Runoff Characteristics and Receiving Water Impacts According to National Bridge Inspection Standards, a bridge is a structure, including supports, erected over a depression or an obstruction, such as water, highway, or railway, and hav- ing a track or passageway for carrying traffic or other moving loads, and having an opening measured along the center of the roadway of more than 20 feet between under copings of abutments or spring lines of arches, or extreme ends of open- ings for multiple boxes. It may also include multiple pipes where the clear distance between openings is less than half of the smaller contiguous opening. For the purpose of the 25-42 study, the panel determined that bridges are highway structures directly discharging over open water.2 As owners of state highways and bridges, DOTs are inter- ested in discerning whether contamination of water bodies from roads and bridges is significant and, if it is, what mitiga- tion is appropriate. The most comprehensive prior research on the topic is NCHRP Report 474 (2002) and a multi-agency study led by the North Carolina Department of Transporta- tion (URS 2010). Bridge Deck Runoff Characteristics The NCDOT report (URS 2010) found “no compelling evidence that bridge deck runoff in North Carolina is higher in [pollutants] typically associated with stormwater runoff as compared to runoff from other roadways.”3 Malina et al. (2005) showed that bridge deck runoff is gen- erally not statistically different from highway runoff.4 Malina’s 2 NCHRP 25-42 panel meeting, project kick-off, December 4, 2012 3 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Legisla- tion Transportation Oversight Committee, North Carolina Department of Trans- portation (2010), p. 4-5 4 Malina et al., Characterization of Stormwater Runoff from a Bridge Deck and Approach Highway, Effects on Receiving Water Quality in Austin, Texas, TxDOT, (2005).

A-3 statistical data comparing bridge deck runoff event mean con- centrations (EMCs) to the approach highway revealed only limited instances when parameters were significantly differ- ent from each other. Pollutant concentrations reflected (same order of magnitude) or were less than average historical high- way runoff concentrations, such that Malina concluded that highway runoff data can be used as a conservative approxi- mation of bridge deck runoff quality.5 At Barton Creek, Malina found that loading of all measured water quality con- stituents was minimal, with “no substantial adverse impact to the receiving streams . . . observed or indicated by bridge deck runoff from the three monitored sites.” 6 Loadings from upstream sources were several orders of magnitude greater. As Nwaneshiudu and others have pointed out, “Most of the pollution found in highway runoff is both directly and indirectly contributed by vehicles such as cars and trucks. The constituents that contribute the majority of the pollu- tion, such as metals, chemical oxygen demand, oil and grease, are generally deposited on the highways.”7 Jongedyk (1999) and Dupuis (2002) list common pollutants in highway runoff as metals, inorganic salts, aromatic hydrocarbons, suspended solids, and materials that are a result of wear and tear on a vehicle, such as oil, grease, rust, and rubber particles.8 Traffic patterns, bridge characteristics, antecedent dry periods, sea- sonal cumulative rainfall, rainfall intensity, and land use are contributing factors as well,9 and atmospheric deposition can be the major source of some parameters, such as trace metals, in urban watersheds (Sabin et al., 2005).10 Splash from surface water on roadways rinses the underside of vehicles and sur- face water carries salt that may have been applied to the road in winter maintenance and pollutants from air deposition to receiving water if sources are not controlled or pollutants are not detained. Metals have acute and chronic toxicity to aquatic life, par- ticulates are the carriers of other pollutants and sedimenta- tion effects on aquatic habitat, nutrients can contribute to eutrophication and salts have aquatic life toxicity effect and affects drinking water supply taste.11 Roadway stormwater runoff data has been independently collected and studied by many sources.12 FHWA’s Effects of Highway Runoff on Receiving Waters—Volume IV Proce- dural Guidelines for Environmental Assessments (Dupuis and Kobringer, 1985) identified several parameters that affect the magnitude of pollution in highway runoff, which can be grouped in the following general categories:13 • Traffic characteristics—speed, volume, vehicular mix (cars/ trucks), congestion factors, and state regulations control- ling exhaust emissions; • Highway design—pavement material, percentage impervi- ous area, and drainage design; • Maintenance activities—road cleaning, roadside mowing, herbicide spraying, road sanding/salting, and road repair; • Accidental spills—sand, gravel, oils, and chemicals. Generally, roadway runoff water quality data is used as an approximation for the pollutant profile of bridge deck run- off (Dupuis 2002). Common highway runoff pollutants and their primary sources include the following, as outlined in multiple studies to date: Particulates Pavement wear, vehicles, atmosphere Nitrogen, Phosphorus Atmosphere, roadside fertilizer application Lead Tire wear, auto exhaust Zinc Tire wear, motor oil, grease Iron Auto body rust, steel highway structures, moving engine parts Copper Metal plating, brake lining wear, moving engine parts, bearing and bushing wear, fungicides and insecticides Cadmium Tire wear, insecticides Chromium Metal plating, moving engine parts, brake lining wear Nickel Diesel fuel and gasoline, lubricating oil, metal plating, brake lining wear, asphalt paving Manganese Moving engine parts 5 Malina et al., Characterization of Stormwater Runoff from a Bridge Deck and Approach Highway, Effects on Receiving Water Quality in Austin, Texas, TxDOT, (2005). 6 Malina et al., Characterization of Stormwater Runoff from a Bridge Deck and Approach Highway, Effects on Receiving Water Quality in Austin, Texas, TxDOT, (2005). 7 Nwaneshiudu, Oke (2004). Assessing effects of highway bridge deck runoff on near-by receiving waters in coastal margins using remote monitoring tech- niques. Master’s thesis, Texas A&M University. Texas A&M University. http:// hdl.handle.net/1969.1/1462 8 Jongedyk, H. 1999. FHWA Environmental Technology Brief: Is Highway Run- off a Serious Problem? Washington DC: Federal Highway Administration and Dupuis (2002) 9 Kayhanian, M., A. Singh, C. Suverkropp, and S. Borroum. 2003. Impact of Annual Average Daily Traffic on Highway Runoff Pollutant Concentrations. Journal of Environmental Engineering 129 (11): 975-990. Kayhanian, M., C. Suverkropp, A. Ruby, and K. Tsay. 2007. Characterization and Prediction of Highway runoff Constituent Event Mean Concentration. Journal of Environmental Management 85 (1): 279-295 10 Kayhanian et al., 2003) and atmospheric deposition can be the major source of some parameters, such as trace metals, in urban watersheds (Sabin et al., 2005) as cited in NCDOT 11 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 11 12 Gupta et al., 1981; FHWA, 1990; Sansalone and Buchberger, 1997; Barrett et al., 1998, Kayhanian et al., 2007. Barrett, M. E., Irish, B. L., Lesso, G. W. III., Malina, J. F. Jr., and Charbeneau, R. J. (1998). “Characterization of highway runoff in Austin, Texas, area,” Journal of Environmental Engineering-ASCE, 124 (2), 131-137 13 FHWA’s Effects of Highway Runoff on Receiving Waters – Volume IV Proce- dural Guidelines for Environmental Assessments (Dupuis and Kobringer, 1985)

A-4 Cyanide Anti-cake compound used to keep deicing salt granular Sodium, Calcium, Chloride Deicing salts Sulphate Roadway materials, fuel, deicing salts Petroleum Spills, leaks or blow-by of motor lubricants, antifreeze and hydraulic fluids, asphalt surface leachate The Impact of Curbed vs. Uncurbed Sections on Bridge Decks Some research has pointed to the accumulation of pollut- ants and sediments along curbed sections of bridge decks. Wu and Allen performed research on stormwater runoff in North Carolina. The majority of Wu and Allen’s (2001) study sites found runoff concentrations from bridges similar to pre- viously published urban runoff data for the Charlotte, NC area, however, the higher levels of pollutants on one bridge prompted the researchers to postulate that the bridge deck curb (railing) wall might be responsible for accumulation of pollutants.14 High traffic and a lack of (pervious surface) were also considered to be factors. Bridge Deck Impacts to Receiving Water A variety of variables (rural vs. urban environment, aver- age daily traffic, curbed vs. non-curbed section, climate, runoff volume, time from previous rainfall event, receiv- ing water chemistry and/or flow) can influence the degree of impact bridge deck runoff may have on receiving waters. The following examination of research literature is divided into that occurring in the last decade or so and pre-2000 research. Early Research Findings In his 1999 work for FHWA, Dupuis (1999) described 19 different methods to manage, assess, and identify bridge deck runoff that could potentially affect receiving waters. Dupuis suggested consideration of average daily traffic in the area, if the bridge is a retrofit or a replacement bridge, and usage and hydrology of the receiving water; e.g., if it is fresh- water, saltwater, drinking water supply, lake, etc.15 A relatively small number of earlier studies focused on bridge deck runoff prior to Dupuis’s research; these included Yousef et al. 1984; Kszos et al. 1990; and Dupuis et al. 1985. A 1998 study by CH2M Hill sampled bridge deck runoff at two sites. Predictive models for highway runoff have estimated water quality based on average daily traffic (ADT), urban vs. rural location, vehicle traffic during storms, and other vari- ables (Barrett et al. 1995; Driscoll et al. 1990, etc.). Caltrans, based on analysis of its own data (Racin et al. 1982) also determined in 1992 that fewer than 30,000 vehicles during a storm, equated to mean 30,000 ADT, would have “. . . little or no impact, because corresponding constituent masses were relatively small.”16 A 1996 Florida study, Effectiveness of a Stormwater Collec- tion and Detention System for Reducing Constituent Loads from Bridge Runoff in Pinellas County, Stoker (1996) found evidence of “first flush” impacts, in particular that:17 • Most constituents measured in stormwater runoff from the bridge were greatest at the beginning of the storm. • Quality of stormwater runoff from the bridge varied with season, runoff volume, and the antecedent dry period. • Maximum values of most measured constituents occurred in the spring of 1994 when rainfall was minimal. • Maximum stormwater loads of nitrogen, iron, aluminum, nickel, and zinc occurred on August 22, 1995, also the date of maximum measured storm volume. In his meta-analysis of existing studies, for his 2002 NCHRP Report, Dupuis said while several studies had shown that direct drainage to some types of receiving waters (e.g., small lakes) could cause localized increases in certain pollut- ant concentrations, most studies did not consider whether such increases adversely affected the biota or other receiving water uses. The only comprehensive study of bridge runoff at that time, FHWA’s I-94/Lower Nemahbin Lake site, found that although direct scupper drainage increased metals con- centrations in near-scupper surficial sediments, biosurveys and in situ bioassays found no significant adverse effects on aquatic biota near the scuppers. FHWA concluded that for lower traffic volume bridges at least, runoff had a negligible impact, based on results of its Phase III program (Dupuis et al. 1985a), which included extensive bioassay testing and field study at three sites that had traffic volume less than 30,000 vehicles per day (VPD). 14 Jy S. Wu and Craig J. Allen, Sampling and Testing of Stormwater Runoff from North Carolina Highways (2001). Wu, J. S., Allan, J. C., Saunders, W. L., and Evett, J. B. (1998). “Characterization and pollutant loading estimation for highway runoff”, Journal of Environmental Engineering-ASCE, 124, (7), 584-592 15 Dupuis, T., National Cooperative Highway Research Program (NCHRP) Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volume 1 Final Report (2002). 16 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Legisla- tion Transportation Oversight Committee, North Carolina Department of Trans- portation (2010), p. 22, citing Report 474, citing Racin, 1998. 17 Stoker, Effectiveness of a Stormwater Collection and Detention System for Reducing Constituent Loads from Bridge Runoff in Pinellas County, Florida Stoker (1996)

A-5 Literature Research over the Last Decade The primary recent US studies assessing the effect of bridge deck runoff on receiving water beneficial uses (NCDOT/ USGS/URS, 2010, NCHRP, 2006, and Malina et al. 2005) concluded that bridge deck runoff is not a primary source of receiving water impairments; however, deicing practices, bridge components (galvanized metal railing), and sensitive or otherwise outstanding resource waters merit further con- sideration. In addition, ADT remains an indicator of poten- tially higher pollutant concentrations in runoff. NCHRP Report 474 (Dupuis 2002) reviewed scientific and technical literature addressing bridge deck runoff and high- way runoff performed by FHWA, USGS, state DOTs, and universities, focusing on the identification and quantification of pollutants in bridge deck runoff and how to identify the impacts of bridge deck runoff pollutants to receiving waters using a weight-of-evidence approach. Dupuis et al. found no clear link between bridge deck runoff and biological impair- ment, though salt from deicing could be a concern. Other conclusions were as follows:18 • Undiluted highway runoff can exceed federal and state ambient water quality criteria, but this alone does not automatically result in negative effects to receiving waters. • The quality and use of receiving waters, as well as the flow path and possible transformations of pollutants in runoff, must be considered independently of runoff loading. • Lead concentrations in highway runoff have significantly decreased since the 1970s due to the phase-out of leaded gasoline. • Direct discharge to some types of receiving streams, pri- marily small streams and lakes, can lead to localized increases in pollutant concentrations in sediment and, in some cases, aquatic biota. However, whether localized effects adversely affected biota was unknown. • Comparison of historic metal toxicity research to present day data may prove difficult due to the measurement of metal toxicity shifting from total metals to dissolved metals. • The ability of sediment to accumulate metals, polycy- clic aromatic hydrocarbons (PAHs), nutrients, and other compounds warrants further research of sediment qual- ity impacts and further development of standards and criteria. • The results of bioassay testing using whole effluent toxic- ity from various studies have been mixed. For the studies that do show some level of toxicity, the runoff samples were high in salt content from deicing activities. The bio- assay methods used by these studies may not be appro- priate for evaluating stormwater runoff. Most bioassays expose the organism being testing continuously to run- off for long periods of time. However, stormwater runoff is delivered to receiving streams in short, intermittent time frames. NCHRP Report 474 noted, “Highways typically constitute a very small fraction of a watershed’s total drainage area, and bridges often constitute a small portion of the highway drainage area. Thus, highways often, but not always, contribute a small fraction of the over- all pollutant load to a given receiving water body, and bridges contribute even less.” According to NCHRP Report 474, “This circumstance pro- vides opportunities to consider and implement common- sense solutions such as providing enhanced pollutant removal somewhere else in the ROW, or even somewhere else in the watershed (i.e., off-site mitigation, or pollutant trading).”19 Perkins and Hazirbaba (2010) also concluded that “contami- nation is slight, unlikely to affect the receiving waters, and not sufficient to warrant concern.”20 Nwaneshiudu assessed the quantity and quality of storm- water runoff from a bridge that spans Clear Creek as a part of highway FM 528 near Houston, Texas. He found that an old galvanized metal bridge railing was contributing to “zinc con- centrations ten times higher than the culvert and creek sam- ples and higher than the USEPA standard.”21 Nwaneshiudu also concluded that: 22 • Total copper and dissolved copper concentrations from the bridge deck runoff were also consistently higher than the USEPA standard. • Total lead and dissolved lead concentrations from bridge deck runoff were orders of magnitude less than the USEPA standard. • Chemical oxygen demand (COD) concentrations from bridge deck runoff were significantly less than values from a nationwide survey of highway runoff data (FHWA 1990). 18 Dupuis, T., National Cooperative Highway Research Program (NCHRP) Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volume 1 Final Report (2002), as cited in NCDOT, 2010. 19 Dupuis, T., National Cooperative Highway Research Program (NCHRP) Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volume 1 Final Report (2002), p. 4. 20 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 4. 21 Nwaneshiudu, Oke (2004). Assessing effects of highway bridge deck runoff on near-by receiving waters in coastal margins using remote monitoring tech- niques. Master’s thesis, Texas A&M University. Texas A&M University. http:// hdl.handle.net/1969.1/1462. 22 Nwaneshiudu, Oke (2004). Assessing effects of highway bridge deck runoff on near-by receiving waters in coastal margins using remote monitoring tech- niques. Master’s thesis, Texas A&M University. Texas A&M University. http:// hdl.handle.net/1969.1/1462.

A-6 • Phosphate concentrations in the creek were on average much higher than concentrations from bridge deck runoff. • Total nitrogen and total Kjeldahl nitrogen (TKN) concen- trations showed no trend, but were sometimes above the USEPA standard. • Total suspended solids (TSS) and volatile suspended solids (VSS) concentrations showed no consistency or noticeable trends and were relatively low. Total suspended solids con- centrations were highest in the creek. Kayhanian et al. (2003) performed a statistical analysis on a Caltrans highway runoff dataset of monitoring data from 83 highway sites over a 4-year period. Kayhanian’s multiple linear regression analysis revealed that ADT, event rainfall, cumulative seasonal precipitation, and antecedent dry period each had a similar, statistically significant effect on pollut- ant concentrations; however, ADT is the only parameter that can be reasonably quantified by transportation agencies in advance of a project being built.23 Thus ADT remains an “indi- cator of potentially high pollutant concentrations and can be useful for locating sites which would benefit from poten- tial BMP retrofit installations. However, because ADT alone cannot accurately predict whether pollutant concentrations at a particular site will be higher than another, it should not be used as a sole indicator of impact.”24 Sabin and Schiff (2008) thought that recent research linking atmospheric deposition of metals to proximity to urban areas and accounting for a significant portion of metal inputs to runoff suggest that defining urban roadways by population is appropriate.25 In 2010, North Carolina DOT concluded a legislatively mandated study of 50 bridges. The objectives of this study were to (1) quantify the constituents in stormwater runoff from bridges across the state, (2) evaluate the treatment prac- tices that can be used to reduce constituent loadings to sur- face waters from bridges, and (3) determine the effectiveness of the evaluated treatment practices.26 NCDOT also summa- rized conclusions from previous studies:27 • Pollutant loadings from bridge decks to a receiving stream can be minimal when compared to pollutant loadings from other watershed sources. • Specific instances of elevated parameters, particularly zinc, may be linked to galvanized bridge materials. • While parameter concentrations in bridge deck runoff can exceed nationwide benchmarks, no widespread link between bridge deck runoff and negative impacts to receiv- ing streams has been shown. • Deicing activities and pollutant accumulation in sediment are potential sources of localized toxicity that require fur- ther study. NCDOT concluded that these observations “support the concept that surface water quality protection may be better served by managing stormwater runoff on a watershed scale as opposed to focusing management efforts specifically on bridges. In addition, there may be opportunities to improve water quality by identifying and controlling the source of pollutants (e.g., by replacing certain bridge materials).” 28 NCDOT also developed a treatment BMP selection frame- work and estimated costs. NCDOT’s study resulted in a number of major observa- tions, including the relatively minor importance of ADT and the relative importance of the urban-rural distinction:29 • Similar to previous studies, ADT showed a small influence on pollutant distributions with only total recoverable zinc, cop- per, and cadmium significantly higher for high ADT bridges. • Differences between total recoverable metals, particularly nickel, aluminum, manganese, iron, chromium, and lead, tend to track significant differences in total suspended solids. These metals tend to be predominantly particulate- bound, with the exception of manganese (Blazier 2003). Therefore, these results may reflect a difference in solids generation by bridge characteristic. • Of all the characteristics investigated, the urban versus rural designation appears to have the most influence on pollutant loading. All solids parameters studied were higher in urban areas, as well as most total recoverable metals and dissolved copper and lead. Similar relationships were also noted for the asphalt versus concrete hypothesis testing (pollutant loading from asphalt surfaces is higher), but most urban bridges were also concrete. • For characteristics in which total recoverable arsenic showed a significant difference (statewide vs. regional, regional vs. subregional, urban vs. rural, piedmont vs. coastal, and blue ridge vs. coastal), the higher arsenic mean and median was 23 Kayhanian et al. (2003) cited in NCDOT 4-34 24 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Legisla- tion Transportation Oversight Committee, North Carolina Department of Trans- portation (2010), p. 4-34 25 Sabin and Schiff, 2008, cited in NCDOT 4-35 26 USGS, North Carolina Water Science Center, Water quality characterization of bridge deck runoff in NC http://nc.water.usgs.gov/projects/bridge_runoff/ overview.html 27 URS, Stormwater Runoff from Bridges, Final Report to Joint Legislation Transportation Oversight Committee, North Carolina Department of Trans- portation (2010), p. 2-3 28 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 2-3 29 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 4-36

A-7 associated with the characteristic opposite to other total recoverable and dissolved metals and TSS. Because a major source of arsenic in stormwater runoff is air deposition from point sources (e.g., coal-fired power plants), total recoverable arsenic loads in highway runoff may be related to atmospheric pathways. Higher total recoverable arsenic distributions were noted for coastal bridges primarily. Even though total recoverable arsenic was significantly higher in regional bridges as compared to statewide and subregional bridges, both regional bridge sites in this analysis are also coastal sites. • Significantly higher nutrients were generally found in pied- mont, regional, and subregional bridges and were associated with asphalt pavement. Surprisingly, only nitrate+nitrite and dissolved orthophosphate were significantly dif- ferent between urban and rural sites. Further, the nitrate+nitrite distribution was higher in urban sites as opposed to rural sites. • Dissolved metals, as a whole, did not exhibit any strong relationship with any one bridge characteristic. Dissolved zinc was only significantly different based on bridge sur- face material, with higher concentrations noted for asphalt bridges. The dissolved lead distribution was also higher in asphalt bridges. Dissolved copper and dissolved lead concentrations were significantly higher in piedmont and urban bridges. Dissolved cadmium concentrations were higher for statewide and regional bridges, but showed no significant difference between urban and rural bridges and high and low ADT bridges. Further studies are underway, such as source assessment and monitoring to determine levels of bacteria from the Virginia Dare Bridge.30 Shellfish contamination and contri- bution of bacteria from bridge decks are of potential con- cern in some coastal states. Monitoring for bacteria in bridge deck runoff was not included in NCDOT’s monitoring plan for their 2010 report because of the logistics required for the short holding times and available certified labs. Impact of Bridge Deck Runoff on Sediment Quality NCDOT (2010) found no statistically significant differ- ences in sediment inorganic or organic concentrations down- stream from no-direct discharge bridges as compared with direct discharge bridges or downstream as compared with upstream locations. Overall, the North Carolina analysis of streambed sediment did not indicate any impacts of bridge deck runoff on sediment quality. Ecoregional differences were observed for some analytes, but these differences seemed to be associated with naturally occurring conditions or upstream anthropogenic influences. Furthermore, where sediment quality benchmarks were exceeded, except for lead and mer- cury, the exceedances were found to be independent of the dis- charge drainage design (i.e., direct versus indirect) and were also found to occur either upstream of the bridge deck, or at similar levels upstream and downstream, implicating sources other than bridge deck runoff.31 Stormwater Quantity Impacts from Bridges Bridge deck runoff quantity can be characterized by runoff volume and peak flow rate, both of which are considerations when evaluating the potential hydrologic effect of bridge deck runoff on receiving streams. NCDOT (2010) discussed how stormwater quantity from bridge deck runoff could neg- atively impact receiving streams.32 The construction of any new transportation facility, whether that facility includes a bridge deck or not, will increase imper- vious area in a watershed. Increasing impervious area increases both runoff volume and peak flow rates. These changes, if not properly mitigated, can negatively impact receiving streams by causing hydromodification, or the alteration of the hydrologic characteristics of a receiving stream that can negatively impact water quality (USEPA 2007). Some characteristics of hydromodification include increased movement and deposition of stream sediment, channel modifi- cation as receiving streams attempt to accommodate larger flows, stream bank erosion, increased stream turbidity, and changes in flow patterns. Such changes to the receiving stream can degrade water quality below intended uses and negatively impact biologi- cal habitat (USEPA 2007). Hydromodification should not be an issue however from bridge decks alone, since the runoff coefficient is identical to rainfall on the receiving water. NCDOT also discussed the risks of increased sediment deposition as runoff flows over- land from the bridge to the stream bank, increasing potential toxicity and reducing available sunlight, with detrimental impacts to aquatic communities. Design is a factor:33 For bridges that drain runoff via gutter flow and bridge end col- lectors or closed drainage systems, there is a possibility for erosion 30 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 8-5 31 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 4-44 32 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 2-2 33 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 2-2

A-8 to occur between the pipe outlet (typically located on the bridge embankment or near a bent) and the receiving stream. For bridges that discharge runoff from the deck using deck drains, lack of energy dissipation at the point of physical impact between run- off and the land beneath the bridge deck may also cause erosion. The likelihood of localized erosion can be verified through the use of professional judgment (e.g., bridge deck height) or ero- sion prediction through calculation of flow velocities in convey- ance and at points of discharge. Bridge Impacts at a Watershed Level Dupuis proposed comparison of pollutant loading from bridge decks to other sources in the watershed as one piece of evidence considered in assessing the potential impact of bridge deck runoff on the receiving waters (Dupuis, NCHRP 2002, vol. 2). Likewise, Malina et al. (2005a; 2005b) compared pollutant loads estimated for bridge decks in Texas to their receiving water loads and concluded that relative contribu- tions from bridge decks were very small and did not result in adverse impact to receiving waters. Watershed contributions were also among the most impor- tant identified in the NCDOT report given its emphasis on urban vs. rural differences.34 The hydrologic and water quality effects of increased storm- water runoff and pollutant loading on receiving waters have been well studied and documented, and these effects have been linked to land use change and urbanization (Burton and Pitt, 2001; Calder, 1993; Urbonas and Roesner, 1993). Effects of increased storm water runoff include stream bank and channel erosion, worsened flooding, and an increased ability for runoff to detach sediment and transport pollutants downstream. Effects of increased pollutant loading include eutrophication of receiving waters and subsequent hypoxia due to excessive nutrients, toxic- ity of aquatic life or inedible fish caused by loading of metals and organics, and limited contact recreation and shellfish consump- tion due to bacteria. In an effort to better mitigate these effects, the National Research Council has recently recommended a shift in stormwater management and regulatory permitting to a more watershed based approach, where discharge permits are based on watershed boundaries rather than political boundaries. (National Research Council 2008) To provide perspective on the relative contribution of runoff quantity and pollutant loads from bridge decks in North Carolina as compared to total watershed contribu- tions, NCDOT’s approach characterized runoff volume from bridges over waterways across the state and compared imper- vious area, runoff volume, peak flow rates, and pollutant loads estimated for selected bridge decks to those amounts estimated for their receiving waters. Three geographically distributed bridge sites with different watershed areas were selected for the site-specific evaluations. To respond to the trend of managing and regulating stormwater according to a more watershed-based approach (National Research Council 2008), NCDOT took a watershed-based perspective on runoff volume to weigh the hydrologic effect of bridge deck runoff.35 To provide this perspective, runoff volume and impervious area attributed to bridge decks were compared to total watershed contributions for three sites: Black River, Little River, and Swan- nanoa River . . . these three sites were also evaluated through direct comparison of concentration thresholds to measured end of pipe values and through mixing analysis. The three sites are spatially distributed in each of the three ecoregions in North Carolina and represent various sized watershed areas (i.e., stream drainage areas at the point of bridge crossing) . . . Deck area for all bridges in each watershed is a small fraction (below 0.05% in all cases) of the total watershed area. With the exception of the Swannanoa River site, the ratio of deck area for all bridges to total watershed areas is well below 1%. Overall, impervious area introduced by bridge decks in these watersheds is relatively small when compared to total imperious area and very small when compared to the total watershed area. NCDOT’s weight-of-the-evidence approach concluded that “bridge deck runoff does not have a widespread effect on receiving waters and that NCDOT’s current use of stormwater control measures for the mitigation of bridge deck runoff is protective of surface waters;” results indicated the following:36 • Quality and pollutant loading in bridge deck runoff is similar to roadway and urban runoff; bioassessments made upstream and downstream of bridges provided similar results; periodic toxicity of bridge deck runoff is possible, but not common (periodic toxicity observed may be linked to roadway deicers); • Bridge deck runoff did not contribute to stresses from organics or nutrient enrichment; • Potential erosion due to concentrated flow from bridge decks could impact receiving waters. • NCDOT currently implements structural stormwater con- trol measures (BMPs) to treat discharges to sensitive waters and BMPs to reduce potential erosion. Consequently, results of the study indicate that NCDOT’s current approach to BMP implementation is protective of state surface waters. Malina’s results were similar. Malina, et al. (2005) concluded that “mass loadings of constituents contributed by the run- 34 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 4-54-4-56 35 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 4-64 36 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 8-1

A-9 off from bridge decks were minimal compared to the mass loads of constituents carried by the respective receiving stream.” NCDOT pointed out that loads should be evaluated for meeting specific stormwater management goals, such as goals associated with waste load allocations, pollutant trad- ing, stormwater banking programs, or off-site mitigation, as required by a particular program or regulation. Linking Bridge Deck Runoff to Receiving Streams Linking bridge deck runoff to conditions in receiving streams is more difficult than measuring constituents in run- off. In its work for NCDOT, URS noted, “Despite a signifi- cant amount of stormwater characterization in the literature, no standard method exists for evaluating post-construction stormwater concentrations in an impairment context;”37 In general, results from a particular stormwater monitoring project are compared to national compendiums of stormwater data or to previous locally collected stormwater monitoring programs. While such comparisons are convenient for assess- ing stormwater runoff concentrations, they do not provide insight into the impacts of stormwater runoff on a particu- lar watershed. Logically, if a particular concentration is not contributing to impairment for a receiving stream with lower water quality standards, no significant stormwater treatment should be necessary. The same concentration profile might require sophisticated BMPs when paired with a high quality drinking water source. Therefore, efficient and cost-effective stormwater management, including BMP selection, becomes a function of evaluating stormwater characterization data against receiving stream surface water quality goals. Linking stormwater runoff to overall degradation in receiv- ing streams is an emerging area in stormwater management research. Fundamentally, it is understood that increased urbanization causes both hydrologic and water quality impairments to receiving streams (Burton and Pitt 2002). However, the specific processes and chemical pathways for the impact of stormwater runoff from transportation facili- ties, isolated from the impact of other nonpoint sources in the watershed are not currently well understood. Runoff Management on Bridge Decks Historically, bridge engineers have designed storm water drainage systems to drain directly into receiving waters through deck drains, scupper systems, or simply open-rail drainage. This was the low-cost, practical way to get water off the bridge quickly and maintain safe driving conditions. Vir- tually all bridges constructed in the United States still have these types of drainage systems. NCDOT’s 2010 report describes typical conveyance methods in the context of a bridge’s physical constraints, as follows: In general, as rain falls on a bridge deck, it drains in the direc- tion of roadway cross slope to the edge of the bridge deck. From there, runoff is conveyed by gutters and either exits through deck drains evenly spaced on the bridge deck or is conveyed off the bridge deck into grated inlets or other collection system. For some bridge drainage systems, runoff will free fall from the deck drains onto the roadway embankment, the overbank, or in some cases, directly into a body of water. Deck drains discharging directly into a water body is common on long coastal bridges, where collection and conveyance of stormwater is not feasible due to the size and cost of systems required. For older bridges, gutters and deck drains were not provided and runoff generally would sheet flow directly off the bridge deck onto the overbank or into a waterway. Requests to treat bridge deck runoff are becoming more common. Some state and local governments now encourage or require new projects to be constructed to drain runoff to land to allow for some form of active or passive improve- ment of the stormwater before it is discharged to the receiv- ing water or infiltrated into the ground without being directly discharged to the receiving water. USEPA has recommended diversion of runoff to land for treatment, restricted use of scupper drains on bridges less than 400 feet in length and on bridges crossing very sensitive ecosystems, or provision of equivalent urban runoff treatment in terms of pollutant load reduction elsewhere on the project to compensate for the loading discharged off the bridge.38 NCDOT effectively avoided the physical constraints of treatment by taking the latter option when the agency constructed a wetland in a rest area and treated 20+ acres to offset 14 bridges. NCDOT quan- tified the costs and benefits and showed this was much more cost-effective than retrofitting the bridges.39 Design Constraints with Stormwater Collection and Conveyance on Bridges General Discussion of Design Challenges Stormwater collection and conveyance is difficult on bridges. In the case of longer, flatter bridges, sufficient elevation does not exist to drain runoff by gravity for treatment at bridge approaches. Even when water can be drained to the 37 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 4-17 and 4-18 38 USEPA, Coastal Zone Act Reauthorization Amendments, B., Management Measure for Bridges http://water.epa.gov/polwaste/nps/czara/ch4-7b.cfm 39 Project interview with North Carolina DOT (Matt Lauffer), USGS (Chad Wagner), and URS Corp., December 20, 2012

A-10 abutment, on-deck spread takes valuable surface area and may require construction of a larger bridge deck, with the physical and carbon footprint that entails. Bridge girders, vis- ibility, and space constraints complicate piping. Maintenance is also more complex, dangerous, and expensive on bridges. Altogether, collection system cost and technical feasibility, in many cases, make conveyance of runoff to the abutment area for treatment impractical and raises questions about the ben- efits in relation to the cost. The current practice for mitigation of bridge deck runoff water quality via treatment typically adds a collection system to bridge deck drains to route the runoff to the abutment area. This is problematic for many installations. For long- span bridges, the conveyance system can become relatively large, introducing engineering, aesthetic and maintenance issues into the bridge design. For example, the design team completed a preliminary study for piping runoff to the abut- ment of the new San Francisco–Oakland Bay Bridge for treat- ment. The preliminary cost of this 16″ steel pipe system was estimated to be close to $4 million. In addition, many bridges have very low longitudinal grade, providing little slope to provide the necessary hydraulic gradient. Some bridges have a negative grade when crossing deeper canyons. Bridges with lifts can also pose problems. Locating BMPs in the touchdown (abutment) area can also be problematic. Space is at a premium and there may be geo- technical concerns with infiltration near the bridge supports or where slopes are steep down gradient. Areas adjacent to bridges often include sensitive riparian and wetland habitat. Areas in natural condition that are neither wetland nor home to threatened and endangered species can be undervalued and lost, even when biodiversity is high. Nevertheless, the use of conventional BMPs in the abutment area has been shown to be protective of receiving water beneficial uses.40 Detailed Discussion of Structural, Physical, and Spatial Constraints Structural, physical, and spatial constraint issues associ- ated with placing conveyance systems and runoff BMPs on bridges were summarized in NCHRP Report 474. Such issues included the following:41 • System Configurations. Typically, a stormwater con- veyance system is comprised of a number of deck inlets each connected to lateral pipe running transversely to the bridge. This lateral pipe conveys deck runoff to a main trunk line running longitudinally along the length of the bridge. The trunk line exits the bridge at the abutments and connects to a treatment system located nearby. The details of the conveyance will depend on the type of bridge under consideration. Given the large number of bridge types, there will be a wide variety of piping layouts, inlet sizes, and support systems. • Bridge Load. If large diameter piping will be required to convey the stormwater, the additional load to the bridge must be considered early in the bridge design. The AASHTO Bridge Design Specifications do not specifically mention this load; therefore, the designer will need to apply project-specific criteria to evaluate this load in com- bination with other bridge loads. • Inlets. Although inlets for bridge deck drainage systems are typically much smaller than inlets on roadway drain- age systems, they can create substantial conflicts with the structural design of the bridge. The inlet typically is cast into a concrete deck. This creates a conflict with the trans- verse and longitudinal deck-reinforcing steel; therefore, additional reinforcing may be required in these locations. Large inlets in positive longitudinal bending locations may necessitate analyzing the bridge deck using a reduced section modulus determined by subtracting the portion of the deck lost to the inlet. Decks with posttensioning steel require special consideration. Inlets can create con- flicts with both longitudinal and transverse posttension- ing. Relocation of longitudinal posttensioning in the field may not be possible. Anchorage zones for transverse post- tensioning may be adversely affected because inlets are typically placed at the edge of the deck where the anchor- age stresses are highest. These details must be considered at the design stage to avoid construction difficulties. Inlet design must consider the deck grooving and grinding that may be performed on bridge decks. This may necessitate casting the inlet below the top of deck level. Some agencies specify a minimum spacing of 10 feet for inlets. • Piping. The piping for bridge stormwater conveyance is typically much smaller than piping used for roadway drain- age systems, in which the minimum pipe diameter is often 18 inches. Bridge deck drainage systems typically incor- porate 6-inch-minimum-diameter pipes. Piping is usually located within the structure of the bridge to satisfy verti- cal clearance requirements. From an aesthetic standpoint, locating the pipes between the girders and thereby hid- ing them from view is beneficial. Piping is usually located inside of box girder bridges for the same reasons. To reduce clogging, large radius sweeps are used at bends in the piping (3-foot radius bends on 6-inch pipes is common). The designer should verify that sufficient space exists for these sweeps; otherwise the piping will conflict with girders or penetrate below the bridge depth. Because of the sweeps 40 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010) 41 NCHRP Report 474, Vol. 2, pp. 69–70

A-11 and the limited headroom available within the bridge, pipe slopes are essentially restricted to the bridge slope. Piping located near the top of vertical curves may have very slight slopes and small flow capacities. Care must be taken to avoid conflicts between the piping and the other utilities on the bridge. For durability and maintenance concerns, the piping must be strong. Welded steel pipe is commonly used. Structural analysis of the piping system is required to verify the pipe supports and to verify that the pipe can span between these supports. • Conflicts with Structural Members. The following struc- tural members can be adversely affected by the conveyance system: – Girder Webs. Penetrations in the girder webs are often necessary for the pipe laterals, because the inlets are nec- essary in complex piping systems and should be made accessible from safe locations. Access to expansion joints is especially important for their maintenance. – Maintenance Travelers. On bridges incorporating main- tenance travelers, the travelers will have to be designed to access the pipes and not to conflict with them. Coor- dination is required with compressed air piping to avoid conflicts. – Vents on Box Girder Bridges. Vents are provided on box girder bridges to allow air to circulate inside the bridge. Often these vents are only 4 inches in diameter. Design- ers should resize the vents or provide additional vents to pass the flow of a broken trunk line pipe within the bridge. For large diameter piping, a steel grate, similar to that used on bridges with pressure pipe water utilities, may be necessary. • Roadway Design. Consideration of bridge drainage and conveyance issues during the geometric design of the roadway will lead to simplified conveyance systems. Most importantly, avoiding sag curves and super-elevation rever- sals on the bridge will greatly reduce the number of inlets and the diameter of piping. Locating the high point of the bridge near the middle of its length may negate the need for inlets and piping on the bridge. Constant width bridges have simpler piping systems than tapered bridges and less likelihood of girder conflicts. All the flow upstream of the bridge should be intercepted to limit bypass flow from entering the bridge and having to be conveyed through the less-reliable bridge conveyance system. Most bridge designs restrict the amount of surface flow that may pass over the expansion joint between the approach slab and the superstructure. • Intermediate Diaphragms and Cross-Frames. The lon- gitudinal trunk line may conflict with these transverse members. • Bent Caps. Integral concrete caps are typically highly rein- forced and will have additional steel at the column/cap joint for joint shear requirements. Because piping often is directed down columns at the bents, the sweeping turns in the piping make this a difficult area to avoid reinforc- ing. Often the column transverse and main reinforcing are spaced more tightly than the diameter of the piping. • Columns. Pipes conveying stormwater down concrete columns typically exit the face of the column just above the footing. When a fixed connection between the column and the footing exists, the pipe will conflict with the trans- verse and main longitudinal column steel just as it does at the bent cap. This necessitates additional analysis and detailing. • Hinges. In concrete bridges, hinges in the superstructure are highly reinforced and experience high bending and shear stresses. Large diameter trunk lines are difficult to fit in this area. • Expansion Joints. Bridges with expansion capability at the abutments or in the spans will require compatible pipe expansion joints. These joints are typically of much larger diameter than the connecting pipe and are difficult to maintain. On very large bridges, the joint may be expected to move over 1 foot under temperature movements alone. Designing expansion joints for such large movements is difficult. An alternative to providing a mechanical joint at abutments is to construct a gapped system in which the piping directs stormwater downward from the super- structure into a small rectangular funnel-shaped reservoir located in the abutment seat. In this manner, the piping in the superstructure moves with the expansion or con- traction of the bridge above the small receiving reservoir, which is sized to always accept water from the piping. • Maintenance. Bridge stormwater conveyance systems, because of their small diameter piping and the nature of highway debris, create a challenge for maintenance staff. Repairing or replacing damaged or worn piping and com- ponents is difficult. This is especially true in enclosed box girder bridges because of restricted access, low working headroom, and low-light conditions. Access hatches or manholes are required in the top or bottom slab of box girder bridges, creating more locations for conflicts with rebar and posttensioning steel. DOTs report that there are other general problems with maintenance of collec- tion systems on bridges. Maintenance “elevated over the ground and adjacent to fast moving traffic” is a safety issue and a “risk to life.”42 In addition, the lifespan of pipes and systems “generously, is 10-20 years to replace the whole sys- tem versus a 50+ year bridge life,”43 making the issues and problems recurring ones, not just in maintenance, but in construction and finance. 42 Project interview with NCDOT, USGS, and URS, December 20, 2012 43 Project interview with NCDOT, USGS, and URS, December 20, 2012

A-12 • Limitation in Right-of-Way. There is no flexibility regard- ing the size of the foot print. There is no lateral right-of- way on which to build mitigation measures. Mitigation measures can be located on the bridge only at substantial cost, or stormwater must be gravity drained back to land. Other Perceived Conveyance System Challenges from DOT Practitioners Unexpected Environmental Impacts from Conveyance System Design In addition to a larger carbon and physical footprint when bridges have to be expanded to accommodate spread from curbside stormwater runoff, and impacts on subaquatic veg- etation, treatment areas sometimes encounter unexpected problems. For example, in North Carolina, a long coastal bridge (Virginia Dare Bridge) discharges to shellfish waters. NCDOT put in collection systems over the wetlands, to dis- charge into filtration basins; however, NCDOT found that the bacterial concentrations they were trying to prevent rose in the collection system because the warm water along with presence of trash and debris attracted animals that routinely produced waste.44 Additionally, two DOTs noted cases in which they had wid- ened a bridge deck in order to transport stormwater off the bridge without it spreading into the travel lane, when run- off could not be drained through scuppers. Undoubtedly, wider design concepts necessitate higher construction cost. NCDOT estimated the additional cost at around $120 per square foot.45 Other agencies, such as MassDOT, doubted that regulators would agree to a wider bridge as an answer, despite the potential that additional shaded area beneath the struc- ture would lessen the impact to regulated resources.46 Conflicts of Conveyance System Design with Stakeholder Interests and Public Aesthetics States on both coasts raised the high public expectations for DOTs to deliver aesthetically pleasing bridges. Maryland’s Chief of Hydraulics explained that on one of the bridges where they discussed having a collection system instead of scuppers, MDSHA proposed running the pipe in a box girder and devised a collection system maintenance approach. However, concerns were raised from the local boating com- munity about the aesthetics of the design and the potential for discharge of bird excrement.47 Thus, if MDSHA imple- mented the piped collection approach, they could possibly achieve nutrient reduction, but in the process, add pathogens. Ultimately, piping was eliminated because it was concluded that the benefits did not outweigh the costs. 48 In this case, MDSHA used a lip that allowed water to run off the deck to treat the first flush and the agency made the shoulders bigger to handle the spread. Consequently, transportation capacity was reduced49 Bridge Deck Runoff Mitigation Strategies Safety and other reasons require prompt removal of water from travel lanes on bridges. Thus, public safety concerns have dominated the discussion on drainage or runoff manage- ment until more recently. This section reviews how structural treatment BMPs and operational source control practices are typically incorporated into runoff mitigation strategies. DOTs tend to focus on the approaches to the bridge if on-site stormwater treatment is included or added to a project. While older bridges tend to drain untreated through the deck or scuppers, runoff from newer bridges that drain to sensitive waters or priority areas may be treated. Use of Structural Treatment Controls on Bridges Considerations and Limitations to Treatment Identified by DOTs In the interviews for NCHRP Project 25-42, DOTs identi- fied the following considerations related to runoff mitigation strategies: • Resource Agency Requirements/Specifications. Nearly every DOT interviewed said that the difficulties with on- bridge modifications are such that bridge runoff tends only to be treated if resource agencies specifically require it. For example, the Nebraska Department of Roads (NDOR) will treat bridge runoff “when it is requested by Game & Parks/Fish and Wildlife Service through project coordina- tion.”50 Likewise, the Louisiana Department of Transporta- 44 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012 45 URS/NCDOT 2010, Chapter 7. 7-9, note G 46 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012 47 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012 48 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012 49 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012 50 Personal communication with Gabe Robertson, Nebraska Department of Roads, December 17, 2012

A-13 tion Development (LADOTD) treats runoff “in accordance with permit.”51. TxDOT also said it treats bridge deck run- off only if there is a regulatory requirement to do so; “typi- cally, this is tied to 401 certification of very large Individual 404 permits (more than 1,000 linear feet or 3 acres of impact to waters of the US), a rare event.”52 In North Carolina, the decision to treat is based on specific considerations like water quality classifications of the waters to which the bridge discharges, any ESA issues, and whether the bridge is being newly constructed. Other regulations with which NCDOT must comply and which potentially drive treat- ment include the Clean Water Act 401 certifications, the NPDES program, state stormwater program, and state reg- ulations on nutrient sensitive waters, covering one-third of the state. For new bridges, state water quality agencies issue a 401 certification and have greater authority than on retrofit projects. Other DOTs (e.g., LADOTD, MassDOT) had no special or additional designs beyond the standard HEC 21 guidelines or the state’s stormwater handbook. States emphasized that designs were developed on a site- by-site basis as a result of requirements emerging in the environmental scoping process and what was considered the MEP treatment for the site. MDSHA also does “nothing different for bridges.”53 As previously discussed, in one case, MDSHA raised the lip height of scuppers to avoid direct discharge of the first flush. In other places MDSHA decided not to put in any scuppers if the bridge is small enough. DOTs may treat bridge deck runoff to the Maximum Extent Practicable (MEP), “just like roadway runoff,” but “options for bridge deck runoff are few” and “success in piping” is more possible on shorter spans, according to MassDOT.54 Thus, MEP is different for bridges than conventional roadway sections. MDSHA has also added BMPs at bridge approaches, since runoff is generally untreated in these areas. MDSHA has had the arrangement that “if the bridge surface is one acre, MDSHA needs to find one acre of approach roadway run- off to manage.”55 Managing runoff along the deck is generally considered the main design challenge. • Highly Sensitive Areas. Treatment of bridge deck runoff tends to be confined to highly sensitive areas. For example, LADOTD is transporting and treating runoff from only one (1) bridge site at this time, a case in which a bridge crosses a sensitive water body and drinking water supply.56 TxDOT and MassDOT also referenced the importance of drinking water supplies and treatment in those areas.57 • Bridge Girder Size and the Pipes that Can Be Accommo- dated. If they have to, some DOTs (e.g., FDOT, WSDOT) will pipe stormwater off bridge decks; however, girder size constrains the size of the pipes that can be used under them. For example, due to girder sizes in Washington State, “pipes have to go back and forth through the girders underneath the bridge and the pipe size is constrained,” preventing diversion of more than 91% of the 2-year storm in the sample case provided.58 • Spread and Bridge Elevation. If DOTs can get the water to the end of the bridge and a space where it can be sustain- ably treated, they will do so;59 however, if water spreads into the travel lane, it increases hydroplaning potential and risk of accidents. “On long flat bridges, the spread tends to open up very quickly and the DOT can’t always get the stormwater to the ends of the bridge.” 60 – Gutters: Some DOTs say they have been successfully using a gutter system, draining to detention areas or vegetative swales for treatment. Other research in North Carolina suggested that gutters might be implicated in concentration of pollutants.61 – Source Control like High Efficiency Sweeping: On some of the newer bridges where the DOT cannot get storm- water off the bridge and into treatment (e.g., WSDOT’s new floating bridge, which is very flat), the DOT is using source controls like high efficiency sweeping.62 • Ability to Perform Treatment on Roadsides along Bridge Approaches. Treatment at bridge approaches may include detention ponds, grass swales, or buffers, but treatment at bridge approaches is not always feasible. For example, MDSHA is treating bridge deck runoff to comply with state and federal requirements, in their case relating to imper- vious surfaces, but treatment near the bridge approach is infeasible in certain areas due to the 100-year floodplain and wetland regulation.63 In low-lying coastal areas, the 51 Project interview with Joubert Harris, LADOTD, January 3, 2013 52 Personal communication with Amy Foster, TxDOT, December 2012 53 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012 54 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012 55 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012 56 Project interview with Joubert Harris, LADOTD, January 3, 2013 57 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012. Also, personal communication with Amy Foster, TxDOT, December 18, 2012 58 Project interview with Mark Maurer, PLA, PE, Highway Runoff Program Man- ager, Washington State Department of Transportation, December 18, 2012, Proj- ect interview with Amy Tootle and Rich Renna, Florida DOT, December 18, 2012 59 Project interview with Amy Tootle and Rich Renna, Florida DOT, December 18, 2012 60 Project interview with Amy Tootle and Rich Renna, Florida DOT, December 18, 2012 61 Jy S. Wu and Craig J. Allen, Sampling and Testing of Stormwater Runoff from North Carolina Highways, 2001 62 Project interview with Mark Maurer, PLA, PE, Highway Runoff Program Man- ager, Washington State Department of Transportation, December 18, 2012, Proj- ect interview with Amy Tootle and Rich Renna, Florida DOT, December 18, 2012 63 Project interview with Karuna Pujara, Chief of Hydraulics, Maryland State Highway Administration (MDSHA) December 20, 2012

A-14 floodplain may be wide and wetlands extensive compared to the bridge project, in addition to the difficulties with draining water on long, flat bridges. Such cases are drivers in considering off-site mitigation. Stream buffer regula- tions can also restrict a DOT’s ability to treat stormwater at bridge approaches. NCDOT cited instances of buffer regulations where NCDOT “can’t discharge into Zone 1 (30 feet) and in some cases Zone 2.”64 Only two state DOTs interviewed (WSDOT and SCDOT) said they were treating bridge deck runoff in a vault on a site. WSDOT said they had a bridge (Riverton, WA) where they were using infiltration vaults to treat and infiltrate water off the bridge. • In a case over a shellfish area and outstanding resource water (ORW), SCDOT has a closed system and Stormceptor® device treating drainage from one direction (the other could be piped to an upland detention site); however, SCDOT said the closed system approach, “isn’t very prac- tical. Stormceptors® don’t do much to treat fecal coliform and might even exacerbate it. Sometimes rodents get in closed systems and make water quality worse.” • Availability of Off-site Mitigation. Consideration of off- site mitigation options is becoming a standard part of the process in Florida and Maryland. South Carolina is “try- ing to work out something based on surface area of the bridge.”65 – MDSHA and the Maryland Department of Environ- ment established a water quality bank that allows for permitting highway projects that cannot meet all storm- water water quality requirements. The water quality credit is established through off-site mitigation at the 6-digit HUC watershed level and the currency is acres of impervious surface treated. The positive balance in the bank is kept by implementation of various water quality projects designed to treat unmanaged impervi- ous surfaces. – FDOT tries to partner with co-permittees and “pay for offsite improvements.”66 FDOT is taking advantage of the current political environment to press for off-site treatment; last year, the state legislature passed a bill mandating that the state regulatory community allow flexible treatment approaches for transportation. That bill specifically named watershed level treatment and other strategies. Ultimately, FDOT expects that there will be stormwater banks just like mitigation banks. • Modeling to Show That Bridge Doesn’t Have Enough Surface Area and ADT to Have Detrimental Effects. South Carolina DOT is performing modeling to understand the impacts of bridge deck runoff,67 as is TxDOT. TxDOT has an ongoing project called “Contribution of Bridge Dwell- ing Birds to Bacterial Water Quality Impairments,” for which data is not available yet.68 TxDOT also sponsored a study called “Characterization of Stormwater Runoff from a Bridge Deck and Approach Highway: Effects on Receiv- ing Water Quality” in 2006. • Cost and Technology Development are factors in gener- ating on-bridge treatment solutions. FDOT is exploring further technology development, in particular bioactivated media, but the technology has not developed to the point of availability for use in bridge scuppers. The University of Central Florida is currently testing seven different medias, trying to achieve 1 gallon, per minute, per square foot load- ing through the media, which would enable the size to be reduced by about four times from where it is now. The tech- nology has been licensed and is in use in Florida, Michigan, and New Jersey, in up flow baffle boxes as a pre-treatment system, the bottom of retention ponds, bioswales, bio- retention areas, and improving water coming out of wet detention facilities, with favorable cost factors. Filter media cartridges currently developed last about 5 years, but with- out a cartridge, with more room and four times the material, life expectancy is 20 years.69 Further products are expected by May 2014 and Florida DOT is highly optimistic.70 • Understanding Resource Agency needs and where Treat- ment is Really Necessary. The NDOR mentioned the need for programmatic identification of critical areas, where treatment of runoff might really be necessary. If we, “under- stand when agencies will require treatment of bridge deck runoff, we can be more proactive in design,” said staff at NDOR.71 Despite the barriers discussed above, some DOTs said they were not encountering issues as they “only have to treat in very exceptional cases, such as over a public water supply (very uncommon) or if there is a regulatory require- ment to do so. Typically, this is tied to 401 certification of very large Individual 404 permits (more than 1,000 linear feet or 3 acres of impact to waters of the United States). This doesn’t happen very often.”72 64 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, December 20, 2012 65 Project interview with Sean Connolly, January 9, 2012 66 Project interview with Amy Tootle and Rich Renna, Florida DOT, December 2012 67 Project interview with Sean Connolly, South Carolina Department of Trans- portation, January 2013 68 Personal communication with Amy Foster, TxDOT, December 2012 69 Personal communication, Dr. Martin P. Wanielista, P. E., University of Central Florida, Orlando, Florida, January 31, 2012 70 Project interview with Amy Tootle and Rick Renna, Florida DOT, December 2012 71 Personal communication with Gabe Robertson, Nebraska Department of Roads, December 17, 2012 72 Personal communication, Amy Foster, Texas DOT, December 18, 2012

A-15 Nearly all DOTs contacted said that treatment for new construction projects is determined on a project-by-project basis with resource and regulatory agencies as part of the project planning phase. Where states consider retrofit mea- sures, those may be selected and designed through the DOT’s Highway Stormwater Retrofit Program to meet site-specific water quality goals.73 NCDOT avoids direct discharge off bridge decks whenever possible; they try to discharge to the overbank and collect and convey the stormwater to the stream in a manner that doesn’t cause erosion. On lower ADT secondary bridges, NCDOT is replacing the structures if needed and not adding stormwater treatment mechanisms.74 Level spreaders and energy dissi- paters in the overbank area are the most common method to treat stormwater.75 Higher-level treatment is provided in consultation with regulatory agencies.76 Nearly all of the DOTs contacted are dealing with bridge deck runoff on what they called a “case by case basis.” Treat- ment of bridge deck runoff is far from standard, due to the obstacles treatment entails and the relative benefit that treat- ment can produce. Unique Constraints Associated with Bridge Retrofit Nearly all interviewed DOTs note that retrofit of water qual- ity devices on a bridge is difficult. MDSHA has figured out a workable approach, swapping untreated bridge deck area for treatment of other currently untreated roadway. The state is also exploring off-site/off-alignment mitigation approaches. LADOTD said, “Funding and budgetary strategies are the big- ger challenges. Priority is usually given to new construction and routine retrofit/rehabilitation projects.” FDOT is trying to deal with the issue by partnering with co-permittees and “pay in lieu fees.”77 FDOT is focusing on off-site treatment; last year, the state legislature passed a bill mandating that the state regulatory community allow flexible treatment approaches for transportation. That bill specifically named watershed level treatment and other strategies. Ultimately, FDOT expects that there will be stormwater banks just like mitigation banks. Common Pollutant Removal Mechanisms Treatment BMPs NCHRP Report 565: Evaluation of Best Management Prac- tices for Highway Runoff Control (2006) describes some of the most common pollutant removal mechanisms for roadway runoff, which apply to runoff from bridges as well, especially where runoff can be routed for treatment off-site:78 • Sedimentation—Runoff is detained in a basin so that sus- pended solids and particulate-bound pollutants settle as a function of particle density, particle size, and fluid viscos- ity (under quiescent conditions) to the bottom of the water column. • Filtration and Infiltration—Runoff passes through an engineered media or existing soils where solids and particulate-bound pollutants are physically filtered by the media. If the media has adsorptive properties, dissolved pollutants may be entrained by the media as well. Treated runoff recharges groundwater supplies and reduces vol- umes delivered to receiving streams as surface flow. • Microbially Mediated Transformations—Runoff is con- tained in a microbially diverse environment (e.g., a storm- water wetland, vegetated basin). Microbes decompose and mineralize organic pollutants and transform inorganic pollutants before runoff is released. • Sorption—Runoff is contained in BMP systems (e.g., swales, filtration basins, stormwater wetlands) where substances of one state are incorporated into another substance (absorp- tion) or molecules are bonded onto the surface of another molecule (adsorption). • Uptake and Storage—Organic and inorganic constituents are removed from runoff by plants and microbes through nutrient uptake and bioaccumulation. Source Control Approaches for Bridges General Discussion Non-structural BMPs are often used as source control and management methods. Source control measures can be cost- effective and sometimes more efficient pollutant mitigation compared to treatment control practices. Alternative pave- ments, street sweeping, catch basin and scupper cleaning, deck drain cleaning, deicing controls, traffic management, management of hazardous materials, and spill prevention can be implemented without any structural burden on the bridge. 73 NCDOT, Stormwater Manual, 2008 74 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, Decem- ber 20, 2012 75 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, Decem- ber 20, 2012 76 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, Decem- ber 20, 2012 77 Project interview with Amy Tootle and Rick Renna, Florida DOT, Decem- ber 2012 78 Oregon State University, Geosyntec, et al., NCHRP Report 565, Evaluation of Best Management Practices for Highway Runoff Control (2006), cited in NCDOT, p. 2-10

A-16 Some DOTs confine source control to DOT operations only. For example, during construction and maintenance projects, LADOTD limits materials placed on bridges to only that neces- sary, with special attention to cleaning materials, solvents and/ or fuels. Only non-phosphate solutions are allowed for cleaning bridge structures. During deicing events, minimum amounts of deicing agents are used. MassDOT no longer uses sand for low traction conditions, and many DOTs have dramatically reduced sand usage, for both air and water quality purposes. Other DOTs are contemplating how passing vehicle sources could be better controlled, outside of reducing vehicle spray through greater use of PFC/OGFC. For example, NCDOT leaders would like to see how rumble strips prior to the bridge could shake off pollutants from the undercarriage of vehicles, to minimize pollutants being carried onto bridges and subse- quently falling on bridges or being sprayed off splash, during precipitation events. Engineers have noticed concentrations of oil and grease where there are irregularities in the roadway surface.79 Nearby on-site BMPs could be used to treat runoff from these areas. Alternative Pavements—Permeable Friction Course and/or Open Graded Friction Course (OGFC) Pavement PFC/OGFC may act as a source control for pollutants on the undercarriage and exterior of vehicles by greatly reducing the spray that occurs during precipitation events, and entrap- ment of particles within the overlay matrix. TxDOT and NCDOT have invested in research on the water quality ben- efits of PFC and/or OGFC pavement. Eck et al. and data from North Carolina indicated that the water quality benefits last as long as the structural life of the pavement, even though no maintenance at all was performed.80 NCDOT confirmed that as long as the road has speeds over 45 mph, pavement main- tenance can be avoided without a loss of permeability in the overlay. NCDOT has one more PFC research project under- way.81 WSDOT said they would consider OGFC on the wearing course but OGFC is susceptible to damage from studded tires. OGFC is being used and considered for use on roadway shoulders and for water quality treatment purposes, even where it is not used on the wearing course. NCHRP 25-25/82 will pro- vide design guidance for permeable pavements on roadway shoulders.82 Street Sweeping Street sweeping is one of the most common source control approaches in MS4s and some states are considering applying this measure outside of MS4s, where sufficient bridge height for flow to bridge ends is lacking. The water quality benefit of sweeping is somewhat controversial though. Schilling (2005) indicates that the direct benefit to stormwater qual- ity or effect on receiving waters of this sediment removal has not been conclusively defined.83 NCDOT’s 2010 report states that:84 Additional investigation is needed to establish the effective- ness of bridge sweeping as a BMP (BMP for stormwater) and to provide potential improvements to existing sweeping practices to benefit stormwater quality. NCDOT conducts sweeping practices for many existing bridges throughout the state because of the associated maintenance and safety benefits . . . NCDOT does not currently conduct bridge sweeping to specifically address storm- water quality concerns; . . . (however), because of the potential to remove large amounts of sediment, bridge sweeping should continue to be considered as a potential water quality Level II treatment BMP for bridge decks. Multiple DOTs are looking at bridge sweeping as a viable alternative for stormwater mitigation, particularly when other methods of treatment are not feasible or are cost-prohibitive, which may be the case for long coastal bridges. In addition, potential improvements to existing sweeping practices should be considered, including equipment upgrades and new train- ing for proper disposal of captured solids. Additional study is recommended to further evaluate sweeping as a BMP and to shape sweeping practices (including frequency, type of equipment, and disposal practices) to maximize the benefit for stormwater quality (NCDOT 2010). NCDOT has used sweeping as a negotiated stormwater control measure. On Currituck Bridge, it was not possible to install a collection system for technical reasons. Sweeping was agreed to as a source control measure with the regula- tory agency, with the PPP managing the bridge for 50 years. Sweeping will be performed on a 7-day rotation during the summer, after the peak traffic period, by the private operat- ing company. Other state transportation agencies, such as MDSHA, are trying to determine how they can squeeze the requested sweep- ing cycles (25 in MDSHA’s case) in during the non-freeze/ summer months.85 The anti-icing material is needed on the 79 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, December 20, 2012 80 Eck, Bradley, et al. Water Quality of Drainage from Permeable Friction Course, Journal of Environmental Engineering, ASCE, February 2012, pp. 174 81 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, December 20, 2012 82 Project interview with Mark Maurer, PLA, PE, Highway Runoff Program Manager Washington State Department of Transportation, December 18, 2012 83 Schilling, 2005 84 NCDOT, URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 8-2 85 Project interview with Russ Yurek, MDSHA Maintenance Manager, May 31, 2012

A-17 roadway November–April (when rain might freeze), so sweep- ing during those seasons is counterproductive. Thus, the agency would be “sweeping every week in the summer, and MDSHA would like to validate this interval,” so the agency is “working through the issue internally; they want to make sure they man- age this process well as a public agency.” 86 MDSHA needs to report the pounds of sediment collected by sweeping, which can be difficult for contractual processes going from one county line to another but crossing sev- eral watershed boundaries, before weighing and disposing. MDSHA’s Chief of Hydraulics posed several questions and comments: Do they stop the truck from watershed boundary to watershed boundary and go weigh? Do they pro-rate it? MDSHA supports highway sweeping but perhaps at a fre- quency of 2-3 times per year, an order of magnitude differ- ence than what the regulatory agency would like to see.87 MDSHA is also concerned that they cannot sweep from one watershed to another, meaning the DOT might be able to sweep only a portion of the drainage area in downtown Gaithersburg, contravening expectations that the DOT would sweep an entire stretch of road, such as Main Street. MDSHA is unsure whether they can physically do this type of sweep- ing at an economical level, in a way that is fair to the public. The fairness/justice issue and public expectations are very important because sweeping is a very visible DOT activity that “builds expectation because the public sees this occur- ring every two weeks.”88 There are a number of ways that sweeping may be applied at a DOT. Where sweeping is found to be practical and ben- eficial to deal with particulates, new high-efficiency street sweeping machines may be economical in urbanized areas89 and for some bridges that lack the vertical drop needed for drainage to bridge approaches. Other DOTs, such as FDOT, noted that sweeping was usually done by co-permittees. SCDOT utilizes a compliance matrix/checklist to identify where source controls apply. The evaluation is based on the amount of development in the watershed (e.g., above 5% impervious surface/development) and how much DOT right-of-way is in that watershed.90 If the agency is over a cer- tain percentage, they must perform sweeping and checks for (and removal of) dead animals. 91 Improved Sweeping Technology and Planning:92 • New vacuum-assisted and regenerative air sweepers (which blow air onto the pavement and immediately vacuum it back to entrain and filter out accumulated sol- ids) have greatly increased effectiveness, particularly with fine particles. In terms of improved sweeping methods, tandem sweeping, which is mechanical sweeping followed immediately by a vacuum-assisted machine have shown good increases in percent pollutant reductions (Sutherland and Jelen 1997). • Broom/vacuum combination. In recent studies, a new type of street-sweeping machine called the EnviroWhirl was found to be most effective, reducing TSS loading up to 90% for residential streets and up to 80% for major arte- rials. The actual percent reduction also depended on the number of cleanings per year, with the maximum reduction reported for weekly cleanings. Results for biweekly clean- ings are about 70% for both residential and major arterials. Deicing Controls Reduced salt usage is one of the most profound source control actions a DOT can take. For example, Caltrans implemented a reduced salt-use policy starting in October 1989 that required transportation districts to develop spe- cific route-by-route plans.93 That policy stated that, “Snow removal and ice control should be performed as necessary in order to facilitate the movement and safety of public traffic and should be done in accordance with the best management practices outlined herein with particular emphasis given to environmentally sensitive areas.”94 During the first winter, Caltrans reduced salt usage by 62% statewide as compared to the previous winter, helped by improved control of the appli- cation frequency of deicing salt.95 Alternative deicing practices and compounds that can reduce the loading include using alternative deicing com- pounds (e.g., calcium chloride or calcium magnesium ace- tate), designating “low salt” areas on bridges over sensitive receiving waters, and reducing deicing applications through operator education, training, and equipment calibration. In addition, using deicers such as glycol, urea or calcium 86 Project interview with Karuna Pujara, MDSHA Chief of Hydraulics, Decem- ber 20, 2012 87 Project interview with Karuna Pujara, MDSHA Chief of Hydraulics, Decem- ber 20, 2012 88 Project interview with Karuna Pujara, MDSHA Chief of Hydraulics, Decem- ber 20, 2012 89 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p.2 90 Project interview with Mark Giffin, Project Manager, South Carolina Depart- ment of Health and Environmental Conservation, January 7, 2013 91 Project interview with Mark Giffin, Project Manager, South Carolina Depart- ment of Health and Environmental Conservation, January 7, 2013 92 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 32 93 Venner, Marie, NCHRP 25-25/04: AASHTO Compendium of Environmental Stewardship Practices, Policies, and Procedures, 2004 94 California Department of Transportation, “The Use of Deicing Chemicals on California State Highways.” Caltrans Report to the Legislature in response to Chapter 318. (July, 1992), cited in Venner/AASHTO, 2004 95 California Department of Transportation, “Caltrans Snow and Ice Control Operations.” (March, 1999) 7 pp., www.dot.ca.gov/hq/roadinfo/snwicecontrol. pdf., cited in Venner/AASHTO, 2004

A-18 magnesium acetate (CMA) reduces the corrosion of metal bridge supports that can occur when salt is used. There are no effective methods for removing salt from stormwater, which points toward source control as the only viable alternative for mitigation.96 “Smart” in-vehicle application technology involving GPS and electronic sensing might make it feasible to use special deicers on bridges, or not use them at all, depending on the circumstances. 97 DOT Strategies for Handing Hazardous Material Contamination from Bridges The survey of state DOTs performed for NCHRP Report 474 found that the instances where treatment of bridge deck runoff was required involved drinking water supplies, cross- ing ORWs or national recreation areas, and/or concerns over endangered species or hazardous material spills.98 Spills Every DOT has a process in place for handling hazardous material spills. Most of these were focused on first responder protocols and on promptly containing and controlling the spill after the fact. Sometimes the DOT puts in a retention swale; a 20-foot long structure could include weirs and skim- mers. Florida DOT has instances of a skimmer collecting an entire gas spill.99 Several DOTs noted that “there are so many sensitive receptors out there (fish, wetland, water supplies)” that the DOT has placed a priority on water supplies.100 In the north- east, considerable land is reserved to protect water supplies. Instead of installing valves or boxes, MassDOT has detention areas and first responders have spill kits of plugs, caps, and booms; the focus is on getting first responders out there in a timely fashion. 101 To facilitate this, MassDOT helped develop a hazmat response storm drain Atlas that shows every seg- ment of highway with outfalls and catch basins:102 Instead of building a mechanical system that may fail, the first responders identify where the spill is going to discharge from the storm drain system Atlas and deploy caps, covers and plugs are used to try to contain the spill. There are some outfalls where the highway serves as the embankment of the reservoir. The DOT has catch basin hoods and deep sumps in these areas for containment. DOTs also maximize the use of swales. DOTs referenced the following practices to reduce the risk of contamination from spills off bridge decks: • MassDOT has tried to reduce risk by improving roadway geometry and site distance where applicable. In general, “If the road is straight, MassDOT would do nothing in par- ticular. If the receiving water was an essential water supply or critical in some way or there is higher probability of a spill the location will receive higher priority.” 103 • WSDOT also relies on absorbent booms that would be put around scuppers or drains, to prevent water contamina- tion. In some cases, spills have been directed to a wetland that could be cleaned up before contamination of the lake.104 • MDSHA noted that in one particular watershed they installed shut off valves on a riser structure so they could isolate the pond from the drainage system. • NCDOT has a policy on hazardous spills and Chapter 9 of their BMP toolbox addresses hazmat considerations. • NDOT noted that spills are documented through NDOR’s DIRK (Department Incident Reporting knowledgebase) system. • NCDOT installed a hazardous spill basin on Highway 64 in line with the flue gate. Special consideration has been strongly encouraged but not required, for mussels, often listed as threatened or endangered. Bridge Painting and Washing Practices As noted in NCHRP Report 474, bridge painting is prob- ably the most common bridge maintenance practice and the one with potentially the greatest adverse effects on the receiv- ing water.105 Blasting abrasives and paint chips from paint- ing activities may fall into the receiving waters below the bridge. Surveys have indicated that up to 80% of the bridges repainted each year were previously painted with lead paint. These surveys have also indicated that substantial amounts of 96 Talend, D., Salt: No Easy Answers. Stormwater, the Journal for Surface Water Quality Professionals 10 (7): 2009, pp. 16–28 97 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 2 98 Dupuis, T., National Cooperative Highway Research Program (NCHRP) Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volume 1 Final Report (2002), pp. 24–25 99 Project interview with Rick Renna, Florida Department of Transportation, December 2012 100 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012 101 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012 102 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012 103 Project interview with Alex Murray and Henry Barbaro, Massachusetts DOT (MassDOT), December 2012 104 Project interview with Mark Maurer, PLA, PE, Highway Runoff Program Manager, December 2012 105 Dupuis, T., National Cooperative Highway Research Program (NCHRP) Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volume 1 Final Report (2002)

A-19 used abrasives can be lost to the environment if appropriate containment practices are not used.106 Paint overspray and solvents also may be toxic to aquatic life if they reach the receiving water (Kramme 1985). An AASHTO Center for Environmental Excellence consulting engagement on environmental maintenance practices for the Kentucky Transportation Cabinet and AASHTO’s Environ- mental Stewardship Practices Guide describe bridge painting and washing practices to avoid and minimize environmen- tal contamination, in particular, impacts to water quality.107 Young et al. estimated that the costs of implementing mea- sures to reduce the effects of bridge painting on receiving water quality are an additional 10 to 20% for containment techniques and an additional 10 to 15% for waste disposal,108 both of which are accepted practices now.109 Bridge wash water also generally needs to be tested and/ or treated before being either discharged to the receiving water or otherwise controlled and managed off-site (NCHRP 2002). Recovery of wastes, containment of wastes, and train- ing of maintenance workers to increase their awareness of potential impacts on receiving waters are techniques that can be used to decrease the impacts of bridge maintenance activi- ties on receiving waters.110 Endangered Species Act Implications The ESA is a driver with treatment of bridge deck runoff in the Northwest, where salmonids are a consideration. At WSDOT, “most of WSDOT’s bridge deck treatment is driven by ESA. They have different triggers in the Highway Run- off Manual, which outlines the minimum.”111 The Highway Runoff Manual, the ESA Section 7 biological assessment or biological opinion or the state CWA 401 certification may outline extra requirements in sensitive areas. Sometimes, WSDOT will, “go someplace else to do equivalent area treat- ment; sometimes it is just too difficult to get the water off the bridge.”112 In the southeast, mussels are a consideration. NCDOT reported that the ESA has driven treatment of bridge deck runoff in the western part of North Carolina. The South Carolina DOT said the Carolina Heelsplitter (a mussel) has been an issue during construction but not post-construction. Generally, SCDOT, “complies with Corps requirements.”113 LADOTD noted that any ESA issues that arise in project development would lead to a special request that would then be accommodated by Hydraulic Design. This is consis- tent with other DOTs processes. Florida DOT said they have installed Manatee grates at outfalls but it is not influencing treatment of bridge deck runoff. The box turtle has triggered bridge deck runoff require- ments in Maryland. Threatened and endangered species are less of a factor in the northeast, where states recalled no known instances where the ESA drove treatment of bridge deck runoff. NDOR reported that,“the ESA seems to be the driver in requiring treatment of bridge deck runoff ”114 in the Midwest as well. Texas studies have often occurred in the same area as the Barton Springs salamander, an ESA- listed species. TxDOT has installed PFC to help address water quality and threatened and endangered species in sensitive areas. Other Miscellaneous Source Control Methods and Operation Control Measures on Bridges DOTs have shared a number of other source control prac- tices including the following:115 • High efficiency catch basin cleaning is being considered along with high efficiency sweeping in some states.116 NCDOT is exploring design-related stormwater control measures for bridge decks such as creating guidance on bridge materials that can reduce the concentration or load of pollutants in runoff that enters receiving streams. • Coatings for exposed galvanized metals. Bridge deck run- off studies in Texas found that exposed galvanized metal 106 Young, G. K., Stein, S., Cole, P., Kammer, T., Graziano, F., and Bank, F., “Evaluation and Management of Highway Runoff Water Quality.” Publica- tion No. FHWA-PD-96-032 (June 1996) 107 Venner and Kober, 2002, AASHTO CEE consulting engagement for the Kentucky Transportation Cabinet, Bridge Washing and Painting Practices 108 Young, G. K., Stein, S., Cole, P., Kammer, T., Graziano, F., and Bank, F., Evaluation and Management of Highway Runoff Water Quality.” Publication No. FHWA-PD-96-032 (June 1996) 109 Venner, M., AASHTO Compendium of Environmental Stewardship Practices, Ch. 7: Bridge Maintenance, 2004, online at transportation.environment.org http://environment.transportation.org/environmental_issues/construct_ maint_prac/compendium/manual/7_1.aspx. Also see: Venner and Kober, 2002, AASHTO CEE consulting engagement for the Kentucky Transportation Cabi- net, Bridge Washing and Painting Practices 110 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 34 111 Project interview with Mark Maurer, Washington State DOT, December 2012 112 Project interview with Mark Maurer, Washington State DOT, December 2012 113 Project interview with Sean Connolly, South Carolina DOT, January 9, 2013 114 Project interview with Gabe Robertson, Nebraska DOR, December 17, 2012 115 ADOT, Project interview with Mark Maurer, PLA, PE, Highway Runoff Pro- gram Manager Washington State Department of Transportation, December 18, 2012, Project interview with Amy Tootle and Rich Renna, Florida DOT, Decem- ber 18, 2012 116 Project interview with Mark Maurer, PLA, PE, Highway Runoff Program Manager Washington State Department of Transportation, December 18, 2012

A-20 railings were a source of zinc contamination in runoff. Coatings for exposed galvanized metals have the potential to reduce such discharges. • Configuration of the deck cross section (curbed or open at pavement level) was found to be a factor in a bridge deck runoff by Wu and Allen, as it could impact the buildup of pollutants. • Smart controllers in DOT equipment, applying herbicides and deicing materials using GPS to distribute the amount of chemical according to needs and can reduce the amount of the material that enters the environment and that is ulti- mately washed off of roads. • Catch basin cleaning practice and design that facili- tates cleaning. One design option consists of a series of trays, with the top tray serving as an initial sediment trap; the underlying trays filter out pollutants. 117 Michi- gan Council of Governments (SEMCOG 2009) describes another design option that uses filter fabric to remove pol- lutants from runoff.118 Frequency and consistency of cleaning improves performance. As with sweeping, it is important to remove accumulated sediments before those are flushed downstream. Comparative Effectiveness of BMP Types for Bridges NCDOT’s joint final report with USGS and the state Divi- sion of Water Quality on Stormwater Runoff from Bridges contains one of the most comprehensive summaries of BMP types for bridge deck runoff. BMP Types are categorized as shown in Table A-1. Each BMP summary contains the following information:119 • Snapshot table indicating the BMP category and ratings on identified characteristics based on available literature,120 general knowledge, and best engineering judgment. Judg- ment rationale included consideration of a BMP’s specific bridge application, experience with a particular BMP, understanding of unit processes, and general knowledge of relative costs. – Water Quality: In general, how effective is the BMP at reducing pollutant loads? – Volume Reduction: How well does the BMP reduce the inflow hydrograph volume? – Peak Rate Attenuation: How well does the BMP reduce the peak flow rate? – Groundwater Recharge: How well does the BMP replen- ish groundwater? – Cost: What are the construction costs relative to other BMPs? – Land Requirement: How large is the BMP footprint relative to other BMPs and what is the probability of right-of-way acquisition? – Possible Site Constraints: What is the relative probabil- ity of encountering issues with placement of the BMP at a bridge site based on the physical characteristics of the BMP? – Maintenance Burden: What is the relative level of effort, considering frequency, cost, and scope of maintenance activities required to keep the BMP functioning as intended? 117 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation (2010), p. 32 118 Southeast Michigan Council of Governments, SEMCOG. (2009). Catch Basin Cleaning www.semcog.org Statewide Transportation Improvements Program (STIP). Retrieved Table A-1. BMP types. 119 Adapted from NCDOT 5A-2 120 Particularly important sources were PADEP, 2006. Stormwater Best Manage- ment Practices Manual. Document Number 363-0300-002. Harrisburg, PA: Pennsylvania Department of Environmental Protection, Bureau of Watershed Management; NCDENR, 2007. Stormwater Best Management Practices Man- ual. Raleigh, NC: North Carolina Department of Environment and Natural Resources, Division of Water Quality

A-21 • Description. A brief overview of the BMP, including a description of the basic design concept and functionality. • Bridge Implementation. A description of how and where the BMP is typically put into practice at bridge sites. • Key Considerations. A summary of important informa- tion related to the design, construction, and maintenance of the BMP that should be considered in the selection process. • Cost. A summary of bridge implementation consider- ations follows, adapted from NCDOT’s Appendix 5-A of their joint final report with USGS and the state Division of Water Quality. Where noted, ratings are supplied by the following pre-existing documents:121 a. PADEP, 2006. Stormwater Best Management Practices Manual. 363-0300-002. Harrisburg, PA: Pennsylvania Department of Environmental Protection, Bureau of Watershed Mgmt. b. NCDENR. 2007. Stormwater Best Management Prac- tices Manual. Raleigh, NC: North Carolina Department of Environment and Natural Resources, Division of Water Quality. c. NCDOT. 2010. Stormwater Control Inspection and Maintenance Manual. HSP-2010-01. http://ncdot. org/programs/environment/stormwater/download/ SWControlInspectionMaintJan2010.pdf Bioretention basins can be located down- grade of a bridge deck. Bridge runoff can enter a basin via sheet flow, but typically runoff is directed to a basin via a bridge deck collection system. A pretreatment BMP (i.e., forebay) is recommended upstream of a bioreten- tion basin, especially in cases of high sediment load. Right- of-way acquisition may be necessary due to basin size, location relative to bridge, etc. Bioretention basins should be located where adequate sunlight is available for vegetation. Bioreten- tion basins may require watering during periods of drought and bioretention basins may require more maintenance initially while vegetation is being established. Maintenance frequency of a bioretention basin is a func- tion of the pollutant loads reaching the facility and the type of vegetation specified. In general, maintenance activities include maintaining vegetation; removal of trash, sediment, and debris; and cleaning/flushing of the underdrain system (when present). Periodic replacement of filter media may be required. Construction cost varies by basin size, which is based on drainage area and percent imperviousness. A 2012 NCDOT study of bioretention and swales found that both were effec- tive in treating bridge deck runoff to some degree, though bioretention had greater pollutant removal outside of TSS. When employing the percent load reduction metric, the small bioretention cell achieved 60-90% of the load reduc- tions that were achieved by the large cell for all pollutants except total phosphorus (TP), despite the fact that the small cell only captured 30% of the design storm.122 Undersized bioretention cells are a viable retrofit option to achieve hydrologic and pollutant removal goals since undersized cells achieve volume reductions comparable to full-size systems; however, it is unknown if the reduced benefit associated with an undersized system would justify its lower capital cost.123 Arizona DOT listed the following benefits: Can be very effec- tive for removing fine sediments, trace metals, nutrients, bac- teria and organics as well as suited for impervious areas and widely applicable to different climatic zones and limitations: (1) Pretreatment is necessary to avoid clogging; (2) Not suit- able in climates where soil can freeze; (3) Not recommended for upstream slopes greater than 20%; (4) Not suitable for distance to aquifers less than 6 feet; and (5) Bioretention BMPs can attract mosquitoes and other environmental nuisances.124 ADOT concluded that bioretention might be appropriate along facilities such as port-of-entries and rest areas. 125 Bridge Sweeping. When structural BMPs are found to be impractical, sweeping can provide a practical alternative.126 Bridge sweeping may be an attractive option for longer bridges where collection and con- veyance of stormwater to a treatment BMP is not feasible and costs are pro- hibitive. A program should be developed to optimize water quality benefits (sweeping frequency) rela- tive to costs. Proper traffic 121 Adapted from NCDOT 5A-2 122 S. K. Luell, R. J. Winston, C. E. Wilson, S. G. Kennedy, W. F. Hunt, “Retrofit- ting with Bioretention and a Swale to Treat Bridge Deck Stormwater Runoff,” NCDOT Research Project 2011-12, Final Report, North Carolina Department of Transportation, November 27, 2012, p. 87 123 S. K. Luell, R. J. Winston, C. E. Wilson, S. G. Kennedy, W. F. Hunt, “Retrofit- ting with Bioretention and a Swale to Treat Bridge Deck Stormwater Runoff,” NCDOT Research Project 2011-12, Final Report, North Carolina Department of Transportation, November 27, 2012, p. 88 124 Arizona DOT post-construction BMP manual, Appendix Table B-3, p. 160 of .pdf. 125 Arizona DOT post-construction BMP manual, Appendix Table B-3, p. 160 of .pdf. 126 Shoemaker, L., Lahlou, M., Doll, A, Cazenas, P. 2002. Stormwater Best Man- agement Practices in Ultra-Urban Setting: Selection and Monitoring. Washing- ton, DC: U.S. Department of Transportation, Federal Highway Administration. http://www.fhwa.dot.gov/environment/ultraurb/index.htm.

A-22 control practices must be executed throughout the sweeping process. Air quality impacts should be minimized by utilizing sweepers that include dust control mechanisms. Notably, the introduction of vacuum-assisted and regen- erative air sweepers (which blow air onto the pavement and immediately vacuum it back to entrain and filter out accumu- lated solids) has greatly increased effectiveness, particularly with fine particles. In addition, improved methods such as tandem sweeping (i.e., mechanical sweeping followed imme- diately by a vacuum-assisted machine) have shown marked increases in percent pollutant reductions.127 Bridge stormwater conveyance and collection (BSCAC) involves inspection of bridge stormwater con- veyance and collection systems with the pur- pose of identifying and documenting potential need for improvements to correct, minimize, or avoid erosion problems that could potentially impact receiving waters. NCDOT is considering incorporating this non-structural BMP into existing bridge maintenance inspections, which are per- formed every two years for all bridges in the state. To imple- ment BSCAC statewide, additional training of inspectors and additional effort during the inspection process would be needed for recognition and documentation of potential conveyance and collection problems. Improvement needs would then be forwarded to Hydrau- lics Division staff for prioritization, design, and implementa- tion of a solution and/or potential future retrofit. Catch basin inserts (CBI) can be implemented downgrade from a bridge deck collection system or in catch basins that receive bridge runoff or are con- nected to a bridge drain system. These devices are typically proprietary treatment that consists of a manufactured insert suspended in a catch basin or storm drain to filter pollutants (targeted and removed depending on the design and catch basin configuration). CBIs are a potential alternative for retrofit applications where available land is lim- ited. They cannot remove pollutants as well as other structural treatment BMPs and cannot effectively remove soluble pol- lutants or fine particles, according to USEPA.128 CBIs require frequent maintenance and can become a source of pollutants through re-suspension if not properly maintained, which consists of trash removal and removal of sediment (which may require use of a vactor truck) and/or replacement of a filter bag, cartridge or media. Traffic con- trol may be required for maintenance activities and collected material must be disposed of in accordance with current envi- ronmental regulations. Designed closed sys- tems (pipes) collect bridge deck runoff (a design- storm amount) and con- vey that to a point of discharge for purposes of stormwater management. Closed systems maintain hydraulic conveyance (spread) outside the travel lane and are typically composed of deck drains and hanging pipe systems, and are sometimes utilized for large or long new location or replacement bridges where deck conveyance is not practical. Treatment BMPs are likely to be needed at the outlet to dis- sipate discharge energy and prevent erosion and expansion fittings should be considered in the design at bridge joints and other locations. Significant maintenance burden should be anticipated, including removal of solids, trash, and debris; repairing sepa- rated or broken sections of pipe and eliminating clogs. Cost depends on the system configuration, number of expansion fittings, length, pipe size, pipe material, and other similar considerations. Conveyance chan- nels (stabilized channels to convey runoff from a bridge) prevent erosion and sedimentation by pro- viding a stable conveyance from a bridge to an energy dissipater or streambank structure. Conveyance channels can be implemented down- grade of a bridge deck and receive stormwater from a bridge deck collection system. They typically do not require additional right-of-way and should have a minimum design capacity to handle a 10-year storm event and consider the hydraulic capac- ity of upstream conveyances tributary to the channel. 127 Sutherland and Jelen, 1997, cited in NCHRP Report 474, volume 2, p. 69. 128 USEPA Office of Wastewater Management (OWM). National Menu of Storm- water Best Management Practices. Catch Basin Inserts Fact Sheet. USEPA. http:// www.epa.gov/npdes/stormwater/menuofbmps (Search Catch Basin Inserts).

A-23 Conveyance channels may be provided to collect runoff from bridge scuppers. They are typically lined with riprap since immediate stabilization is required. Maintenance activi- ties include removal of sediment, trash, and debris. Riprap may need to be added or replaced periodically. Construction costs are largely a function of excavation and grading costs and the cost of the lining material, and depend on channel length. Deck conveyance involves widening of the bridge deck to accommo- date collection and convey- ance of bridge deck runoff from a design storm, to keep runoff within the shoulder area and convey it to a treatment location in the abutment area. Though it involves construction of more deck/impervious surface area, deck conveyance is generally more cost effective than piping or closed systems on small new location or replacement bridges. Flow spread criteria must be considered in the widening design and appropriate collection, conveyance and treatment provided where deck conveyance reaches the end of the bridge. Costs include additional deck construction and main- tenance activities include removal of sediment, trash, and debris in the flow path. Dispersion is design to allow bridge deck runoff to discharge into the environ- ment without collection and conveyance. Instead of one or a few point source discharges from a collec- tion system, dispersion encourages diffuse flow over a large area by releasing flow from the bridge deck directly onto well-vegetated areas, open water, channels, or buffer zones. NCDOT considers dispersion on bridges from a height of 12 feet or more and where concentrated flow from other drain- age systems can be diverted away from the dispersion area.129 Dispersion reduces the need for additional ROW. Over ground, the cover beneath the bridge needs to be of a surface material that will withstand impact from dispersed runoff. In addition, the topography of the land receiving dispersed flow needs to encourage sheet flow so that re- concentration of runoff does not occur. The flow path from the bridge surface should be inspected to verify that flow is not obstructed. Maintenance needs can be assessed at the time of bridge inspections. Dispersion is likely to be the low- est cost management approach; since the cost of designing and installing scupper systems are incorporated into the cost of bridge structure construction, dispersion does not repre- sent a separate stormwater treatment expense. Dry detention basins can be constructed downslope of the bridge, receiving storm- water runoff from a bridge deck collection system, or as sheet flow from the bridge deck. A dry detention basin may also be used in series with other controls such as forebays, filter strips, or swales to meet pollutant removal efficiency requirements. Such basins temporarily collect and store stormwater runoff and gradually release it to a receiving stream. Dry detention basins attenuate peak flows, promote settlement of suspended solids and particulate-bound pollut- ants, and reduce erosive velocities downstream. These basins are typically designed to capture storm water and release it through a primary outlet control structure over a two to five day period and designed to remain dry in between storm events. ROW acquisition may be necessary due to basin size, location relative to bridge, and other pos- sible site constraints, and sediment basins that are used dur- ing construction can be converted into dry detention basins once construction is completed. The required minimum design surface area with a length-to-width ratio needs to be considered during site selection; construction cost varies by basin size, which is determined by drainage area and percent imperviousness. Vegetation maintenance (mowing) may be required. Energy dissipaters (e.g., riprap in preformed scour holes or rock aprons) can be implemented downgrade of a bridge deck and can receive stormwater from several sources including bridge deck collection sys- tem, underneath bridge scuppers, or downgrade of another BMP. They should be designed to reduce velocity of the out- fall to a non-erosive rate for the design storm of the contrib- uting facility or the 10-year event, and should be sited on level grade, where possible. At minimum, the downgrade edge of the dissipater must be level perpendicular to the flow line. Maintenance activities include removal of sediment, trash, and debris. Riprap may need to be added or replaced periodically. 129 Henderson, D. R. 2002. Bridge Deck Drains Memo to Hydraulics Project Managers, NCDOT. Dated July 18, 2002. North Carolina Department of Trans- portation, Hydraulics Unit. http://www.ncdot.org/doh/preconstruct/highway/ hydro/pdf/BridgeDeckDrainGuidelinesJuly02.pdf

A-24 Environmentally sensi- tive design (ESD) utilizes natural topography down- grade of a bridge deck to receive stormwater from a bridge deck collection sys- tem. ESD techniques infil- trate, filter, store, evaporate, and detain runoff close to its source and promote the natu- ral movement of water within a watershed. The Maryland Department of Environment (MDE) has characterized ESD as “a comprehensive design strategy for maintaining pre- development runoff characteristics and protecting natural resources” (MDE 2009). North Carolina also promotes this strategy, focusing on utilizing existing areas, natural or previously disturbed, to treat stormwater with little or no modification. A dry deten- tion basin created using a naturally existing depressed area may be described as an environmental design basin. The natural topography needs to match the final graded needs of the BMP to which ESD is being applied. (In most cases, energy dissipation will be needed upgrade of a natural ESD.) In addition to energy dissipation, other ESDs may require a slight retrofitting such as an outlet structure. ESDs reduce construction effort and cost as well as require less mainte- nance in most cases. Vegetated Filter strips can be implemented down slope of a bridge deck, under neath the bridge deck, or along bridge deck embankments in areas receiving storm water runoff from a bridge deck runoff collection system, or via sheet flow. For instances where runoff is supplied by a bridge deck collection system, a level spreader, preformed scour hole, or weir is required to promote diffuse flow upgrade of the filter strip. Sheet flow across the filter strip is treated through infiltration into the soil. Filter strips are typi- cally located in series with other devices that promote diffuse flow, such as level spreaders or preformed scour holes. Any natural vegetated area may be adapted for use as a filter strip, though additional right-of-way may be required to pro- vide sufficient flow length and gradient. Vegetation should reasonably tolerate standing water, resist erosion, resist exces- sive bending when subject to runoff flows, and be as dense as possible. Soil and groundwater conditions that allow for high infiltration rates will increase the water quality benefits. Filter strips are relatively inexpensive BMPs with the majority of the associated cost due to grading and planting costs; how- ever, filter strips should be periodically inspected and main- tained to sustain good vegetative cover and to remove rills formed by concentrated flow and excessive sediment deposi- tion. The formation and maintenance of sheet flow across the filter strip is critical to their successful operation. Filtration basins can be installed downgrade of the bridge, receiving storm- water runoff from a bridge deck collection system or as sheet flow from the bridge deck; however, a pretreat- ment BMP is recommended upstream of a filtration basin, especially in cases of high solids load, as this treatment-type BMP detains and routes stormwater through filter media. As stormwater infiltrates through the amended soil, sand, or engi- neered media of the filter, pollutants are filtered and adsorbed onto particles. Stormwater vacates the basin through an under- drain system and is directed back to the receiving stream. A filtration basin can be used in areas where the soils are not suitable for infiltration systems. They may require addi- tional ROW due to the filter size, location relative to bridge, and other possible site constraints. Filtration basins have underdrain systems (designed to resist clogging) with clean- outs to facilitate inspection and maintenance activities. The filter bed will have an outlet control device to collect under- drain flows and direct flow to the receiving stream, typically designed to discharge flow above the prescribed treatment elevation. Maintenance frequency of a filtration basin is a function of the pollutant loads reaching the facility. In gen- eral, maintenance activities include removal of trash, sedi- ment, and debris and cleaning/flushing of the underdrain system. Periodic replacement of filter media may be required. Infiltration basins can be implemented down- grade from the outlet of a bridge deck runoff collec- tion system or underneath the bridge deck, where the hydraulic conductivity of the site soils is adequate for infil- tration. Underdrain systems are not incorporated. Pollutant removal capacity can be high because most pollutants associated with water quality volume are filtered or adsorbed by surficial soils, though highly soluble pollutants may persist in groundwater. Due to the size of infil- tration basins, additional right-of-way may be necessary. If the existing soil does not have a high infiltration rate, the surface area of the basin may become prohibitively large. Site soils must be able to infiltrate stormwater in the basin within a specific period of time. Pretreatment with an upstream BMP is necessary to remove solids that can clog the infil- tration basin and reduce the infiltration rate. Care must be

A-25 taken during construction activities to protect the infiltration basin area from construction traffic, material laydown, and other activities that can compact soils and reduce infiltra- tion capacity. Construction cost varies by basin size, which is based on drainage area and percent imperviousness, and hydraulic capacity of in-situ soils. The frequency of mainte- nance will largely depend on the pollutant loads to the infil- tration basin. DOTs mush remove debris, trash, and sediment buildup from the basin as necessary to maintain the perme- ability of the soil. Arizona DOT’s evaluation of BMPs favored infiltration for treating runoff collected at a single point as it can theoretically achieve 100% removal of dissolved and col- loidal pollutants to surface water bodies while reducing peak flows and eliminating downstream bank erosion.130 However, ADOT uses infiltration with caution due to sev- eral known limitations, including (1) High failure rates due to improper siting, design, and lack of maintenance, especially when no pretreatment is included; (2) Clogging likely under high suspended solid loading; (3) Lack of suitability below steep slopes; (4) Possible groundwater contamination; violation of Aquifer Protection Permit (APP) standards; and (5) Requiring complete stabilization of upstream drainage area. 131 Level spreaders can be implemented downgrade of a bridge deck or under- neath the bridge deck with both applications receiving stormwater runoff from a bridge deck collection sys- tem. A level spreader can also be used in series with other treatment options to optimize hydraulic and water quality benefits, providing a non-erosive outlet for runoff by distribut- ing concentrated water uniformly across a large area of a stable slope (dispersion). The structure consists of a level concrete or vegetated trough with a non-erosive lip that discharges into a vegetated area. Level spreaders can also be used in series if the slope is too steep from the trough to the stream. Verification of sur- ficial soils may be necessary using soil surveys or existing geotechnical reports to determine if site can support a vege- tated trough. Maintenance activities include removal of sedi- ment, trash, and debris from the level spreader trough and other components. Inspections must confirm that the level spreader lip is not damaged and that flow is not bypassing the lip as the lip must create uniform, diffuse flow for the con- trol to function properly. Construction cost is a function of type of level spreader (concrete lip or vegetated trough) and length/configuration. The spreader lip must be absolutely level to operate correctly; this requirement is often difficult to achieve in practice. Preformed scour holes (PFSH) or pre-shaped, riprap-lined basins can be located directly downgrade of an outfall or underneath the bridge deck. The basin is stabilized with filter fab- ric and riprap to absorb the impact of the discharge (i.e., energy dissipation) and to prevent additional erosion. Once runoff has filled the shallow basin, it overtops the pre- formed scour hole and is redistributed as diffuse flow to the surrounding area. By inducing diffuse flow conditions, pre- formed scour holes reduce downgrade erosion and promote infiltration and filtration in the flat, vegetated land surface. PFSHs require a small footprint and typically do not require additional right-of-way. The ground downgrade of the PFSH must be at a gradient that maintains a non-erosive flow velocity and diffuse flow. If the PFSH is used for energy dissipation only, runoff can exit the preformed scour hole to an alluvial channel. Maintenance activities include removal of sediment, trash, and debris. The riprap base should be inspected to ensure that no rock has been dislodged or removed. Off-site stormwater mit- igation is a design-related BMP that describes a system of providing offsite treat- ment to compensate for sites where treatment is not practicable or when there is a larger environmental and economic benefit from implementing stormwater controls in other areas of the watershed. Bridge sites may be limited by site constraints that reduce BMP construction feasibility and negatively impact cost-effectiveness. Off-site stormwater mitigation provides an avenue for stormwater control to be provided elsewhere in the watershed where implementation is more practicable. With proper planning, offsite stormwater mitigation can reduce construction and maintenance costs. An accounting program to track offsite mitigation activities must be devel- oped. If desired, the multiple other ecological benefits of off- site mitigation can also be tracked. A DOT may further reduce costs through advance planning and coordination with mul- tiple projects and units of government. Off-site mitigation can be somewhat unique in its capacity to deliver high levels of water quality, volume reduction, peak-rate attenuation, and groundwater recharge at low cost. 130 Arizona DOT post-construction BMP manual, Appendix Table B-3, p. 160 of .pdf 131 Arizona DOT post-construction BMP manual, Appendix Table B-3, p. 160 of .pdf

A-26 Stormwater wetlands can be implemented down- grade from the outlet of a bridge deck runoff collec- tion system. These engi- neered wetlands with dense vegetation remove pol- lutants primarily through biological processes, evapotranspiration and infiltration. Stormwater wetlands sometimes evolve from failed infiltra- tion basins.132 Like infiltration basins, stormwater wetlands improve both water quality and help mimic pre-development hydrology; the extent of the improvement in water quality depends on the soils and vegetation. Stormwater wetlands designed specifically for storm- water treatment are distinguished from naturally-occurring wetlands by having distinct inlet and outlet structures. Like infiltration basins, right-of-way acquisition may be neces- sary as stormwater wetlands require a large surface area. Soils with low permeability should be present because a constant water level should be maintained in the wetland. If soils have a high permeability, a liner will be needed. Dry weather flow may be necessary to keep vegetation alive. Plant selection in stormwater wetlands is critical, and it should be specified by an appropriate professional. Maintenance activities include removal of trash, dead or undesirable vegetation, and debris. Dead or dying vegetation and undesirable vegetation should be removed. A 2012 study found high bacterial counts at the out- fall of a treatment wetland designed for bridge deck runoff. The researchers concluded that the wetlands are being utilized by raccoons, squirrels, deer, birds and other wildlife and that their feces were re-contaminating the water, negating any bacterial removal that the wetland initially provides to the water enter- ing it though “the wetlands might provide other nutrient or water quality enhancement benefits such as serving as settling areas for sediments or removing nutrients or other pollutants from the stormwater runoff coming from the bridge deck.”133 Other limitations are: possible release of nutrients during the fall season and discharges from constructed wetlands may be warmer than the temperature of receiving surface water body (heat sink effect).134 ADOT listed the following benefits for wetlands: high aesthetic value; improved treatment over dry detention and retention; flood attenuation; reduction of peak flows; and limits downstream bank erosion.135 Stream bank drop struc- tures are implemented downslope at the discharge point of collection and conveyance facilities of the bridge to safely convey bridge deck and/or road- way runoff into a water- way. The structure minimizes erosion caused by concentrated storm-water flows when existing surface cover does not provide adequate protection, preventing sedimentation in the receiving water and thus the bridge impact on the stream. Stream bank drop structures generally consist of riprap or concrete sloped or vertical drops to locally protect the stream bank from erosion. The contributing factors to stream bank erosion must be evaluated and identified in order to select the most appropri- ate stabilization method; the possibility of utilizing vegeta- tive stabilization in conjunction with structural stabilization should be evaluated. At minimum, structures should be designed for the 10-year storm event. The hydraulic capacity of upstream conveyances should be considered in the design. Swales typically have denser vegetation and flatter slopes than most flood man- agement drainage channels. Swales can be implemented downgrade of a bridge deck and receive stormwater from a bridge deck collection sys- tem. These broad and shal- low channels with dense vegetation convey and treat peak runoff by decreasing stormwater runoff velocity and promot- ing infiltration and physical filtration. They fit well in linear areas, usually along roadways and medians, and tend to be better suited to smaller drainage areas due to the maximum allowable discharge velocities. Check dams may be required depending on the longitudinal slope. Swales are typically sized to treat frequently-occurring storm events and the length is generally related to the size of the drainage area. Maintenance activities include removal of sediment, trash, and debris and the repair of eroded areas and sometimes mowing for aesthetic purposes. Costs tend to be low (mostly excavation during construction). When incorporated into roadway or facility design as part of the conveyance system, swales can provide water quality benefits and be aesthetically pleasing. Swales remove coarse sediment better than fine sediment and since highway stormwater runoff is expected to contain relatively coarse sediment,136 132 NCHRP 25-40 interview with Karuna Pujara, Maryland SHA Hydraulics Chief, May 31, 2011 133 Fleckenstein, E., Final Report for Task Order #40: Bacterial Sampling— NCCF, Preliminary Assessment of Potential Bacterial Loading off the Virginia Dare Bridge in Dare County, NC. 2010 134 Arizona DOT Post-Construction BMP Manual, p. 160 of the .pdf document. Appendix B. Table B-3 135 Arizona DOT Post-Construction BMP Manual, p. 160 of the .pdf document. Appendix B. Table B-3 136 Sansalone, J. J., Koran, J. M., Smithson, J.A, and Buchberger, S. G. (1998). Physical characteristics of urban roadway solids transported during rain events. Journal of Environmental Engineering, 124(5), 427-440

A-27 swales are an appropriate means of TSS reduction from high- way runoff.137 More information about swales is available in NCDOT’s Bioretention and Swale study.138 Emerging BMPs for Potential Use on Bridges Interviewed DOTs staff expressed interest in emerging BMPs and alternative treatment approaches given the diffi- culties inherent in treating on-site with known methods. Most DOTs did not have emerging treatment mechanisms to share with others, though some said they were “working on it.” Emerging BMPs identified in the interview process and lit- erature review included the following: Planning Mechanisms • Louisiana is performing a statewide assessment of water quality at existing and future bridge crossings to assist with the development of runoff management strategies. The state regulatory agency is undertaking the study, to which LADOTD will add bridges, so that the agencies can identify bridge crossings over sensitive waters. LADOTD plans to proactively address bridge deck runoff wherever possible, in those areas. • Some DOTs have identified off-site mitigation areas where treatment could be accomplished more efficiently on a watershed level. Florida DOT has taken the lead in convening the state’s Water Management Districts to talk about cooperative stormwater opportunities with the state DOT.139 FDOT also has a provision for off-site compensa- tion (See Attachment A-1). North Carolina’s Division of Water Quality asked NCDOT to develop a proposal for off- site mitigation as a solution for effective watershed storm- water mitigation where BMPs for bridges are problematic or not practicable, where more stormwater mitigation could be gained for dollar spent, and where retrofit proj- ects are to be constructed. Off-site stormwater mitigation and treatment practices are currently implemented in other states, such as California, Delaware, Florida, and Maryland. Alternatives to Treatment of Bridge Deck Runoff • Stormwater re-use by municipalities. This would involve municipalities taking FDOT’s stormwater and re-using it for irrigation or groundwater recharge. FDOT is working on stormwater re-use statewide, to team up with munici- palities to do stormwater re-use. Dispersion • Dispersion on vegetation. In the mountains on relatively high bridges, NCDOT has used the scupper dispersion method where vegetation underneath acts as an interceptor. • Equivalent area treatment enables a DOT to treat a sim- ilar area to the (then untreated) bridge deck in another untreated spot in the corridor. This is a standard approach in Maryland. • Investment in off-site mitigation, whether regional treat- ment or natural resource areas that offer filtration benefits for pollutants of concern, may be accepted by regulators in exchange for avoidance of less cost-effective on-site treatment. As an innovative approach, Washington has developed a watershed-based process for addressing stormwater (and other resource impacts) that includes leveraging funds for higher- priority local stormwater projects, water quality enhancement at an off-site wetland, and cost sharing on regional treatment off site. Other states that mentioned they use compensat- ing mitigation include Rhode Island, Maine, Massachusetts, and Delaware. According to a memorandum of understand- ing between the Delaware DOT and the state environmental agency, stormwater banking is used by Delaware for non- bridge construction projects to reduce the inefficient use of small mitigation systems. For example, one large pond may be constructed to mitigate other stormwater sources (highway or urban). The ultimate outcome of stormwater banking and compensating mitigation is the overall reduction of pollutant loads to a watershed. Furthermore, the cost is lower, and the mitigation systems used are typically more effective. Filtration Technologies • Bio-sorption activated media. FDOT is working with the University of Central Florida on bio-sorption activated media to absorb nutrients, an approach that is “showing great promise,” according to FDOT hydraulics and water quality staff.140 Bio-sorption Activated Media (BAM) can control nutrients, metals, and bacteria and requires less area to accomplish stormwater management relative to other options.141 BAM materials have the dual characteristic of sorption properties as well as sites for biological growth.142 At this point, BAM has residence time issues and limited 137 NCDOT bioretention and swale study, 2012, p. 11 138 NCDOT bioretention and swale study, 2012, pp. 22–25 139 Project interview with Amy Tootle and Rich Renna, Florida DOT, Decem- ber 18, 2012 140 Project interview with Amy Tootle and Rich Renna, Florida DOT, December 18, 2012 141 Martin Wanielista, Ni-Bin Chang, Manoj Chopra et al., Demonstration Proj- ect for Bio-sorption Activated Media for Ultra-Urban Stormwater Treatment, submitted to FDOT, May 2012. p. 2 142 Martin Wanielista, Ni-Bin Chang, Manoj Chopra et al., Demonstration Proj- ect for Bio-sorption Activated Media for Ultra-Urban Stormwater Treatment, submitted to FDOT, May 2012. p. 2

A-28 ability to handle larger flows, but FDOT notes that “bridge inlets don’t intercept large flow rates. FDOT might use bio- activated media on a bridge deck if the agency is pressed to perform retrofits.”143 Researchers anticipate being able to raise the flow rate by a factor of four in the next year, for bridge applications.144 FDOT has already supported research using BAM in retention areas such as swales and pipe-in-pipe wet detention pond harvesting applications.145 The technol- ogy could conceivably be used for ultra-urban environments or instead of a bridge collection system. FDOT would like to see the technology evolve to the point it is “plug and play.”146 FDOT and university researchers are also performing testing to determine the lifespan of BAM for the removal of both nitrogen and phosphorus compounds and then demonstra- tion projects to document pollutant removal. • Pier cap wetland treatment areas. WSDOT investigated pier cap treatment areas up to the point of testing, which would have cost an additional $500,000. FHWA funded the first two phases of this research and then decided the idea did not have broad enough appeal. Nevertheless, WSDOT believes this approach, which involves the construction of small wetland treatment area around bridge pier/pile caps, has potential. • New product evaluation. LADOTD and other DOTs noted they continually evaluate new products for con- sideration and placement on the departments “Qualified Products List.” Methods and Tools to Identify and Select Appropriate Mitigation Strategies for Bridge Deck Runoff NCHRP Report 474 Methodologies for Evaluating Need for Mitigation of Bridge Deck Runoff and Kind of Treatment NCHRP Project 25-13(01) developed a “practitioner’s” process for DOTs to address whether bridge deck storm water runoff will affect the receiving water; if runoff does have an impact, whether mitigation is necessary; and if mitigation is necessary, what kind is needed, considering: (1) state and federal regulatory requirements; (2) state and federal regula- tory agency and interested party concerns with the impact of the bridge—for example, water quality, spills, and endan- gered species; (3) receiving water characteristics and desig- nated uses, particularly high quality waters and Outstanding National Resource Waters; and (4) bridge deck characteris- tics.147 When combined, these factors underscore the point of many of the DOTs interviewed for NCHRP Project 25-42: that consideration of bridge deck runoff was considered and negotiated on a case by case basis, because every bridge and deck discharge condition and multiple uses of a receiving water is unique. The NCHRP Project 25-13(01) practitioner’s guide reports the conclusion that:148 Low-traffic rural highways do not cause significant adverse effects on aquatic biota . . . Similarly; many highway agencies do not oppose avoiding direct discharge of stormwater to receiv- ing waters in cases in which bridges are small enough or conve- niently enough configured for that to be accomplished at a rea- sonable cost and without compromising public safety. Again, the research team does not propose that the process described here should supersede such rational, commonsense approaches, nor should it cause the practitioner to employ a more complicated process than necessary to come to a judgment. NCHRP Project 25-13(01) developed 19 general method- ologies for bridge deck runoff analysis and mitigation, which are summarized in Appendix A-3. Method 13: Nonstructural and Structural BMP Evaluation (has been determined that some type of BMP may need to be implemented for bridge deck runoff) is more relevant to the current discussion and thus is described in more detail here. Precursor methods to determine if mitigation is necessary (e.g., calculation of pol- lutant concentrations at zone of initial dilution, bio-criteria, sediment pollution accumulation model, loading analysis, in situ toxicity testing, comparison to other source loadings in the watershed) are described in Appendix A-3. As NCHRP Project 25-13 points out, “nonstructural, structural, or insti- tutional BMP approaches, or a combination thereof, could be implemented.”149 Many factors must be considered in selecting a BMP approach, including the BMP capabilities and limitations, appropriateness for the site, pollutant loading benefits, maintenance requirements, and cost. Safety is also a factor. For example, snoopers will be used to maintain below deck piping, and a confined space entry will be required to main- tain piping that is located within the bridge deck structure. 143 Project interview with Amy Tootle and Rich Renna, Florida DOT, Decem- ber 18, 2012 144 Project interview, Dr. Martin P. Wanielista, P. E., January 31, 2013 145 FDOT research project BDK78 977-02 146 Project interview with Amy Tootle and Rich Renna, Florida DOT, Decem- ber 18, 2012 147 Dupuis, T. V., Pilgrim, K., Mischuk, M., Strum, M., Abere, D., and Bills, G., Research Results Digest 235: Assessment of Impacts of Bridge Deck Runoff Contaminants on Receiving Waters. Transportation Research Board, National Research Council, Washington, DC (January 1999) 148 Dupuis, T., National Cooperative Highway Research Program (NCHRP) Report 474: Assessing the Impacts of Bridge Deck Runoff Contaminants in Receiving Waters, Volume 2, Practitioner’s Guide, p. 3 149 NCHRP Report 474, Vol. 2, p. 23, 69

A-29 The final report, published as NCHRP Report 474, advo- cates that, “Nonstructural mitigation techniques should always be considered before structural measures because they are cost-effective and sometimes more efficient pollutant removers.”150 The practitioner’s guide provides a simplified evaluation process to lead the practitioner through the BMP selection process.151 Further information is available in the NCHRP Report 474, Volumes 1 and 2 (see also Figure A-1).152 • Step 1. Define the need (e.g., heavy metals concentration reduction, discharge elimination). • Step 2. Define the constraints (e.g., site, cost, and organiza- tional and physical constraints). • Step 3. Eliminate obviously inappropriate techniques (e.g., if the concern is hazardous material spills only, street sweeping will not address the concern). • Step 4. Begin evaluation of nonstructural BMPs. One by one, determine the benefit and cost of each technique and answer the following: Will the technique achieve the required water quality benefit in whole or in part? Project benefits are based on projected pollutant reduction and/or projected flow reduction. Determine costs for BMPs using literature values or other internal estimates. Nonstructural BMPs that are potentially applicable to bridges include: – Street sweeping – Inlet box/catch basin maintenance – Maintenance management – Deicing controls – Traffic management (e.g., high occupancy vehicle lanes, and mass transit). • Step 5. If one or a combination of several nonstructural BMPs would not achieve the required benefits, begin eval- uation of institutional BMPs (i.e., pollutant trading and mitigation banking). Evaluate whether either of these tech- niques, or a combination of any techniques evaluated up to this point, would achieve the desired water quality benefit. Determine costs of the institutional BMPs. • Step 6. If the nonstructural and institutional BMPs can- not provide the desired water quality benefit/protection, structural BMPs should be evaluated to determine which methods are appropriate and to assess the cost-effectiveness of potential methods. A critical component of the BMP analysis includes engineering evaluations related to the type of drainage and stormwater conveyance needed, and the effects these systems could have on the structural design of the bridge. In selecting an appropriate BMP, required pollutant removal benefits, site constraints, main- tenance constraints, and potential environmental or aes- thetic enhancements need to be considered (Dorman et al. 1996; Shoemaker et al. 2000; Young et al. 1996, Table 33). Once a narrowed list of BMPs is selected, the costs for each should be calculated and an appropriate economic analysis made (Brown and Schueler 1997). Appropriate BMPs may include simple drainage back to land for relatively small bridges in cases in which this is practical. In most cases, the drainage would be to a grassy area or pond prior to discharge to the receiving water. NCHRP Project 25-13 advises the practitioner to, “deter- mine if pollutant trading, off-site mitigation, and mitigation banking exist in the appropriate geographic context (i.e., usu- ally within the watershed) of the bridge project. If none exist, it may be beneficial to consider establishing one . . . ”153 Gathering such information places the bridge runoff in con- text. The loading from the bridge can be estimated by model- ing pollutant loads or collection of site-specific runoff quality data (NCHRP Report 474 v. 2 Methods 11 and 12). Ideally, information on the watershed and waterway can be ascer- tained from the resource or regulatory agency. EPA’s “How’s My Waterway” website and application (http://www.epa.gov/ mywaterway) uses GPS technology or a user-entered zip code or city name to provide information about the quality of local water bodies, including the water’s status and the type of pol- lution reported for that waterway, as well as what states and EPA have done to reduce pollution. Additional reports and technical information is available for many waterways.154 Dupuis et al., recommend using these and other watershed specific information from USGS (see WATERS, ATTAINS, and the National Atlas of Sustainability—“Atlas”). If needed, more in-depth modeling can draw on EPA’s Better Assess- ment Science Integrating point and Nonpoint Sources (BASINS) modeling framework (http://www.epa.gov/ost water/BASINS/). BASINS is a multipurpose environmental analysis system designed for use by regional, state, and local agencies in performing watershed and water quality-based studies. Geographic information supports analysis of a vari- ety of pollutants at multiple scales and point and nonpoint pollution management. The web-based BASINS enables watershed and water quality analyses drawing on national databases; evaluation tools for evaluating water quality and point source loadings at a variety of scales; utilities includ- ing local data import, land use and DEM reclassification, watershed delineation, and management of water qual- ity observation data; watershed and water quality models including PLOAD, NPSM (HSPF), SWAT, TOXIROUTE, and QUAL2E; and post-processing output tools for interpreting 150 NCHRP Report 474, Vol. 2, p. 67 151 NCHRP Report 474, Vol. 2, pp. 69–70 152 NCHRP Report 474, Vol. 2, pp. 69–70 153 NCHRP Report 474, Vol. 2, p. 67 154 USEPA, “How’s My Waterway,” http://www.epa.gov/mywaterway

Figure A-1. Process drivers and treatment identification: NCHRP Report 474, bridge deck runoff practitioner’s guide.

A-31 model results.155 BASINS has an open-source MapWindow GIS interface, a Data Download Tool, project builder, water- shed delineation routines, and data analysis and model output visualization tools as well as plug-in interfaces for well-known watershed and water quality models SWMM5, WASP7, and SWAT 2005. It includes a data extractor, projector, project builder, GIS interface, various GIS-based tools, a series of models, and custom databases; a web data extractor provides a tool for dynamic downloading of GIS data and databases from the BASINS web site and a variety of other sources.156 The user specifies a geographic area of interest and the soft- ware downloads appropriate data from EPA, USGS and other locations on the Internet. After the GIS data are downloaded, they are automatically extracted, projected to the user specified map projection, and a project file (“.apr” for ArcView/BASINS 3.1 and “.mwprj” for MapWindow/BASINS 4.0) is built. This Web Data Download tool then allows the user to add additional data to the BASINS project from a variety of data sources, and to check for more recent data and updates as appropriate. The Auto- mated Geospatial Watershed Assessment (AGWA) tool features the USDA-ARS models KINEROS and SWAT. The Kinematic Runoff and Erosion Model (KINEROS) is an event oriented, physically based model that may be used to determine the effects of various artificial features such as urban developments, small detention reservoirs, or lined channels on flood hydrographs and sediment yield. Rosgen’s Bank Erosion Hazard Index has been incorporated in the pollutant loading model as PLOAD-BEHI; this model is useful for simplified analyses of sediment issues. AQUATOX receives and automatically formats output from HPSF or SWAT in order to integrate watershed analysis with the likely effects on the aquatic biota in receiving waters. The new Parameter Estimation (PEST) tool in WinHSPF automates the model calibration process and allows users to quantify the uncertainty associated with specific model predictions. Considerable guidance exists on BMP selection as noted in the NCHRP Project 25-40 literature review.157 Available resources include the BMP database, online at http://www. bmpdatabase.org with BMP performance data, cost data, BMP monitoring guidance, and protocols for BMP perfor- mance assessment and the proprietary BMP performance data more uniquely accessible through the Massachusetts Stormwater Evaluation Project (MASTEP) at http://www. mastep.net. BMP selection and design strategies are discussed in Strecker et al. (2000),158 for work conducted in the evalua- tion and testing of monitoring equipment and strategies for highway runoff for the FHWA. Other guidance documents including Strecker et al. (2005) and NCHRP (2006) provide general guidance on BMP selection and design.159 Most state DOTs or state environment agencies have developed catalogs and/or fact sheets of treatment BMPs, such as those included in the AASHTO Compendium of Environmental Stewardship Practices (2004) and guidance and manuals issued by Caltrans, WSDOT, and NCDOT, with information on BMP perfor- mance, cost, space requirement, suitability, and/or mainte- nance requirements that can be useful selecting BMPs.160 Typically, no single answer exists to the question of which BMP (or BMPs) should be selected for a site; there are usually multiple solutions ranging from standalone BMPs to treat- ment trains of multiple BMPs to achieve the water quality objectives within physical site constraints. The first step in BMP selection is identification of physical characteristics of a site including topography, soils, contributing drainage area, groundwater, base flows, wetlands, existing drainage ways, and development conditions in the tributary watershed (e.g., construction activity).161 DOTs use physical variables (slope, area, velocity) to select and prioritize potential BMPs, in the design process. As the Denver Urban Drainage and Flood Control District notes:162 Maintenance should be considered early in the planning and design phase. Even when BMPs are thoughtfully designed and properly installed, they can become eyesores, breed mosquitoes, and cease to function if not properly maintained. BMPs can be more effectively maintained when they are designed to allow easy access for inspection and maintenance and to take into consideration factors such as property ownership, easements, visibility from easily accessible points, slope, vehicle access, and other factors . . . Costs are a fundamental consideration for BMP selection, but often the evaluation of costs during planning and design phases of a project focuses narrowly on up-front, capi- tal costs. A more holistic evaluation of life-cycle costs including operation, maintenance and rehabilitation is prudent . . . Designers are advised to “fully consider how and with what equipment BMPs will be maintained in the future” and 155 NCHRP Report 474, Vol. 2, p. 77 156 USEPA Basins website, Better Assessment Science Integrating point and Non- point Sources, http://water.epa.gov/scitech/datait/models/basins/basinsv3.cfm 157 NCHRP 25-40 Interim Report (Literature Review Results), June 2012. 158 Strecker, E. W., Mayo, L., Quigley, M. M., and J. Howell (2000) Guidance Manual for Monitoring Highway Runoff Water Quality. Federal Highway Administra- tion, Unpublished Draft, Contract DTFH651-94-C-00108. FHWA (August 2001). Urban Drainage Design Manual, Hydraulic Engineering Circular No. 22, Second Edition, prepared by S. A. Brown, S. M. Stein, and J. C. Warner. Federal Highway Administration, Washington, D.C. 159 Strecker, E. W., W.C. Huber, J. P. Heaney, D. Bodine, J. J. Sansalone, M. M. Quigley, D. Pankani, M. Leisenring, and P. Thayumanavan (2005). Critical Assessment of Stormwater Treatment and Control Selection Issues, Water Envi- ronment Research Foundation (WERF); Report No. 02-SW-1. ISBN 1-84339- 741-2. 290p. 160 Venner, M., Compendium of Environmental Stewardship Practices in Con- struction and Maintenance, AASHTO, 2004. NCHRP 25-25/04, maintained online at AASHTO’s Center for Environmental Excellence. 161 Urban Drainage and Flood Control District, Urban Storm Drainage Cri- teria Manual Volume 3, August 2011 http://www.udfcd.org/downloads/pdf/ critmanual/Volume%203%20PDFs/USDCM%20Volume%203.pdf, p. 2-13. 162 Urban Drainage and Flood Control District, Urban Storm Drainage Cri- teria Manual Volume 3, August 2011 http://www.udfcd.org/downloads/pdf/ critmanual/Volume%203%20PDFs/USDCM%20Volume%203.pdf, p. 2-13

A-32 acquire “clear, legally binding written agreements assigning maintenance responsibilities and committing adequate funds for maintenance.” Sustainability of performance over time is a design con- sideration, whether the amount of supplemental irrigation required for the chosen vegetation or clogging of infiltration BMPs when there is upstream development. Design of BMPs must be informed by operation and maintenance perfor- mance in the field. NCHRP Report 565 on BMP Selection NCHRP Report 565 advocated the practice of considering BMP unit operations (i.e., treatment mechanisms or pro- cesses) in the design and selection of structural BMPs and BMP treatment trains. Physical processes, such as sedimenta- tion and filtration, can be used to remove a significant por- tion of the pollutant load when a pollutant is predominately particulate bound. However, more complex chemical and biological unit operations may be required to treat pollutants that are dissolved or readily change from within the aque- ous environment as a function of redox, pH, and available partitioning sites (i.e., solids load or media characteristics).163 Historically, BMP selection and comparison involved cal- culating pollutant removal efficiencies, or the ratio of efflu- ent concentration to influent concentration expressed as a percentage; however, Geosyntec noted that this concept of “effectiveness” has key shortcomings:164 • Pollutant removal efficiencies for many BMPs that remove and sequester pollutants are largely a function of the influ- ent stormwater pollutant profile (Wright Water Engineers and Geosyntec Consultants 2007; CASQA 2003; USEPA 2009b). Since influent stormwater conditions can be site specific and are rarely verified by monitoring, using this criteria alone as an estimate of effectiveness may not be appropriate. • Comparing pollutant removal of BMPs that use volume reduction as the primary unit operation (e.g., infiltration basins) to BMPs that promote sedimentation or filtration can be difficult, since volume reduction BMPs remove a portion of the pollution load instead of reducing the concentration. • BMPs that provide diffuse flow or otherwise prevent ero- sion have the potential to provide a widespread water qual- ity benefit, but their effectiveness is not easily benchmarked against other BMPs. It can be complicated to quantitatively compare the theoretical load prevented (e.g., erosion pre- vented by riprap in the bridge overbank) to the theoreti- cal load removed (e.g., solids removed in a dry detention basin). Current research focuses on developing procedures for selecting BMPs based on compiled irreducible concentra- tions and well-defined receiving stream goals (such as benthic macroinvertebrate health ratings).165 Until widely accepted procedures exist for identifying effective BMPs based on a distribution of effluent concentrations proven to protect receiving stream quality, many regulatory agencies require BMPs based on surrogate strategies including mandating certain BMP types under certain circumstances, providing assumed pollutant removal credits for BMPs based on type, and assumed surface water quality protection for a suite of BMPs specific to certain receiving stream classifications or sensitive watersheds.166 DOT Mitigation Methods and BMP Selection Strategies Most DOTs contacted for this study said that they have not developed methods or tools to identify and select mitigation strategies for bridge deck runoff. DOTs concurred that the primary challenge was getting the water off the bridge for treatment; a number of DOTs said that a wide range of con- ventional stormwater treatment BMPs would be considered once flow is conveyed to the abutment area. Selection Processes and Matrices for Traditional Roadside Post-Construction BMPs Selection processes and matrices for general roadside post-construction BMPs are not uncommon at DOTs. An example selection matrix from MassDOT is included in Attachment A-1. Likewise, Arizona DOT’s Post-Construction BMP Manual has matrices for the following evaluation by BMP type: Site Specific Considerations: • Area typically served (acres) • Percent of site area required for BMP (%) • Configuration • Soils • Minimum hydraulic head (ft) • Maximum upstream slopes (%) 163 OSU et al., Evaluation of Best Management Practices for Highway Runoff Control, NCHRP Report 565, 2006 164 OSU et al., Evaluation of Best Management Practices for Highway Runoff Control, NCHRP Report 565, 2006 165 McNett et al., 2010 cited in NCDOT/URS, 2010 166 NCDENR, 2009; NRC, 2008. cited in NCDOT/URS, 2010

A-33 • Fracturing geology present • Minimum depth to groundwater • Approximate percent (%) removal efficiencies for select parameters • Safety reference(s) Environmental Stewardship Considerations • Urban areas • Setback requirements • Streambank erosion • Sensitive water bodies • Sensitive wildlife habitats • References Climatic Zone Restrictions • Peak flow reduction • Temperature extremes (cold climate, arid semi-arid climates) Georgia DOT’s Stormwater Management Manual out- lines a similar screening process to assist the site designer and design engineer in BMP selection. Georgia DOT considers the following factors, in order: • Stormwater Treatment Suitability: Capability to provide water quality treatment, downstream channel protection, overbank flood protection, and extreme flood protection). • Water Quality Performance: Ability to accept hotspot runoff and provide TSS, nutrient and/or bacteria removal). • Site Applicability: Drainage area, space required (space consumed), slope, minimum head (elevation difference from inflow to outflow—particularly important in the case of bridges), and water table. • Implementation Considerations: Including construction cost and maintenance level of effort are considered. Since watershed considerations are not seen as often in these matrices and since NCHRP Report 474 advocates watershed level context analysis for consideration of tradeoffs, Georgia DOT’s table for watershed considerations is included in Table A-2. Georgia DOT’s matrices for each of the above are summa- rized in a final matrix in Table A-3. Interviewed States Point to Case-by-Case, Negotiated Mitigation Strategies The DOTs interviewed for NCHRP Project 25-42 did not have bridge deck runoff specific BMP selection matrices. Rather, resource agency requirements tended to guide and to instigate design for treatment of bridge deck runoff. Florida DOT said their BMP/mitigation selection tool consisted of a simple four step process: 1. Drain it off the bridge and get it to a collection system. 2. Direct discharge 3. Compensatory treatment 4. Last choice: collection system using fiberglass pipe. WSDOT’s stormwater management plan contains a pri- oritization process for where the agency will do retrofits, taking into account sensitive areas and where the agencies can achieve the greatest benefit. Every project must evalu- ate whether they are triggering minimum requirements. NCDOT’s “Merger Process” with state and federal resource agencies targets natural resource mitigation to places in the watershed where they will accomplish the greatest environ- mental good:167 The Merger Process allows for a site-specific stormwater control measure (SCM/BMP) selection process to address the environmental concerns of the various agencies. Site-specific goals for stormwater control measures (SCM) should be based on regional water resource management strategies and should be linked to the designated uses and water quality standards of receiving waters (National Research Council 2008). How- ever, linking stormwater discharges from bridges to receiving water effects can be difficult. Future efforts should continue to develop a process that more closely links stormwater discharges from bridges to receiving water effects and bases SCM selection on their effectiveness. Both TxDOT and LADOTD commented that mitigation strategies and/or BMPs are primarily driven by environ- mental regulations such as permits, administrative orders, etc., with consideration of the sensitivity of the water body. Even North Carolina, which has arguably invested the most in an analytical approach for bridge deck runoff treatment, stressed that they identify and select appropriate mitigation strategies for bridge deck runoff on a case by case basis, in a negotiated fashion.168 URS concluded:169 As more is known about the relationship between pollutant gen- eration, BMP function, and receiving stream effects, stormwater management programs can be developed that select BMPs based on more definitive effectiveness criteria. Using the best available information collected at this time, quantitatively determining the ability of a BMP included in this study to meet receiving stream or water quality objectives, and thus determine its effective- ness, is not feasible. Since determination of site-specific water 167 NCDOT/URS p. 8-2 168 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, Decem- ber 20, 2012 169 NCDOT/URS, 2010, p. 6-2

A-34 quality goals is also not feasible without extensive studies, the (team) relied upon existing water quality regulations established by NCDENR to define sensitive receiving waters. Thus, the assumption of effectiveness for a BMP type is determined by its ability to be applied to a specific location discharging to a sensi- tive stream as defined by NCDENR’s water quality regulations (emphasis added). NCDOT stresses that the process of determining a BMP’s effectiveness for a particular bridge or roadway project is not straightforward, for several reasons:170 • It is generally not possible to identify site-specific pollutants- of-concern (POCs) at each bridge or roadway project. Because many of the storm event characteristics that influ- ence pollutant load are related to variable and unpre- dictable precipitation attributes (intensity, duration, antecedent dry periods), properly characterizing storm- water runoff at a site requires a monitoring program that captures a number of storm events. This sort of monitor- ing is time and cost intensive and is not feasible at every project site. • Even when monitoring data is available, it can be difficult to understand the significance of the magnitude of POC concentrations. The relationship between end-of-pipe concentrations of typical POCs and the degree of resulting impact on a receiving stream is poorly understood (Burton and Pitt 2002). For example, many of the identified POCs in this study had concentrations elevated above surface water quality thresholds, but no expression of toxicity was identified for concurrent bioassays. Exceedances of thresh- olds do not necessarily signify that BMPs are needed to protect water quality. Table A-2. Georgia DOT physiographic and watershed considerations for BMPs. 170 NCDOT/URS. Stormwater Runoff from Bridges, Final Report, July 2010, p. 6-2

A-35 • BMPs are not typically selected based on identified POCs (see first bullet) and are still widely evaluated based on pol- lutant removal efficiency and not the irreducible concentra- tion, as is recommended in the International Stormwater BMP Database (Wright Water Engineers and Geosyntec Consultants, 2007). As previously discussed, pollutant removal efficiency is largely a function of influent con- centration. Because site-specific POCs are generally not identified, nor their typical concentrations known, pol- lutant removal efficiencies do not provide useful infor- mation for determining concentrations reduced or mass load removed for a particular POC. Further, discussions of BMP pollutant removal efficiency or effectiveness without understanding project-specific water quality goals does not provide useful information on BMP performance. • For impaired receiving streams that have been subject to a TMDL where a particular load reduction target has been established, the load reductions are provided in terms of an annual mass load reduction. However, because it is gen- erally difficult to predict the annual mass load removed from a BMP without knowing (1) the influent pollutant profile, (2) typical effluent concentrations from the BMP, and (3) hydrologic characteristics that determine flow rates and volume in advance, it is difficult to accurately select BMPs that can effectively meet TMDL requirements. NCDOT decided that Stormwater Management Plans would be required for all new location and replacement bridge projects.171 • Level I treatment BMPs discussed in the NCDOT treat- ment report were recommended to be implemented on projects under certain conditions: – The bridge project crosses a water body on the 303(d) list as maintained by NCDENR. – The bridge project is located in a TMDL area; treatment requirements will be determined in accordance with Part III, Section C of NCDOT’s NDPES permit (see discussion in section 6.3). – The bridge project is located in an endangered species area and through biological assessments, the U.S. Fish and Wildlife Service (USFWS) has rendered a biologi- cal opinion that a Level I treatment BMP is required to mitigate potential impacts. Table A-3. Georgia DOT general application structural control alternatives for BMPs. 171 NCDOT/URS, 6-22 and 6-23

A-36 – The bridge project is located as part of a roadway with anticipated average daily traffic greater than 30,000 vehi- cles per day. – The bridge project is a new location bridge and located in a water quality sensitive area. – The bridge project is a replacement bridge that is wid- ened more than one travel lane and located in a water quality sensitive area. – Requiring BMPs on projects with an anticipated average daily traffic of 30,000 vehicles per day or higher is not cur- rently included as part of NCDOT’s post-construction stormwater program (PCSP). However, NCDOT does currently focus retrofit implementation in areas where facilities cross sensitive streams with high ADT loads (NCDOT 2008b). This ADT split of 30,000 vehicles per day originates from an FHWA study that showed road- way sites with ADTs higher than this benchmark had higher stormwater pollutant loads than lower ADT sites (FHWA 1990). The researchers theorized that ADT did not directly affect pollutant loads, but might be an indi- cator of atmospheric quality differences between urban and rural land uses. In addition, ADT is currently used to determine BMP treatment requirements for other departments of transportation (WSDOT 2008). The use of ADT as an indicator of pollutant load is still being evaluated in the literature, and statistical analysis of pro- visional bridge runoff data suggests that a roadway site’s urban or rural classification per the FHWA Functional Classification Guidelines may be a more appropriate indicator of pollutant load. For the purposes of devel- oping a statewide BMP cost-estimate, the use of ADT to determine BMP needs is an appropriate estimating tool. However, the use of ADT as a trigger for BMP treatment on a project-by-project basis should be investigated fur- ther before being incorporated into the PCSP for all types of transportation runoff. • Level II treatment BMPs will be implemented on all projects. • Maintenance BMP will be implemented for all projects fol- lowing construction and concurrent with routine bridge maintenance activities. • Bridge sweeping (Maintenance BMP) will continue to be implemented as appropriate (further investigations on implementation for water quality preservation and pro- tection are needed). • Design-related BMPs will be implemented as appropriate to support the no-direct discharge policy or stormwater mitigation. NCDOT’s 2010 report states that: For bridges where water quality of bridge deck runoff may not be a concern, use of scupper drains to disperse runoff over a large area could be a significant cost savings when compared to implementation of deck conveyance or collection systems (installed to support NCDOT’s no-direct discharge policy); these savings could be significant for long coastal bridges and, if combined with off-site stormwater mitigation, could result in a more effective water quality benefit. NCDOT should com- plete investigations into the applicability of dispersion of bridge deck runoff, including developing with DWQ a specific bridge criteria where dispersion of bridge deck runoff is an acceptable practice and assessing the effects on overbank areas, wetlands, and receiving waters.172 This process represents an evolution of the primary requirements outlined in Chapter 9 of NCDOT’s Stormwater BMP Toolbox and no-direct discharge policy (2002), sum- marized as follows:173 • Bridges crossing streams within river basins with buffer rules shall not have deck drains that discharge directly into the water body or buffer zones; deck drains may discharge into the buffer zone if 12 feet above natural ground. • Bridges over sounds or water bodies of the Intracoastal Waterway may be allowed to discharge directly into receiv- ing waters because the volume of stormwater runoff from deck drains is small relative to the volume of the water bodies and sites for effective treatment are scarce, unless advised otherwise by the regulatory agencies. As most of these bridges facilitate boat passage, the bridge height and winds help disperse stormwater from the bridges. • For bridges over other waters (perennial or tidal streams), direct discharge into the water body should be avoided to the maximum extent practicable (MEP). In addition, discharge from deck drains in over bank areas similar to stream buffer areas should be avoided. • Where closed systems are utilized to achieve no-direct dis- charge, the discharge point shall be as far away from the surface water body as practical. Preformed scour holes or other devices were recommended to promote diffuse flow. NCDOT has been implementing these policies for new bridges as well as replacement bridges throughout the state since 2002. No-direct discharge is typically achieved through widening of the bridge to accommodate stormwater flow (deck conveyance) or through the use of closed systems. NCDOT has ongoing research on the performance of PFC pavements, bioretention cells, grassed swales, and environ- mental site design. This research will evaluate irreducible concentrations and removal of dissolved metals and other parameters-of-concern identified in this study. 172 NCDOT, 2-10–2-11 173 NCDOT/URS, p. 6-21

A-37 Methods to Identify an Appropriate Whole Life Cost-Benefit Strategy for Bridge Deck Runoff Mitigation As of December 2012 there were 607,380 bridges in the United States as defined within 23 CFR 650 Subpart C.174 Of those, 504,563 are listed as crossing some type of water- way.175 With this number of bridges to maintain and state and federal budgets as they are, funds must be directed to where they will produce a tangible and worthwhile benefit. Management of bridge deck runoff water quality requires practical solutions, which are easy to retrofit to existing infrastructure, maintainable by DOTs using existing per- sonnel, equipment and techniques, and which will have the lowest possible whole-life cost. State DOT Methods MDSHA’s programmatic approach ensures that the agency can cost-effectively respond to the need for treatment by extending treatment in the highest priority areas, usually not at a bridge deck. FDOT ensures consideration of cost by fol- lowing their tier of preferences for (1) runoff drainage off bridge, (2) direct discharge, (3) compensatory treatment, and (4) collection system and piping. WSDOT’s State Stormwater Strategy includes a prioritiza- tion equation to guide their BMP retrofit program. Additional details are available in WSDOT’s permit. This approach may be most accessible in Alaska DOT & PF’s Bridge Deck Run- off study, which contains descriptive summaries of (scoring within) each element of the equation:176 P-Score A B C1 D C2 E1 E2 E3 E4 E5 E6 F.  [ ] ( )( ) ( ) = + + + + + + + + + Where: A = Type and size of receiving water body. B = Beneficial uses of receiving water body. C = Pollutant loading. D = Percentage contribution of highway runoff to watershed. E = Cost/pollution benefit. F = Values trade-off. Alaska DOT started with a stormwater outfall prioritiza- tion system WSDOT developed, which compares the impacts of one outfall with another and makes an assessment of their overall impacts to determine cases in which retrofitting is warranted. The Alaska DOT adds factors from the ACWA, STIP, and several other Alaskan environmental parameters, to indicate bridges where the impacts of bridge deck runoff on the receiving water should be considered. When consider- ing the benefits of constructing a new BMP or modifications to existing BMPs, the weight can be given to the bridges with highest prioritization score, called their “Modified P-score,” which is formulated as follows:177 = + + + + + +M P-score P score P S T V W X Where:178 • P = ADFG score to prioritize some waters over others to pro- tect critical fish bearing resources (High = 5, Medium = 3, Low = 1) • S = Maximum state priority score given by ACWA. Element shows the waters identified by the ACWA as high prior- ity. Waters are nominated and scored by DF&G, DEC, and DNR state agencies, and factored into the calculation by their highest score from one of these agencies. (High = 5, Medium = 3, Low = 1) • T = Traffic type. Heavy truck traffic = 1, No heavy trucks = 0 • V = Salty water. To be aware of the biological environment under the bridge in general, a column described the water underneath the bridge as salty or fresh. It is scored as -1, if it is salty water and scored as 1, if it is fresh water. • W = Silty water. Element identifies whether silty water goes under the bridge. Gathered from the Juneau Department of Transportation as silty/not silty, -1/1. • X = Dimension of the bridge. The bridges were grouped into three sections depending on their length. If the bridge is longer than 400 ft, it is considered long and scored as 5. If the length is between 200 and 400ft, its score is 3, and if it is less than 200 ft, it is a short bridge, and scored as 1. Alaska DOT also advanced the following steps to help engineers to make a decision whether a BMP should be con- sidered for a bridge. • Is it in Urbanized Area? Alaska is considering BMPs for all bridges within UAs. • Is it in Statewide Transportation Improvement Program (STIP)? It is less expensive to construct a retrofit BMP 174 Project interview with Doug Blades, P. E., Structural Engineer, FHWA, Office of Bridge Technology Washington, DC, January 29, 2013 175 Project interview with Doug Blades, P. E., Structural Engineer, FHWA, Office of Bridge Technology Washington, DC, January 29, 2013 176 Alaska DOT & PF’s Bridge Deck Runoff study, p. 54` 177 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, pp. 59–60 178 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, pp. 62–64

A-38 while other construction is underway so if the bridge is in STIP, then BMP options should be considered to handle deck runoff prior to the completion of the project. • What is State ACWA (Combined water body sensitiv- ity) Score? Under ACWA, ADNR, ADFG and ADEC have developed a water body nomination and ranking process. ADNR hydrologists provide factor-ratings for water quan- tity, whereas biologists in ADFG provide aquatic habitat factor ratings, and ADEC provides water quality ratings. Each water body is assigned a high, medium, or lower pri- ority. This provides a general notion of how “sensitive” a water body is. Criteria include the statutory criteria as well as severity of pollution and uses to be made of the waters, per the Clean Water Act § 303(d) (1)(A). Most waters that are listed as impaired are ranked as high priority in ACWA. • Is the bridge over the waters that feed critical habitat (e.g., Cook Inlet)? The National Marine Fisheries Service proposes to designate a critical habitat under the Endangered Species Act for the Cook Inlet Beluga whale. This would result in all discharges to upper Cook Inlet coming under scrutiny. • What is Modified Prioritization Score (PS)? If a bridge in this analysis gets a very high score, BMP should be consid- ered. If it is low, there may not be any need for a BMP. Most of the bridges that have high modified P scores will require BMP consideration based on one of the four proceeding criteria, but a few may not. Here the Alaska DOT will need to set the threshold based on the score. Aside from the threshold, the modified priority score serves as an index of importance of BMP for that bridge and allows relative rankings between bridges. According to the Alaska DOT’s study, the number of bridges requiring treatment are likely to be as follows:179 • In an Urbanized Area: 66 bridges in Anchorage or Fair- banks, or Mat-Su • In the STIP: 61 additional bridges are slated for construc- tion in the next five years • A state priority according to resource agencies: 118 additional bridges were give a priority by ADF&G, ADEC, or ADNR, indicating such in the Alaska Clean Water Actions document • Over waters that feed Cook Inlet: 10 additional bridges in Beluga Whale habitat About 255 of the state’s 703 bridges should be considered for BMPs based on the defined criteria. For the other bridges, the priority score might indicate it should be evaluated for BMP. In that case, however, the cut off score is not defined by regulation. Using the median score, there would be an additional 10 bridges that should be considered for a BMP. For the remainder, the priority score might indicate a relative ranking, but, absent bridge-specific issues, a BMP is not required.180 During the planning of these projects, “the priority score can be used to rate the bridge regarding its likely contribution to receiving water contamination. Thus the priority score can aid decision making regarding the likely benefits of any given BMP; that is, less expen- sive BMPs would be indicated for lower priority scores.”181 If a BMP is indicated at the end of the bridge BMP selec- tion process, a checklist for BMP type is presented as follows. I. Flow into the river via drains or sides a. Can it be changed to flow to ends? i. Unlikely—major engineering/construction project ii. Perhaps if very short? b. Can it be fitted with pipes to ends or treatment? i. Unlikely—major project ii. Little evidence of success in cold regions iii. Further study c. BMP, non-structural i. Public awareness ii. Trash prevention iii. Deicing changes iv. Street sweeping v. Snow management vi. Melting II. Flow to ends a. Non-structural BMP, same as I above b. Structural BMP i. Vegetation ii. Swales iii. Treatment iv. Other Alaska DOT’s project developed a database of all the state’s bridges and their parameters relevant to stormwater runoff. From those parameters a numerical rating was developed for each bridge. This rating, together with certain regula- tory thresholds, is used to determine if BMPs are required. According to Alaska DOT, the best solution for “each bridge is not defined in law, but requires selection by the Alaska DOT after consideration of the bridge characteristics, costs and benefits of candidate BMPs, and practicalities of construc- tion. In general, there are far fewer options for bridge runoff as compared to a standard highway section, and fewer yet that will work in a climate as cold as Alaska. The options can also be quite different for a bridge that is in service versus a bridge that will undergo major repairs or new construction.” Unless the water body is impaired by the bridge runoff—and Alaska 179 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, pp. 2–6 180 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 2 181 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 3

A-39 DOT’s project did not find any bridges where that was the case—there are a wide variety of BMPs that might be applied, ranging from low cost items such as public education and review of de-icing practices, to more costly items such a street sweeping or drainage modifications.182 When asked about their methods to identify an appropri- ate cost/benefit strategy for bridge deck runoff mitigation, NCDOT emphasized that “it comes back to a case-by-case situation: what is practical; what can be installed. It’s a quali- tative assessment. We don’t have a generic process in place to evaluate cost-benefit. We evaluate (that) on every project.”183 NCDOT takes into account the context of the receiving water and anticipated benefit; “in some cases bridge deck discharges are of higher quality than the receiving water,” from their empirical data on mixing in base flow and storm flows. 184 Interviewed DOTs tended to agree that “investing limited resources to clean relatively clean water [from bridge decks] is not ideal,” though all agencies were committed to complying with laws and regulations. Other DOTs indicated they were not doing any sort of cost-benefit assessment. Two DOTs indicated that they were merely treating bridge deck runoff treatment as an “environmental commitment requirement,” where regula- tory agencies made those stipulations (NE, TX). SCDOT indi- cated that they were evaluating costs. The state DOT’s contact at the South Carolina Department of Health and Environment stated that they “worked with SCDOT on what is feasible, prac- tical, and realistic. SCDOT notes that using structures on a bridge can increase the cost tremendously.”185 In North Carolina, Section 25.18 (c) of Session Law 2008- 107 required NCDOT to determine the costs of each treat- ment BMP and the costs of implementing effective treatments on new bridge construction projects as well as existing bridge retrofit projects for all bridges over waterways in the state. This information was provided in NCDOT’s 2010 report, “Quanti- fying capital outlays and annual expenditures associated with various SCMs is vital in supporting informed choices by envi- ronmental stakeholders during the planning process.”186 NCDOT’s 2010 guide utilizes BMP effectiveness as a guide, comprised of (1) site-specific water quality goals, and (2) which BMPs are capable of source control or treatment of storm- water runoff from a particular land use; a BMP “is considered effective if it can be reasonably deduced from available evi- dence that (its) capability for treatment or pollution preven- tion can provide cost-effective and sustainable mitigation for the effect of stormwater to meet receiving stream or water quality” (emphasis added).187 Thus, cost-effectiveness and sustainability/maintainability are central considerations. NCDOT seeks to provide systematic training for designers and engineers associated with selection and implementation of bridge BMPs, considering cost-benefit; the agency’s 2010 report calls for:188 Additional training for designers and engineers (that) should also include optimal selection and implementation of bridge BMPs (and) should both promote understanding of unit pro- cesses for stormwater treatment and encourage value engineer- ing. Measures that promote the most water quality benefit for dollar spent should be emphasized as part of training for design- ers and engineers, including implementation of environmental site design concepts, design aspects that facilitate construction and maintenance, and others, as deemed appropriate . . . It is difficult to introduce costs into the equation due to the variability of key factors and also due “to the limited guid- ance on costing, with several studies only focusing on specific BMPs and in some cases, providing conflicting evidence on unit costs and scale effects.”189 Costs for SCMs have been shown to vary widely due to the influence of climate; site conditions; regulatory requirements, such as environmental and labor issues; aesthetic expectations; public versus private funding; and other influences (Lambe et al., 2005). Many studies have focused on establishing construction costs for specific SCM types, based on analysis of historical con- struction costs of similar projects, or by the development of a bottom-up cost estimate (Wossink and Hunt 2003; Caltrans 2004; Narayana and Pitt 2006). In general, economies of scale have been recognized in observed construction costs for SCMs (Lambe et al. 2005), which could be correlated to a unit size, such as a drainage area (Wossink and Hunt 2003). Cost estimates are generally more reliable when based on local cost information; a common approach is to use engineering estimates to develop an understanding of material and labor requirements and to use local sources for unit cost data (Lambe et al. 2005). It should be noted in the planning process that retrofitting a SCM into an existing site could also involve substantially larger capital outlays than at a new construction site (NRC 2008). Operating and maintenance costs are a substantial portion as well. As the North Carolina interagency team and consultant URS noted: “There have been relatively few studies into these recurring costs, and relatively little cost information is currently available.”190 Nevertheless, when the team examined itemized costs within the budgets for the 10 retrofit projects under con- sideration, they concluded that design costs for BMPs associ- ated with new construction projects were approximately 40% of the design costs for BMP retrofit projects. 182 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, p. 1 183 Eck, Bradley, et al. Water Quality of Drainage from Permeable Friction Course, Journal of Environmental Engineering, ASCE, February 2012, pp. 174 184 Eck, Bradley, et al. Water Quality of Drainage from Permeable Friction Course, Journal of Environmental Engineering, ASCE, February 2012, pp. 174 185 Project Interview with Mark A. Giffin, Project Manager, SC Department of Health and Environmental Conservation, Division of Water Quality, January 7, 2013 186 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation, 2010 187 NCDOT/URS, p. 8-4 188 NCDOT/URS, p. 8-4 189 NRC, 2008, cited in URS/DOT, p. 7-2 190 NRC, 2008; Lambe et al., 2005; Wossink and Hunt, 2003). Cited in URS/ DOT, p. 7-1

A-40 NCDOT developed cost estimates for each of 50 pilot study sites to characterize costs for the particular SCMs at each site and to provide an additional means of identifying costs for bridge SCMs; when actual construction costs or preliminary construction estimates were unavailable, some known data, typically impervious drainage area, were used to estimate construction costs.191 NCDOT described how they calculated operating costs:192 Operating costs represent the costs necessary to inspect, oper- ate, and maintain an SCM. Typical operating cost estimates were derived from the following sources (in order of preference): local data and information, regional or national estimates or models, and best engineering judgment where other data was not avail- able. For each SCM type, operating costs reported include the cost of an annual inspection, the costs of routine maintenance, and the costs of infrequent maintenance. A cost of $100 per annual inspection was assumed based on inspection require- ments . . . and estimates from NCDOT SCM inspection units for the cost of time, materials, and equipment required for inspection and reporting. Routine maintenance costs were based on proce- dures expected to be performed on a regular basis to maintain the proper working order of an SCM, such as vegetation manage- ment, trash and debris removal, and minimal grading and repairs. Infrequent maintenance costs were based on maintenance tasks anticipated to be performed periodically but less frequently than routine maintenance. Examples of infrequent maintenance include accumulated sediment removal; soil media, mulch, and riprap replacement; and larger scale grading and repairs. Where reported, sweeping costs per linear foot of bridge deck ranged from $0.80 to $1.23 per Division swept.193 The NC team considered cost-estimation methods used in other studies as well, including the relationship of construction cost versus water quality volume, which was used in the Caltrans BMP Retrofit Pilot Program (Caltrans 2004), and the relationship of construction cost versus total drainage area, which was typically used in the Water Environment Research Foundation (WERF) models (Lambe et al. 2005) and is discussed below, along with WERF’s 2012 update. NCDOT found the most appropriate relationship to be that of construction cost versus water quality volume because BMPs, particularly NCDOT’s Level I treatment, is typically sized based on the water quality volume, so that relationship was simplified to be that of construction cost versus impervi- ous drainage area (see Figures A-2 and A-3). While retrofit 191 NCDOT/URS, p. 7-10 192 NCDOT/URS, p. 7-7 193 NCDOT/URS, p. 7-8 194 NCDOT/URS, p. 7-14 Figure A-2. Comparison of construction cost to impervious drainage area. (NCDOT, 2010) 194 Figure A-3. Arizona DOT BMP construction and maintenance cost data.

A-41 Table A-4. Arizona DOT annual maintenance BMP costs. construction costs were found to be only 17% higher than comparable new construction, without incorporating higher unit costs for materials on smaller projects, retrofit design costs were found to be two and a half times those of non- retrofit SCM design costs. Enhanced maintenance and inspection costs have not been calculated, but are anticipated to be related to addi- tional training of inspectors and additional effort during the inspection process for recognition and documentation of potential conveyance and collection issues, should such a program move forward.195 Arizona DOT estimated available construction costs for various BMPs and is just beginning to note and compile annual maintenance costs, as shown in Table A-4.196 WSDOT’s extension of AVL-GPS to the remainder of their fleet and integration with the state’s labor/maintenance management tracking system will enable the state to collect actual costs to maintain BMPs, starting in 2013. In the DOT interviews for this project, NCDOT noted that problems with the use of traditional BMPs for bridge deck runoff mitigation extend beyond capital cost and space constraints. The existing mitigation approach is administra- tive, capital, and maintenance intensive. For example, piping 195 Project interview with Matt Lauffer, Hydraulics Unit, North Carolina Depart- ment of Transportation, Jan. 28, 2013 196 Arizona DOT Stormwater Manual, p. 162, Table B.6

A-42 runoff to the abutment for treatment requires a structural BMP as well as an outlet structure to the receiving water, which requires environmental permitting and potentially an engineered energy dissipater. The whole-life cost of the tra- ditional approach is high compared to passive methods on an at-grade highway cross section, such as engineered vegetative filter strips. NCHRP Project 25-40 Information on the Whole Life Costs of BMPs NCHRP Project 25-40, to be completed in 2014, will pro- vide further BMP performance and cost information, build- ing on the initial literature review results discussed below. While design, site, and cost information is relatively sparse in the International BMP Database (BMPDB), there are some studies where this ancillary information is documented. Only 43 studies out of the 133 (32%) contained any con- struction cost information, and 10 studies (7.5%) contained maintenance costs. A summary of costs including averages and ranges of construction and maintenance costs, where available, by BMP type is provided in NCHRP Project 25-40; the median effluent concentrations for 10 selected BMP studies are then compared to the categorical median efflu- ent concentrations presented for some selected constituents; and performance trends based on the time series of avail- able influent/effluent data pairs for the individual studies are then evaluated. The wet retention pond and the media filters contained the highest average construction costs and the manufactured device contained the lowest construction cost. Out of the three BMP types with maintenance cost information avail- able, the manufactured device contained the lowest average maintenance costs per year. As Table A-5 shows, there was a large range of construction costs for each type of BMP and the sizes of projects and drainage areas differed as well as the number of studies available. Table A-6 summarizes the con- struction and maintenance costs according to drainage area and impervious drainage area. WEF (2012) and Lampe et al. (WERF 2005) produce life cycle cost analyses for a variety of BMP types.197 Some of the concluding highlights are noteworthy:198 • Maintenance costs of wet basins make up almost 50% of the whole life cost when basins are implemented in high- visibility locations, where aesthetics are at a premium. Dry basins tend to be easier and less expensive to maintain because there is little or no standing water in the facility. Wet and dry basins cost the same to construct. • The primary maintenance cost of bioretention is associ- ated with vegetation management. The frequency of this activity was assumed similar to swales, but with a greater cost because many bioretention facilities would require weeding, mulch replacement, and other activities beyond the mowing required for most swales. • For swales and filter strips, water quality benefits can effectively be considered as no cost if these areas are already maintained. BMP Type No. of Studies with Construction Costs No. of Studies with Maintenance Cost Average Construction Cost (Range) Average Maintenance Cost/yr (Range) Vegetated Swale 6 0 $101,250 ($60,000 - $140,000) N/A Dry Detention Basin 5 0 $299,566 (77,389 - $819,852) N/A Vegetated Strip 3 0 $213,333 ($110,000 - $300,000) N/A Manufactured Device 17 8 $38,290 ($320 - $180,000) $932/yr ($80 - $3,000) Bioretention 1 1 $150,000 $3,000/yr Media Filter 10 1 $341,505 ($100,000 - $476,106) $3,000/yr Wet Retention Pond 1 0 $691,496 N/A Table A-5. Summary of construction and maintenance costs from BMPDB. 197 Lampe, Barrett, et al. Water Environment Research Foundation (2005) Per- formance and Whole Life Costs of Best Management Practices and Sustainable Urban Drainage Systems; Project 01-CTS-21Ta; Water Environment Research Foundation: Alexandria, Virginia. 225 p., 2005 198 Barrett, WEF, 2012, pp. 502–509

A-43 • Infiltration trenches may require little routine mainte- nance outside of litter and debris removal. The whole life cost driver is the frequency with which the trench must be rehabilitated. Intervals of 4, 8, and 12 years were assumed based on low, medium, and high scenarios, at which time the cost is essentially the same as the original construction cost. For infiltration basins, the capital cost and routine maintenance are essentially the same as those for a dry basin, but an infiltration basin can incur much higher costs associated with maintaining sufficient infil- tration rates. In addition to sediment removal, an infil- tration basin may require additional activities to remove and replace clogged soils on the floor of the basin. The frequency of this activity is largely dependent on the ini- tial soil texture and the rate at which sediment accumu- lates in the basin. • With pervious pavement in the same location as a conven- tional surface, the cost for the water quality control facility is the incremental cost difference between a conventional pavement and pervious pavement. DOT interest has been fostered regarding permeable thin lift overlays through safety and livability co-benefits offered: better visibility and traction in storm events, reduced splash and hydro- planing, and reductions in deflected noise from highway traffic. Now porous asphalt overlays are being used in Georgia, California, and Utah as well. The use of perme- able overlays (PFC) was up to 8.1% of all pavements in Texas in 2010. The overlay is assumed to need replacement more frequently (every 25 years vs. 35 and 40 years) at a cost equal to original construction. Water quality moni- toring of three locations in the Austin area indicates up to a 90% reduction in pollutant discharges from PFC com- pared to conventional pavement. This reduction is the result of accumulation of pollutants within the pavement and the reduction in pollutants washed off vehicles during storm events.199 The NCHRP Project 25-40 interim report points out that with the exception of infiltration trenches, which may clog/ fail and require total reconstruction on a shorter timeframe than many other facilities, the higher level maintenance cost scenario is driven by aesthetics and local expectations for frequency of mowing, rather than functioning of the water quality facility. In initial NCHRP Project 25-40 interviews, Maryland SHA Hydraulics staff reported, “infiltration BMPs are failing more quickly and the reasons are not always clear. Removing the top layer of soil, some infiltration facilities can be restored to initial conditions, but some do not. Facilities may prematurely fail due to generally poor soil characteris- tics, rising groundwater or groundwater mounding.”200 BMP Type Average Construction Cost per Acre of Drainage Area (Range) Average Construction Cost per Acre of Impervious Drainage Area (Range) Average Annual Maintenance Cost per Acre of Drainage Area (Range) Average Annual Maintenance Cost per Acre of Impervious Drainage Area (Range) Vegetated Swale $89,483 ($5,503 - $187,890) $94,671 ($5,789 - $194,849) N/A N/A Dry Detention Basin $60,282 ($27,147 - $97,870) $127,283 ($39,253 - $291,038) N/A N/A Vegetated Strip $164,828 ($95,957 - $222,577) $164,828 ($95,957 - $222,577) N/A N/A Manufactured Device $50,926 ($980 - $428,491) $51,251 ($980 - $428,491) $1,824 ($138 - $7,142) $1,825 ($153 - $7,142) Bioretention $75,879 $94,848 $1,518 $1,897 Media Filter $252,121 ($80,681 - $569,090) $279,754 ($144,714 - $674,477) $4,669 $4,669 Wet Retention Pond $164,611 $341,267 N/A N/A (According to Drainage Area and Impervious Drainage Area) Table A-6. Average construction cost and maintenance cost per year. 199 Bradley J. Eck, Ph.D., P. E., J. Brandon Klenzendorf, Ph.D., Randall J. Charbeneau, Ph.D., P. E., Michael E. Barrett, Ph.D., P. E. Investigation of Storm- water Quality Improvements Utilizing Permeable Friction Course (PFC), September 2010 200 Project interview, Karuna Pujara, MDSHA, December 20, 2012

A-44 DOTs can preserve functioning and extend the life cycle of BMPs if they prevent sedimentation of permanent BMPs during construction on the project or upstream, as “the majority of sediment problems” in permanent controls are caused by inadequate erosion and sedimentation control from construction upstream of the structure. In a stable urban watershed, WEF estimates that normal annual accu- mulation of sediment would be less than 1 cm per year.201 A UK survey identified that no upstream pretreatment was provided in 85% of the stormwater controls where sediment was a problem, a particular issue in the more expensive main- tenance involved in wet basins.202 Heavier solids, leaves, trash, and debris frequently outweigh the load based on total sus- pended solids.203 To facilitate comparison of costs among BMP types, Barrett et al. normalized the whole life cost for each system for high, medium, and low maintenance scenarios for each BMP type, based on the equivalent water quality volume. The team identified a number of important caveats and les- sons. First, water quality benefits from some controls, such as swales and strips, can effectively be considered free when compared to conventional drainage systems, and when the maintenance is performed by the property owner. Further, “a bare-bone, marginal maintenance program (e.g., inspections every 3 years and little vegetation management) does not save that much money compared to a maintenance program at the medium level.”204 Higher-level maintenance costs were often driven by aesthetics more than performance requirements. In line with these conclusions, in initial Project 25-40 inter- views, at least one DOT noted that BMPs located in prox- imity to frequent callers and/or influential people received a higher level of maintenance. Table A-7 shows whole life costs of common BMPs per cubic meter of stormwater treated. Media filters had the highest average construction cost based upon drainage area; however, the wet retention pond (only 1 study) had the highest average construction cost based upon impervious drainage area. The manufactured devices had the lowest average construction cost based upon drainage area and impervious drainage area. In general, man- ufactured devices (which include a wide variety of practices including hydrodynamic devices and cartridge filters) tended to be a cheaper type of BMP to treat highway/roadway, park and ride, or maintenance station stormwater runoff. How- ever, these BMPs also tend to be among the worst performers with respect to pollutant removal. According to the limited maintenance cost information available, bioretention (only 1 study) has the cheapest average maintenance cost per acre of drainage area and media filters (only 1 study) have the highest cost. Manufactured devices (8 studies) have cheaper average maintenance cost per acre of impervious drainage area. Clearly, cost information is extremely limited, so care should be taken when generalizing about BMP construction and maintenance costs. It is important to note that unit construction cost estimates are far from the whole story of DOT costs. With whole life costs, as summarized with Lampe et al. in 2004 and Barrett/ WEF in 2012, maintenance costs are included in the present value analysis. This significantly increases the unit costs for maintenance intensive controls. Further, the manufactured devices from the International BMP Database include a wide range of devices including catch basin inserts, cartridge filters, oil/water separators, hydrodynamic devices, MCTTs, etc.; this results in a huge range of unit costs ($980–$430,000). More detailed cost analysis of the individual manufactured devices 201 Barrett, WEF, 2012, p. 433 202 WERF, 2005, cited in Barrett/WEF 2012, p. 434 203 California Department of Transportation (2004) BMP Retrofit Pilot Pro- gram, Final Report; CTSW-RT-01–050; California Department of Transporta- tion: Sacramento, California 204 Michael Barrett et al., Design of Urban Stormwater Controls Manual of Prac- tice (MOP 23), Water Environment Foundation, June 2012. https://www.e-wef. org/Home/ProductDetails/tabid/192/Default.aspx?ProductId=18172. Stormwater Control Whole Life Cost ($/m3) Low Maintenance Medium Maintenance High Maintenance Swales/Strip 500 660 2200 Wet Ponds/Wetlands 520 600 925 Dry Extended Detention Basins 330 375 575 Sand Filter 450 520 670 Bioretention 1900 2200 5100 Infiltration Trench 1200 1600 2700 Infiltration Basin 330 400 700 Permeable Pavement 570 640 1400 (WEF, 2012) Table A-7. Whole life costs of common BMPs per cubic meter of stormwater treated.

A-45 is necessary to get an improved range for the various device types, but even then the data are limited (only 17 studies with construction cost information; only eight with maintenance cost information). NCHRP Report 474 on Cost-Benefit Strategies for Bridge Deck Runoff NCHRP Report 474 has relatively little on cost-benefit eval- uation strategies for bridge deck runoff. Volume 2 notes that annual maintenance costs for structural BMPs are impor- tant cost considerations; “BMPs that are not maintained can quickly lose any pollutant removal capabilities. Furthermore, a BMP that is not maintained could pose a hazard to the high- way or bridge where lack of maintenance has reduced the BMP’s capacity to handle the volume of runoff planned.”205 They recommend “methods that directly consider operation and maintenance costs over the life of the bridge . . . to ensure that this often critical cost is not overlooked in the analy- sis.”206 As the NCHRP Project 25-13 research team states, the various methods they describe in their report are general and well known; “consequently, bridge engineers and designers are generally already knowledgeable about these methods.”207 Present value analysis (a component of most of the meth- ods discussed later) provides a framework for comparing the direct costs and benefits of project alternatives by accounting for the “time value” of money and opportunity costs (the cost of giving up the opportunity to use or invest the resource). Because net present value combines the effects of costs and benefits, it would not be as useful as benefit/cost analysis in estimating the relative efficiency of various projects. Benefit/cost analysis focuses on the efficiency of project alternatives. It is a basis for comparing and ranking projects with different goals or varying scales. Benefit/cost analysis also includes an estimate of the relationship of all benefits and costs to society by translating indirect costs and benefits into dollars (the sum of all direct and indirect costs borne by or accrued to everyone). If all costs and benefits were direct, net present value and benefit/cost analyses would yield iden- tical results. Using dollars as a common denominator allows conflicting objectives to be compared. Because benefits and costs often accrue in different patterns over time, it is usu- ally necessary to discount them to a present value. The cost parameters associated with the alternatives can be defined to include both initial investment costs and the present value of maintenance costs anticipated over the life of the facilities. Cost-effectiveness analysis is primarily useful when com- paring the costs (and determining the least-cost approach) of different ways of achieving the same measurable goal. This method rests on the assumption that any additional benefits beyond meeting the goal and any nonmonetary costs are insignificant. If those benefits or costs are significant, a tech- nique that focuses on efficiency, such as benefit-cost analy- sis, would be preferred. Cost-effectiveness analysis would, therefore, be most useful in evaluating situations in which a single goal exists rather than multiple goals. One important consideration for all projects is the economic quantification of environmental value. Many stakeholders view economic quantification of environmental resources as controversial. For stormwater BMPs, the common cost-effectiveness metric is the BMP cost per unit mass of pollutant removed (Brown and Schueler 1997). Life Cycle Cost Analysis takes into consideration the total cost of constructing and implementing a facility for its useful life. Historical cost curves, useful life, replacement costs, and operating cost histories for similar facilities are used to aid decision making. In some cases, this type of analysis might identify bridges that should be retrofitted to help establish prioritization of limited funds. Understanding the life cycle stage of retrofit projects competing for highway agency dol- lars makes it possible to consider such factors as these in the resource allocation process: • Projected changes in annual maintenance costs through- out the remainder of the useful life of the equipment or structure. • Opportunities to extend the useful life of the facility through early restoration or rehabilitation. • Risk of significant increases in the cost of implementing the mitigation measures if they are delayed 1 year, 5 years, or some other interval of time. Analysis of life cycle cost can be combined with benefit/ cost analysis or other related methods in developing compo- nents for evaluating mitigation strategies. Production theory optimization, a stormwater BMP eco- nomic optimization method based on production theory and marginal benefits and costs, has been used for a number of combined sewer overflow and stormwater control projects. Production theory optimization analysis is a quantitative method of comparing candidate BMPs to arrive at an optimal solution. It relies on information developed in the technolo- gies evaluation steps, including performance (i.e., pollutant removal effectiveness), cost, and interactions of individual BMPs. It is most useful in cases in which multiple BMPs are considered. CH2M Hill has developed a computerized pro- gram (BEST) that simplifies what otherwise would be a labo- rious evaluation process. This approach may be applicable to larger bridge projects in which the potential costs and benefits warrant this degree of sophistication. 205 NCHRP Report 474, Vol. 2, p. 71-73 206 NCHRP Report 474, Vol. 2, p. 71-73 207 NCHRP Report 474, Vol. 2, p. 71-73

A-46 BMP ranking procedure. The objective of using a BMP is often to protect aquatic life or prevent spills from entering a receiving water. When faced with limited resources to pro- tect receiving waters, it may be worthwhile to evaluate BMPs with regard to pollutant removal efficiency and cost. A series of steps and formulas were developed by Caltrans (Pilgrim 2001), similar to CH2M Hill’s production theory optimiza- tion concept, to rank a list of proposed BMPs by evaluating the ratio of cost to effectiveness for each BMP. The Caltrans method is different than production theory optimization in that a numerical evaluation of BMP removal efficiency is weighted by giving greater value to the removal of pollut- ants that are of particular concern. Once the optimal BMP is identified, it can be compared with mitigating stormwater with similarly evaluated BMPs at other sites in the watershed (e.g., mitigation banking, pollutant trading). Hence, this procedure can also be used to identify when treatment of bridge runoff is not practical—that is, if significantly greater benefits could be realized by treating runoff, for the same or lower costs, from impervious areas that discharge into other locations within the same body of water or watershed. BMPs are ranked by calculating a selection value according to the following formula SV C M E AF ( ) = + + where SV = selection value (lowest value = best BMP option) C = BMP cost M = present worth of maintenance cost (10 years used by Caltrans) E = present worth of environmental monitoring costs (10 years used by Caltrans) A = area of watershed treated by BMP F = pollutant removal factor The pollutant removal factor is a composite value for sev- eral stormwater runoff constituents and spills and is based on the following equation: = + + + +. . .1 2 3 i spillsF f p f p f p f f where f1 - i = weighting factor for each pollutant of interest fspills = weighting factor for spills p = pollutant removal efficiency (% removal/100) There are a number of potential approaches to develop- ing weighting factors. Professional judgment could be used to assign “f” values for each pollutant of interest. For example, if sediment is considered the most problematic pollutant, a large “f” value (e.g., 100) would be assigned to sediment. If aquatic toxicity was the primary concern, large “f” values could be assigned to metals such as copper and zinc (e.g., 100), whereas lower values (e.g., 40) would be assigned to sediment. Clearly, street sweeping would not be a viable option for spill con- tainment, and in this case, the pollutant removal factor (p) would be zero. A more quantitative method would be to use monitoring data and water quality criteria to identify the problematic pollutants. In this case, the “f” factor could be calculated as the frequency, in percent, with which a particu- lar runoff pollutant exceeds water quality criteria. This would link the “f” factor to the protection of the designated use of the receiving water body. The “f” factor for spills may be based on professional judgment and might include consideration for the risk of spills, downstream drinking water sources, and the nature of the receiving water (i.e., how quickly it flushes). Availability of Bridge Deck Runoff Data To assist in these efforts, the research team asked inter- viewed DOTs whether they had bridge runoff datasets that could be shared. Nearly all states said they did not. One said they had no way to collect such information. Another had some data but the DOT had to make it unavailable when they found there were some problems in the data. The DOT is doing QA/QC and this data could be available later. NCDOT indicated that all of the data collected for their report are available in the USGS report and on the USGS web- site.208 The station numbers are in the report and the USGS data can be queried for that. The USGS report contains data on bridge runoff, quality, quantity and stream quality—more comprehensive than NCDOT’s data. The USGS report includes appendices with the data summarized in different ways, includ- ing Excel spreadsheets. The appendices contain water-quality concentrations and loads, bed-sediment concentrations and bridge deck runoff and in-stream discharge data. NCDOT has biosurvey and bioassay reports, bridge sweeping sediment quality data, traffic counts, and additional bioassays and biosur- veys beyond what is in the report, none of which changed the report’s conclusions in the report. USGS noted that researchers can also download the data directly from the USGS National Water Information System web site at the following links:209 • Water- and bed sediment-quality data (http://nwis.water data.usgs.gov/nc/nwis/qwdata) • Discharge and rainfall data (http://nwis.waterdata.usgs. gov/nc/nwis/sw) 208 USGS report entitled “Characterization of Stormwater Runoff from Bridges in North Carolina and the Effects of Bridge Deck Runoff on Receiving Streams” http://pubs.usgs.gov/sir/2011/5180/ 209 Project interview with USGS, Chad Wagner, Hydrologic Modeling and Inves- tigations Section, U.S. Geological Survey, Raleigh, NC, December 20, 2012

A-47 209 NCHRP 25-40 Interim Report (Literature Review Results), June 2012 209 Strecker, E. W., Mayo, L., Quigley, M. M., and J. Howell (2000) Guidance Manual for Monitoring Highway Runoff Water Quality. Federal Highway Administration, Unpublished Draft, Contract DTFH651-94-C-00108 FHWA (August 2001). Urban Drainage Design Manual, Hydraulic Engineering Circular No. 22, Second Edition, prepared by S. A. Brown, S. M. Stein, and J. C. Warner. Federal Highway Administration, Washington, D.C. 209 Strecker, E. W., W. C. Huber, J. P. Heaney, D. Bodine, J. J. Sansalone, M. M. Quigley, D. Pankani, M. Leisenring, and P. Thayumanavan (2005). Critical Assessment of Stormwater Treatment and Control Selection Issues, Water Environment Research Foundation (WERF); Report No. 02-SW-1. ISBN 1-84339-741-2. 290p 209 Venner, M., Compendium of Environmental Stewardship Practices in Con- struction and Maintenance, AASHTO, 2004. NCHRP 25-25/04, maintained online at AASHTO’s Center for Environmental Excellence 209 Urban Drainage and Flood Control District, Urban Storm Drainage Cri- teria Manual Volume 3, August 2011 http://www.udfcd.org/downloads/pdf/ critmanual/Volume%203%20PDFs/USDCM%20Volume%203.pdf, p. 2-13. 209 OSU et al., Evaluation of Best Management Practices for Highway Runoff Control, NCHRP Report 565, 2006 209 McNett et al., 2010 cited in NCDOT/URS, 2010 209 NCDENR, 2009; NRC, 2008. cited in NCDOT/URS, 2010 209 NCDOT/URS p. 8-2 209 Project interview with Matt Lauffer and Kathy Herring, NCDOT, and Michelle Mayfield, Alex Nice (URS Corp), and Chad Wagner, USGS, December 20, 2012 209 NCDOT/URS, 2010, p. 6-2 209 NCDOT/URS. Stormwater Runoff from Bridges, Final Report, July 2010, p. 6-2 209 NCDOT/URS, 6-22 and 6-23 209 NCDOT, 2-10–2-11 209 NCDOT/URS, p. 6-21 209 Project interview with Doug Blades, P. E., Structural Engineer, FHWA, Office of Bridge Technology Washington, DC, January 29, 2013 209 Alaska DOT & PF’s Bridge Deck Runoff study, p. 54 209 Perkins, R., and Hazirbaba, Y., Alaska UTC and DOT & Public Facilities, Bridge Deck Runoff: Water Quality Analysis and BMP Effectiveness, December 2010, pp. 1–6, 59–60, 62–64 209 Eck, Bradley, et al. Water Quality of Drainage from Permeable Friction Course, Journal of Environmental Engineering, ASCE, February 2012, pp. 174 209 Project Interview with Mark A. Giffin, Project Manager, SC Department of Health and Environmental Conservation, Division of Water Quality, Janu- ary 7, 2013 209 NCDOT/URS, Stormwater Runoff from Bridges, Final Report to Joint Leg- islation Transportation Oversight Committee, North Carolina Department of Transportation, 2010 209 NCDOT/URS, p. 8-4 209 NRC, 2008, cited in URS/DOT, p. 7-2 209 NRC, 2008; Lambe et al., 2005; Wossink and Hunt, 2003. Cited in URS/ DOT, p. 7-1 209 NCDOT/URS, p. 7-14 209 Project interview with Matt Lauffer, Hydraulics Unit, North Carolina Depart- ment of Transportation, Jan. 28, 2013 209 Arizona DOT Stormwater Manual, p. 162, Table B.6 209 Lampe, Barrett, et al. Water Environment Research Foundation (2005) Per- formance and Whole Life Costs of Best Management Practices and Sustainable Urban Drainage Systems; Project 01-CTS-21Ta; Water Environment Research Foundation: Alexandria, Virginia. 225 p., 2005 209 Barrett, WEF, 2012, pp. 502–509 209 Bradley J. Eck, Ph.D., P. E., J. Brandon Klenzendorf, Ph.D., Randall J. Char- beneau, Ph.D., P. E., Michael E. Barrett, Ph.D., P. E. Investigation of Storm- water Quality Improvements Utilizing Permeable Friction Course (PFC), September 2010 209 Project interview, Karuna Pujara, MDSHA, December 20, 2012 209 Barrett, WEF, 2012, p. 433 209 WERF, 2005, cited in Barrett/WEF 2012, p. 434 209 California Department of Transportation (2004) BMP Retrofit Pilot Pro- gram, Final Report; CTSW-RT-01–050; California Department of Transporta- tion: Sacramento, California 209 Michael Barrett et al., Design of Urban Stormwater Controls Manual of Prac- tice (MOP 23), Water Environment Foundation, June 2012. https://www.e-wef. org/Home/ProductDetails/tabid/192/Default.aspx?ProductId=18172 209 NCHRP Report 474, Vol. 2, p. 71–73

A-48 North-West Florida Water Management District Compensatory Treatment Guidelines Compensating Stormwater Treatment Occasionally, applicants find that it is impractical to con- struct a stormwater management system to capture the runoff from a portion of the project site due to on-site conditions such as extreme physical limitations, availability of right-of-way, or maintenance access. Two methods have been developed to com- pensate for the lack of treatment for a portion of a project. The first method is to treat the runoff that is captured to a greater extent than required by rule (i.e., “overtreatment”). The sec- ond method is to provide treatment for an off-site area which currently is not being treated (i.e., “off-site compensation”). Either of these methods will only be allowed as a last resort and the applicant is strongly encouraged to schedule a pre- application conference with agency staff to discuss the project if these alternatives are being considered. Other rule criteria, such as peak discharge attenuation, will still have to be met if the applicant utilizes these methods. Each alternative is described in more detail in the following sections. Overtreatment Overtreatment means to treat the runoff from the project area that flows to a treatment system to a higher level than the rule requires to make up for the lack of treatment for a portion of the project area. The average treatment efficiency of the areas treated and the areas not treated must meet the pollutant removal goals of Chapter 62-40, F.A.C., (i.e., 80% removal for discharges to Class III waters and 95% removal for systems that discharge to OFWs.) To meet these goals, the area not being treated generally must be small (less than 10%) in relation to the area that is captured and treated. Staff can aid in determining the proper level of overtreatment for a particular situation. Off-site Compensation Off-site compensation means to provide treatment to compensate for the lack of treatment for portions of the pro- posed project. The following conditions must be met when utilizing off-site compensation: (a) The off-site area must be in the same watershed as the proposed project, and in the closest vicinity practical to the location of those untreated stormwater discharge(s) requiring compensating treatment; and (b) The applicant shall use modeling or other data analy- sis techniques that provide reasonable assurance that the compensating treatment system removes at least the same amount of stormwater pollution loading as was estimated from the untreated project area. Flexibility for State Transportation Projects and Facilities Due to the unique limitations of state linear transporta- tion projects and facilities, subsection 373.413(6), F.S. (2012) requires the agency, during the review of such activities, to consider and balance the expenditure of public funds for stormwater treatment with the benefits to the public in pro- viding the most cost-efficient and effective method of achiev- ing the treatment objectives of stormwater management systems. To do so, alternatives to onsite treatment for water quality will be considered, which may include regional storm- water treatment systems. A T T A C H M E N T A - 1 Sample Off-Site Compensatory Treatment Guidelines

A-49 A T T A C H M E N T A - 2 DOT BMP Selection Matrices Table A-8. MassDOT BMP selection matrix.

A-50 Table A-9. Georgia DOT, general application BMPs.

A-51 Methodologies for Discerning Appropriate Treatment of Bridge Deck Runoff Summarized from NCHRP Report 474 Vol. 2, Practitioner’s Guide: METHOD 1: CALCULATION OF IN-STREAM POL- LUTANT CONCENTRATION AT THE ZONE OF INITIAL DILUTION. This method provides a conservative approach to calculating in-stream concentrations of pollutants within a limited region in which stormwater and receiving water mix. The mixed concentration is calculated at the edge of this mixing region, generally called the zone of initial dilution (ZID). State water quality standards usually provide meth- odologies for the determination of ZID size. Some states do not allow ZIDs and instead compare acute criteria to end- of-pipe concentrations, which in this case would be direct stormwater from the bridge. Acute criteria protect against short-term, lethal effects. Chronic criteria protect against longer-term effects such as growth and reproduction impair- ment. Another option for states that do not use a ZID con- cept for acute criteria is to assume complete mixing with a design stream flow specific to acute criteria. In these cases, the complete-mix approach (see Method 2) should be used. If the estimated undiluted runoff concentration for a given param- eter is less than the applicable in-stream criterion, there is no reason to undertake mass balance calculations. If the back- ground concentration exceeds the criterion, the practitioner should proceed to methods for cases in which sources of pollutants other than the bridge need to be considered (e.g., Methods 11 and 15). METHOD 2: FULLY MIXED IN-STREAM POLLUTANT CONCENTRATION. Method 2 is applicable to analysis of any situation in which it is assumed that the discharge is fully mixed with the receiving water or some specified fraction thereof. This situation could include acute and chronic aquatic life (e.g., acute effects are short-term lethality, and chronic effects are impairment of growth and reproduction), wildlife, and human health toxicity criteria. Method 2 also is appli- cable to other water quality standards (i.e., substances such as salts and color). Calculations of fully mixed in-stream pollut- ant concentrations have been traditionally used to determine whether pollutants from a continuous point source, such as a municipal or industrial discharge exceed chronic or human health water quality criteria. Because stormwater discharges are intermittent, aquatic organisms as well as humans will experience intermittent exposure to pollutants. Hence, cal- culation of in-stream pollutant concentrations from average or peak stormwater flows will overestimate (and thus provide a conservative estimate of) the potential for runoff to have a toxic/human health effect. Although this method generally is very conservative, it will nonetheless often demonstrate min- imal likelihood of toxicity from specific chemicals in runoff from a bridge deck. If this method predicts an exceedance of one or more criteria, it does not necessarily mean that there will be a real impact in the receiving water. Biological test methods (Methods 4 and 5) can also be used to assess toxic- ity and may be preferable. Detailed calculation methodolo- gies for each receiving water type (streams and rivers, coastal areas, lakes, wetlands, and reservoirs) are provided along with considerations for chloride discharge and multiple bridges. METHOD 3: SEDIMENT POLLUTANT ACCUMULA- TION MODEL. One ultimate sink for pollutants is sedi- ment. Once incorporated in sediments, pollutants can either bioaccumulate or cause toxicity to organisms that live in or near the sediment layer. Therefore, comparing sediment cri- teria to sediment pollutant concentrations near the bridge can identify potential long-term impacts of the bridge. This method describes relatively simple models for the calcula- tion of sediment pollutant concentrations downstream from, or near, bridge deck stormwater discharges. A load- ing estimate is required for each of the models described in this method. The models assume that the loading is con- tinuous; therefore, appropriate adjustments are necessary to account for the intermittent nature of bridge discharges. As in Method 2, the models apply to streams and rivers, coastal systems, lakes, wetlands, and reservoirs. Until states and EPA A T T A C H M E N T A - 3

A-52 adopt sediment criteria and implementing procedures are published and widely adopted into state water quality regu- lations, it can reasonably be argued that practitioners should not be expected to evaluate sediment impacts. METHOD 4: BIOASSAY METHOD. This method does not generally apply to new bridge construction unless an existing bridge is used as a surrogate for the new bridge. The potential for adverse effects on receiving waters is often related to storm duration, volume, time between storms (in some cases), traf- fic volume, and mixing with the receiving water. The area of a receiving water that is potentially affected by toxicity is also a function of mixing. If the runoff is funneled to a single point discharge, the amount of mixing in the receiving water will be less efficient than if the runoff were discharged from multiple points across the bridge deck (analogous to a diffuser). The objective of this test method is to determine if bridge deck runoff has the potential to be acutely or chronically toxic to freshwater or marine organisms under simulated runoff con- ditions. To meet this objective, a sampling and toxicity testing program has been developed specific to bridge deck runoff that will assess the toxicity of runoff for time-variable expo- sures. The laboratory bioassay methods described here will provide a scientifically sound and fairly low-cost way to assess the toxicity of bridge runoff to aquatic organisms. Although the methods suggested and organisms to be used are mostly consistent with standard USEPA testing protocols, several deviations are needed to address the time-variable compo- nent of storm events. The methods and materials needed for this test are described in detail. METHOD 5: BIOSURVEY METHOD. This method does not generally apply to new bridge construction unless an exist- ing bridge is used as a surrogate for the new bridge. Two inte- gral factors in assessing potential impacts from bridge deck runoff are the intermittent nature of rain events and the initial concentrations of contaminants. As described in Methods 1 through 3, conservative models of pollutant concentrations rely on an assumption of continuous point source input. The biosurvey method, like Method 4, takes into consideration the intermittent nature of rain events and the initial concen- trations of contaminants in receiving waters. Method 4 is bet- ter suited for assessment of potential impact within an event and for the total event. Method 5 is better for measurement of potential long-term impact. The USEPA and specific state documents should be consulted before conducting a biosur- vey program. Methods are generally organism specific, each having advantages and disadvantages. The biosurvey methods presented rely solely on the use of benthic macroinvertebrates as the indicator organisms of choice. METHOD 6: RECALCULATION OF HUMAN HEALTH AND WILDLIFE CRITERIA WITH SITE-SPECIFIC DATA. The USEPA’s Great Lakes Water Quality Initiative (GLI), promulgated in March 1995, provides a method by which bioaccumulation is directly incorporated into ambient water quality criteria for protection of human health and wildlife (USEPA 1995a). USEPA considered the bioaccumulation concepts and methodologies in the GLI to be reflective of the most current science available for criteria development. In general, the numeric criteria developed by USEPA can be used without site-specific modification. In the event that site-specific modification is warranted, the GLI provides guidance on procedures and data requirements for that pur- pose. Additionally, the GLI describes in detail how human health and wildlife criteria are to be derived for both organic and inorganic substances and provides default values for key parameters, such as food chain multipliers, that are to be used in the absence of substance-specific or site-specific data. The GLI also explicitly identifies 22 substances that are considered to be both persistent and bioaccumulative, referring to them as Bioaccumulating Chemicals of Concern (BCCs). The typical pollutants of highway runoff, including metals such as lead, cadmium, copper, zinc, nickel, and chromium, are not identi- fied BCCs. The USEPA also has developed computer models that can be used by the practitioner for food chain bioaccumu- lation assessment (see Method 9). METHOD 7: FIRST-ORDER DECAY MODELS. The term “decay” normally refers to the loss, reduction, or atten- uation of a non-conservative pollutant in a receiving water by assimilative processes such as bacterial decomposition. The simple first-order decay approach is widely used and described in numerous water quality evaluation texts and relevant USEPA guidance documents. FHWA describes this approach for the highway practitioner. The first-order decay method will often need to be combined with the simple dilu- tion calculations described in Methods 1 and 2. First-order decay processes are included in most of the computerized fate and transport models described in Method 9, but the analyses can also be readily performed with a calculator or spreadsheet. For the case of multiple bridges, the first-order decay model, again usually combined with dilution calcula- tions, can be used to determine if there is a need to consider cumulative impacts (e.g., whether pollutant concentrations reach background levels before the next downstream bridge is reached). METHOD 8: SEDIMENT SAMPLING. This method does not generally apply to new bridge construction unless an existing bridge is used as a surrogate for the new bridge. Two main types of devices are used to collect sediment samples: grab samplers and core samplers. Both devices can be used in toxicity testing and in evaluating chemical and physical properties of the sediment. Core sampling can also be used to evaluate historical sediment records. Location of sites for taking samples will depend on the objectives of the study. However, samples are typically taken from an area of poten- tial contamination and a reference area.

A-53 METHOD 9: FATE AND TRANSPORT MODELS. If a more rigorous analysis of fate and transport of pollutants is warranted (i.e., for long-term pollutant loading effects and sediment accumulation), a more complex water quality mod- eling program can be used to assess the effects of short- or long-term loadings on a receiving water. This type of analysis requires significantly more effort than a basic steady-state model or equation approach. However, the results from this type of analysis can be much more accurate and precise in terms of effects on sediment and water column, as well as in terms of potential effects on water intakes. Several USEPA- supported fate and transport models are available from the Center for Exposure Assessment Modeling (CEAM). A description of the most applicable models is provided. METHOD 10: LAKE MODELS. There are many com- puter-modeling techniques available to predict the effects of stormwater runoff discharges on receiving waters. In the case of lakes, a simplifying complete mix assumption can be used to predict pollutant concentrations. An equation for doing so is provided. That equation and the procedures described for lakes, reservoirs, and wetlands in Methods 1 and 2 focus on conservative substances such as metals and salts. In some cases, bridge runoff effects on eutrophication of these types of water bodies will need to be addressed by practitioners. In these cases, methods outlined by FHWA for highways will be suitable for bridges (Young et al. 1996). However, these water bodies will almost always be subject to nutrient loads from sources other than bridges. Thus, the relative loading analy- ses, pollutant trading, and stormwater banking options are all viable approaches for nutrients (see Methods 11 and 13). METHOD 11: POLLUTANT LOADING. Two methods of calculating pollutant loads from a bridge deck are described under Method 11. These are a simple method and an intensity- correlation method. Both require knowledge of pollutant concentrations in runoff. The NCHRP Project 25-13 literature review revealed only a limited number of studies of bridge deck runoff quality; however, the pollutant concentrations reported may be comparable with stormwater quality data for totally impervious highways—that is, studies in which storm- water was monitored directly from pavement. Therefore, impervious highway runoff quality data likely can be used to supplement bridge deck runoff quality data. Although a comprehensive and edited database of bridge and impervi- ous highway runoff quality does not currently exist, multiple sources of highway runoff quality data do exist. These include reports from FHWA, the US Geological Survey (USGS), and state DOTs; academic publications; and state DOT monitoring studies that were performed for compliance with federal and state NPDES stormwater permit requirements. METHOD 12: COLLECTION OF SITE-SPECIFIC RUNOFF QUALITY DATA. If a more precise, site-specific pollutant concentration and loading is desired, field data can be collected for an existing bridge, or a surrogate bridge with similar attributes as the bridge in question. In 1985, the FHWA published a guidance manual for highway runoff and receiving water monitoring (Dupuis et al., 1985a). In general, the methods described remain valid and applicable today. In addition to their previous studies, the FHWA has recently sponsored development of an updated monitoring guidance document for highway runoff. METHOD 14: IN SITU TOXICITY TESTING. This does not generally apply to new bridge construction unless an existing bridge is used as a surrogate for the new bridge. In situ toxicity studies use a unique method in which organisms that occur as natural populations within the system under study are used as test organisms. In these studies, the end- point is usually some measure of survival (percentage alive compared with a reference/control group). METHOD 15: COMPARISON OF BRIDGE DECK LOADING TO OTHER SOURCE LOADINGS IN WATER- SHED. Comparison of the pollutant loading from a bridge deck with other sources in the watershed can provide an idea of the relative impact from the bridge. Additionally, infor- mation needed for pollutant trading, off-site mitigation, and stormwater banking programs can be obtained through such an analysis. Loadings from the bridge deck can be determined by use of Methods 11 and 12, whereas loadings from other sources can be obtained in a variety of ways. The preferred approach is to obtain these estimates from an agency or entity that has already developed them for other reasons (e.g., a local TMDL program). It will generally be useful to be able to compare the antici- pated or predicted pollutant loads from the bridge deck with those from other sources in the watershed. This not only places the impact of the bridge in a relative context, but it also can provide the information needed for pollutant trading, off-site mitigation, and stormwater banking programs. The loading from the bridge can be estimated using Methods 11 and 12. Obtaining estimates of pollutant loadings from other sources in the watershed can be done in a variety of ways. The preferred approach is to obtain these estimates from an agency or entity that has already developed them for other reasons; however, when these data are inadequate, the responsible agency may consider implementing a water qual- ity monitoring program. With the recent increase in water- shed based programs, including TMDLs (see NCHRP Research Results Digest 235 [Dupuis et al. 1999]), there will be a rapidly expanding database on sources and loads for receiving waters in the United States. Some data will be specific to a particular watershed; other data will be statewide or regional and cover a variety of land uses. These will become increasingly acces- sible to the practitioner, as evidenced by USEPA’s Surf Your Watershed Internet access database (http://www.epa.gov/surf). Other sources of applicable water quality information may

A-54 include USGS, watershed councils, or university extension offices. Such watershed-specific information will be superior to nonspecific literature values for particular land use types that have often been used in the past (Dupuis et al., 1985b). This is particularly true for agricultural sources, which vary widely because of differences in climate and agricultural practices. Other methods available to the practitioner range from the very simple (e.g., export coefficients) to sophisticated watershed models (Lahlou et al. 1996; USEPA, 1992a). For most bridge projects, simpler methods should suffice in cases in which load- ings data are not already available from other agencies. If a more in-depth modeling approach is indeed appro- priate, a recommended starting point would be USEPA’s BASINS modeling framework (http://www.epa.gov/ostwater/ BASINS/), which is based on a geographic information sys- tem (GIS). According to USEPA’s BASINS web page (http:// www.epa.gov/ostwater), BASINS, originally released in 1996, addresses three objectives: (1) to facilitate the examination of environmental information; (2) to provide an integrated watershed and modeling framework; and (3) to support analysis of point and nonpoint source pollution management alternatives. It supports the development of TMDLs, which require a watershed-based approach that integrates both point and nonpoint sources. BASINS can support the analysis of a variety of pollutants at multiple scales, using tools that range from simple to sophisticated. The heart of BASINS is its suite of interrelated components essential for performing watershed and water quality analy- sis. These components are grouped into five categories: 1. National databases; 2. Assessment tools (target, assess, and data mining) for eval- uating water quality and point source loadings at a variety of scales; 3. Utilities including local data import, land use and dem reclassification, watershed delineation, and management of water quality observation data; 4. Watershed and water quality models including pload, npsm (hspf), swat, toxiroute, and qual2e; and 5. Post processing output tools for interpreting model results. Basins’ databases and assessment tools are directly integrated within an ArcView GIS environment. The sim- ulation models run in a Windows environment, using data input files generated in ArcView. METHOD 16: ASSESSMENT OF HAZARDOUS MATERIAL SPILLS. Spills on bridges obviously have the potential to adversely affect aquatic life in the receiving water. Given that most highway spills are of limited volume and duration, the primary concern is acute (i.e., mortality) effects. Oregon has developed documentation of a hazardous material spill risk assessment (Kuehn and Fletcher 1995) that applies to drinking water supplies. The Oregon document and other studies were used to develop a hazardous material spill risk assessment methodology that consists of three parts. The assessment methodology can be found in the full description of Method 16 provided in the Appendix to this volume and applies to any numeric water quality criterion, whether it be drinking water or acute aquatic life. Another topic relevant to the mitigation of hazardous material spills is “restoration- based compensation,” in which the timing of a restoration project in relation to a hazardous material spill is important. For instance, if a restoration project is performed after a spill, the “time value” of the spill must be considered in determining the extent of the project. By the same token, if restoration is prior to the spill, a certain amount of credit becomes available the longer the time is between restoration and the spill event. METHOD 17: MICROCOMPUTER SPILL MODELING. In rare situations, a bridge project may warrant a more sophis- ticated assessment of the effects of a spill. In these cases, the practitioner (or consultant) can use a software package such as the Spills Analysis Workstation (SAW) developed by the Danish Hydraulics Institute in Denmark or other specialized programs. METHOD 18: RETROFIT PRIORITIZATION METH- ODOLOGY. This method does not generally apply to new bridge construction unless an existing bridge is used as a surrogate for the new bridge. Retrofitting bridges with struc- tural stormwater BMPs is technically difficult and can be very costly. Therefore, it is likely that this method would be used on only a limited number of existing bridges. A prioritization method can be used to identify the bridges where bridge deck runoff is substantially affecting the receiving water and where the greatest benefit could be gained by retrofitting. Retrofit- ting can include the construction of new structural BMPs or modifications to existing BMPs. WSDOT developed a storm- water outfall prioritization system, which uses a rating system to compare the impacts of one outfall to another and makes an assessment of their overall impacts to determine when retro- fitting is warranted (WSDOT 1996). WSDOT’s outfall pri- oritization methodology has been modified in Method 18 to address prioritization of bridge deck runoff discharges only. METHOD 19: ANTIDEGRADATION ANALYSES. All states are required by the Clean Water Act to have an antideg- radation policy in their water quality standards. The policy is especially intended to protect high-quality waters from new or increased sources of pollution. Additionally, state waters are not allowed to degrade from their existing condition without appropriate analysis, justification, and public input. Although no standardized national protocols exist for antidegradation analyses, many states have specific procedures and methods that must be followed for new or increased discharge of pol- lutants. The practitioner is thus advised to investigate these restrictions very early in the bridge-planning process.

A-55 A T T A C H M E N T A - 4 BMP Inspection and Reporting Tables (WERF 2005) Table A-10. Inspection, reporting, and information management for swales and strips (WERF 2005). Vegetation Management

A-56 Table A-11. Vegetation management with trash and minor debris removal for swales and strips (WERF 2005).

A-57 Table A-12. Intermittent maintenance for swales and strips (WERF 2005).

A-58 Table A-13. Summary method to estimate effort for inspection, reporting, and information management for wet ponds (WERF 2005). Wet Ponds

A-59 Table A-14. Summary of vegetation management and trash and minor debris removal for wet ponds (WERF 2005).

A-60 Table A-15. Vegetation management with trash and minor debris removal for dry extended detention ponds (WERF 2005). Dry Extended Detention Ponds

A-61 Table A-16. Summary method to estimate costs for inspection, reporting, and information management for media filters (WERF 2005).

Table A-17. Filter maintenance for media filters (WERF 2005). Filter Maintenance

A-63 Table A-18. Trash and minor sediment and debris removal for infiltration trenches (WERF 2005). Infiltration Trenches

Table A-19. Sediment removal for infiltration trenches (WERF 2005). Table A-20. Street sweeping and trash and minor debris removal practices for pervious pavement (WERF 2005). Street Sweeping

A-65 Table A-21. Intermittent facility maintenance: structural repairs for pervious pavement (WERF 2005). Pervious Pavement

A-66 Catch Basin Cleaning Many municipalities, especially those with combined sewer systems, have catch basins that maintain a permanent pool of water. These inlets retain sediment and floatables, which must be periodically removed. As material accumulates in the catch basin, pollutant retention decreases. According to Aronson et al. (1983), catch basins should be cleaned. It is also a good idea to inspect and clean all catch basins that serve as a tribu- tary to a wet basin or wetland when that facility is cleaned to reduce sediment loading to the fore bay. One study of catch basins in Alameda County, California, found that increasing the maintenance frequency from once per year to twice per year could increase the total sediment removed by catch basins on an annual basis (Mineart and Singh 1994). The study found that annual sediment removed per inlet was 25 kg (54 lb) for annual cleaning, 32 kg (70 lb) for semiannual and quarterly cleaning, and 73 kg (160 lb) for monthly cleaning. Although catch basins are relatively inexpensive to install, the real cost is associated with long-term maintenance cost. An educator truck (or Vactor truck), the most common method of catch basin cleaning, can cost up to $250,000 (U.S. dollars). Typical trucks can store between 10 and 15 m3 (10 and 15 cu yd) of material, which is enough storage for three to five catch basins (WERF 2005). Typically, using a crew of two, the average catch basin takes 30 minutes to clean. Severely polluted catch basins, which typically result from illegal dumping, may take several days of repeated cleaning. (WEF 2012, p. 478)

Next: Appendix B - Simple and Complex Assessment Methods and Worked Example »
Bridge Stormwater Runoff Analysis and Treatment Options Get This Book
×
 Bridge Stormwater Runoff Analysis and Treatment Options
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 778: Bridge Stormwater Runoff Analysis and Treatment Options presents information and an analysis process for identifying cost-effective, pollution-reducing strategies for management of stormwater runoff from highway bridges.

Six spreadsheet analysis tools accompany the report:

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!