Highway Hydraulic Engineering State of Practice (2020)

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.

5 Introduction The literature review summarizes the results from searching state DOT manuals, additional policies, procedures, and technical memoranda for topics as defined in the scope. The informa- tion collected from the states was informed by the results from the survey—the responses from the states provided a guide to policy and practice that was researched. The states highlighted overall in the literature review are shown in Figure 1, and for each topic, the states highlighted within the literature review are shown in Table 1. This literature review contains information based on available documents and assumes that an updated manual corresponds with an update in the practice. There may be a lag between the state of practice and when the state of practice is documented in a policy document or manual; for example, DOTs may be implementing a state of practice that is not yet documented in a policy/manual. Additionally, practice may have advanced but the DOT may not have updated its manual. Roadway Drainage Proper roadway drainage is imperative for safety and traffic flow during storm events. Road- way drainage is typically achieved through the installation of inlets and culverts, but to operate effectively, they must be clear of debris. Spread criteria (i.e., allowable depth and width of flow on the road surface) are essential to conveying peak discharges with minimal disruption to traffic in locations where inlets catch excess runoff (Urban Drainage Flood Control District 2016). Posi- tioning and malfunctioning of inlets through clogging can lead to reduced inlet efficiency (Leitao et al. 2015). This literature review is limited to the subject topics that include spread criteria—both temporary and permanent—for roadway drainage. NCHRP Project 15-55, “Guidance to Predict and Mitigate Dynamic Hydroplaning on Roadways” provides information on water depth on pavement in regard to hydroplaning. The state policy and practices with respect to spread criteria are included in the following paragraphs, along with inlet design and reduced inlet efficiency. Some highlights are that Minnesota considers shoulder-to-shoulder width in some cross-section designs, Colorado con- siders hydroplaning potential in designs, and Ohio designs for temporary and permanent drain- age conditions. Also, to account for reduced inlet capacity, North Carolina assumes 50% debris blockage for inlet capacity and 50% blockage for inlet spacing design depending on the condi- tion, Virginia designs for off-site drainage interception before it reaches the roadway, and Utah requires a 6-inch upsizing of pipe culverts with fill heights of 8 ft and greater. This review is not comprehensive, but it is intended to be representative of practices and policies at the state level. C H A P T E R 2 Literature Review

6 Highway Hydraulic Engineering State of Practice Colorado Colorado DOT considers design storm frequency and roadway type in determining spread criteria. In Colorado, highway classification (along with highway speed) are major consider- ations as public expectations are that a certain amount of water will be on a given road surface. Design speed is selected after evaluating hydroplaning potential, but some leeway is provided as a likelihood exists that this speed will be exceeded (Colorado DOT 2013). Colorado uses design storms to determine spread criteria, using 2-year through 10-year design storm frequency and 50-year storms for a sag point on arterials (Colorado DOT 2013). Indiana Indiana DOT specifies a clogging factor of 50% is to be used for inlet spacing design, but this is not always required. Exceptions are made for special cases, a full list of which can be found in Chapter 203 of their drainage manual. Slotted drain inlets are no longer considered for use in sags as there is a tendency to collect debris (Indiana DOT 2013). Minnesota Minnesota DOT bases spread criteria on average daily traffic (ADT), lane width, and speci- fied design storms. Lane width may either be the parking lane width, shoulder width, or driving lane width (Minnesota DOT 2000). A memorandum introduced in 2016 expands spread criteria to consider shoulder width with an additional factor of safety for varying roadway types. The document further details design frequency and allowable spread for specific roadway types such as Trunk Highway Turn Lanes and Roundabouts. Design storm events currently include the 5-year event and above (Minnesota DOT 2016), representing an increase from the previously specified 3-year event (Minnesota DOT 2000). Created with mapchart.net © Figure 1. States highlighted in the literature review for NCHRP Project 20-05/ Topic 50-02, “Highway Hydraulic Engineering State of Practice.”

Literature Review 7 State Roadway Drainage Culverts Aquatic Organism Passage Bridges Scour Computations and Countermeasures Advanced Hydraulic Modeling Regulatory Requirements Floodplain Impacts and Mitigation Coastal Hydraulics Alternative Project Delivery Methods Alaska x x California x x Colorado x x x x Delaware x x Florida x Georgia x x x Hawaii x Illinois x Indiana x Louisiana x Maine x x Maryland x x x x Michigan x x Minnesota x x x x x Nebraska x New Hampshire x New Mexico x New York x x x North Carolina x x x Ohio x x x Oregon x x x Pennsylvania x x x Rhode Island x South Dakota x Tennessee x x Texas x x x x Utah x x x x Vermont x x Virginia x x x x Washington x x x West Virginia x Wyoming x Table 1. States highlighted within the literature review.

8 Highway Hydraulic Engineering State of Practice North Carolina In North Carolina, permanent spread criteria for major arterials depends on the speed limit: speeds less than 45 mph and sag points use the shoulder width plus 3 ft, while speeds greater than 45 mph use shoulder width only, with both using a 10-year design storm. For minor arterials and speeds less than 45 mph with sag points, spread is calculated using half of a travel lane width; spread for those arterials greater than 45 mph are calculated using shoulder width only (North Carolina DOT 2016). An assumption of 50% debris and potential blockage is required when computing inlet capacity for sags. During construction, the allowable spread is based on traffic volume, road classification, speed limit, and lane width. The drainage manual does not provide specific values for temporary spread conditions (North Carolina DOT 2016). Ohio Ohio DOT incorporates different spread criteria for permanent drainage and temporary drainage for construction. For temporary conditions, a dry lane width of 10 ft is necessary for each traveled lane, determined using a 2-year design storm. Any drainage system designed incorporates both permanent and temporary solutions, with permanent drainage preferred for the project. Temporary systems may include inlets, storm sewers, culverts, ditches, catch basins, conduits, French drains, and pavement saw cut openings, among others. These tem- porary systems may be removed if not included in the final design. Ohio DOT specifies that final designed sizes should be used where feasible and temporary storm sewers must have a minimum diameter of 12 inches, while temporary culverts must have a minimum diameter of 18 inches (Ohio DOT 2018). South Dakota South Dakota DOT limits the use of standard grate inlets to minor sag points as they are susceptible to clogging, with oversized grate inlets recommended for major sag points and in areas prone to clogging. Alternatively, combination curb-opening and grate inlets may be placed in sags to accommodate reduced inlet efficiency in the event of excessive debris. The most com- monly used inlet by South Dakota DOT is a combination inlet with a grate featuring curved vanes sloped in the direction of flow to increase hydraulic capacity. Drop inlets are typically designed with sumps to collect debris (South Dakota DOT 2011). Utah Utah DOT references HEC-9 (FHWA 2005) for sediment and debris control for pipe culverts to account for reduced inlet efficiency. Pipe culverts with fill heights 8 ft or greater are to be upsized by 6 inches. Additionally, minimum pipe diameters that can be specified are 24 inches for off-site flows and 18 inches for on-site flows (Utah DOT 2018b). Virginia Virginia DOT prefers curb-opening inlets to grate inlets as more debris can be handled. Any runoff from land off site is intercepted from cut slopes and other areas draining toward the road- way before it reaches the curb and gutter section. This minimizes sediment and debris build-up on the roadway, as well as reduces the amount of water flowing into an inlet. To account for clogging, an efficiency of 50% is adopted when grates are used in a sag (Virginia DOT 2019a). Culvert Aquatic Organism Passage Information reported here pertains to culvert replacements, as little guidance for culvert rehabilitation for aquatic organism passage (AOP) was available; rehabilitation may simply consist of retrofitting the existing culvert. Advances in culvert AOP consist of estimating local

Literature Review 9 and average velocities using flume and computational fluid dynamics models for low flows in large culverts. Culvert assessment and asset management systems have been developed to account for AOP. Stream simulation methods for low-impact designs, such as embedded or open bottom culverts and countersinking, have also been used for AOP. Alaska, Georgia, Maryland, Michigan, Minnesota, Vermont, and Wisconsin DOTs in Alaska, Georgia, Maryland, Michigan, Minnesota, Vermont, and Wisconsin partici- pated in a FHWA pooled fund study to fill the need for national attention placed on designs for successful AOP. To accurately predict fish movements (and impediments), the study developed a method to reliably estimate local and average velocities for low flows in large culverts, as well as velocity variations in embedded and non-embedded culverts, to replace unreliable velocity esti- mates that have typically been determined from bankfull conditions or even higher flows. Culvert materials, shapes, lengths, slopes, bed material, and entrance and outlet conditions were examined. Full-scale flume experiments and computational fluid dynamics (CFD) simulations were per- formed, with CFD modeling calibrated to physical results. The proposed procedure will allow the design engineer to reliably estimate average and local velocities to produce a velocity distribution over the cross section modeled. The method was developed for circular culverts and could be expanded to other shapes (FHWA 2014). The study concluded in 2014 with the goal of incorpo- rating the results in future versions of HEC-26 and other guidance documents (FHWA 2016b). Maryland Maryland DOT State Highway Administration (SHA) has indicated that currently AOP design is case by case as agreed upon with environmental agencies, and future designs will be based on policy guidelines being developed by the Maryland DOT SHA/MDE (Maryland Department of the Environment) Hydraulics Panel. The Maryland DOT SHA/MDE Hydraulics Panel has delegated AOP to a subcommittee to identify regulations, policies, and procedures that should be updated and developed for integration of effective AOP practices (Maryland Hydraulics Panel 2018). Minnesota Minnesota DOT has a new AOP guide, Minnesota Guide for Stream Connectivity and Aquatic Organism Passage Through Culverts (2019), that includes some changes that Minnesota DOT has made over the last 10 years to address fish passage. This guide was created from existing literature, design documents, expert input, and a survey to identify best practices that will allow Minnesota DOT to: 1. Design the culvert to be similar to the stream channel (reference reach) by matching its slope, alignment, bankfull width, and flow depth to maximize AOP; 2. Provide a continuous sediment bed with roughness similar to the channel, while maintaining continuity of sediment transport and debris passage; 3. Design for public safety, longevity, and resilience; 4. Improve AOP; 5. Account for sediment transport; 6. Reduce long-term maintenance costs; and 7. Increase culvert life span. Design procedures incorporated in this document include both geomorphic simulation (US Forest Service Stream Simulation) and hydraulic simulation (FHWA HEC-26) (Minnesota DOT 2019). The report considers not only design but also site assessment parameters, and hydraulics including sediment transport, floodplains, energy dissipation structures, water level controls, and cost considerations (Minnesota DOT 2019).

10 Highway Hydraulic Engineering State of Practice New Hampshire New Hampshire DOT follows New Hampshire Department of Environmental Services (NHDES) rules for structure permitting and the New Hampshire Stream Crossing Initiative Field Manual that was created from a partnership of state agencies. Through the expertise of the partnership, stream crossing survey protocols and data management measures were refined for a multi-agency effort to score stream crossings characterized by size, safety, and flooding. The data is stored in the Statewide Asset Data Exchange System (SADES) online ArcGIS geodatabase. The culvert assessment initiative is aimed at collecting coarse level data during a stream survey to rank culverts in terms of: 1. Geomorphic Compatibility (GC; structure fit with river form and processes); 2. AOP (ranking of whether the structure is a barrier to animal passage); 3. Condition (New Hampshire DOT asset condition score); and 4. Hydraulic Capacity of the structure to transport predicted flows under storm events (NHDES et al. 2019). New Hampshire DOT creates sensitive designs for aquatic organism passage on a case-by-case basis, as evidenced by several projects and coordination efforts between the New Hampshire Bureau of the Environment and the Natural Resource Agency (New Hampshire DOT Bureau of the Environment 2015). Oregon Oregon DOT uses hydraulic methods and/or stream simulation. The method defined in Stream Simulation: An Ecological Approach to Providing Passage for Aquatic Organisms at Road-Stream Crossings by the U.S. Department of Agriculture (2008) intends for unrestricted movements of aquatic organisms by designing crossings that mimic the stream. Additionally, FishXing software and Hydrologic Engineering Center’s River Analysis System (HEC-RAS) aid in low-impact design as they can simulate flow in embedded or open bottom culverts. A stream assessment is conducted for stream simulation, and the goal is to provide a crossing that minimally inhibits the natural channel flow. There is no set rule for stream simulation, and the designer should be aware that this is a dynamic approach as stream simulation guidelines are always changing. Considerations for design include natural alignments, careful design of transi- tions, size and mobility of bed material, and pool depth and location. The Stream Simulation: An Ecological Approach to Providing Passage for Aquatic Organisms at Road-Stream Crossings guide provides many examples and scenarios for consideration (U.S. Department of Agriculture 2008), including the following alignment examples: “(a) Matching culvert alignment to stream alignment. (b) Realigning the stream to minimize culvert length. (c) Widening and/or shortening the culvert.” Pennsylvania The Pennsylvania highway design manual includes policy for risk assessment and coordina- tion with the PA Fish and Boat Commission regarding “potential for changes to the ecology or the aquatic habitat of the stream channel and floodplains.” Specifications for “channel con- struction involving fishable streams” include considerations for erosion protection, natural channel relocation, bench widening, turbidity, and open bottom structures. It is stated that the specifications listed in the highway manual are not all-inclusive and that the Engineering Dis- trict Office will make recommendations for the hydrologic and hydraulic report (Pennsylvania DOT 2015).

Literature Review 11 Virginia The Virginia DOT Drainage Manual requires countersinking of culverts, whereby the bottom of the culvert is placed below the bed of the stream, as prescribed necessary in a streambed under jurisdiction of the U.S. Army Corps of Engineers (USACE). The USACE countersinking method must be followed and is outlined in Virginia DOT Drainage Manual Chapter 8 (Virginia DOT 2019a). Design for fish habit requires passage accommodations, and VDOT design criteria are speci- fied in the following: • An Analysis of the Impediments to Spawning Migrations of Anadromous Fish in Virginia Culverts (Pages 61 through 66) August 1985, by Mudre, Ney and Neves; and • Nonanadromous Fish Passage in Highway Culverts Report No. VTRC 96-R6 October 1995 by Fitch (Virginia DOT 2019a). Virginia DOT will evaluate any project where circumstances do not allow for countersinking on a case-by-case basis (Virginia DOT 2019a). Washington The Washington State DOT Hydraulics Manual specifies guidance for fish passage in addi- tion to the Washington Department of Fish and Wildlife Water Crossing Design Guide- lines Manual (WDFW 2013) and the Fish Passage Inventory, Assessment, and Prioritization Manual (WDFW 2019). The design engineer should inspect for aquatic habitats with particu- lar attention to large woody material (LWM; also known as large woody debris). LWM has a significant effect on aquatic habitat in Washington, and the Washington State DOT hydraulics manual outlines detailed design and placement criteria for its use (Washington State DOT 2017b). Figure 2 shows the U.S. Department of Agriculture (USDA) Natural Resource Con- servation Service (NRCS) Thunder Road fish barrier project in La Push, Washington. This project restored floodplain connectivity and allowed fish to pass four barriers. Wyoming Wyoming DOT uses “FishXing” software for assessment and design of culverts for AOP. The models constructed by this software provide several aids in the design of parameters that allow for unobstructed fish passage, such as comparing culvert hydraulics with organism capa- bilities, performing water surface profile calculations for different culvert shapes, and calculating leap condition velocities (FishXing 2006). Bridge Scour Computations and Countermeasures Information from several states is summarized pertaining to bridge scour. Design scour depths could be reduced through new methods based upon the unconfined compressive strengths of the natural soils (Illinois DOT), advanced calculations on bridges in tidal waterways (Maryland SHA), rigorous hydrologic and hydraulic modeling (New Mexico DOT), detailed guidance for various bridge/culvert types (Texas DOT), and use of rock quality designation (Virginia DOT). New countermeasures used are standalone riprap installations, matrix riprap, geobags or geotextile containers (Minnesota DOT), elimination of the filter fabric wrap within the toe for easy maintenance and repair (Illinois DOT), inspection and asset management for critical structure determination (Maine DOT), and 3D sonar infrastructure mapping for rapid assessment (Minnesota DOT).

12 Highway Hydraulic Engineering State of Practice Illinois Illinois DOT conducted a research study on contraction and pier scour prediction in cohesive soils: Pier and Contraction Prediction in Cohesive Soils at Selected Bridge in Illinois (Illinois Center for Transportation, et al. 2010). One of the results was that calculated scour depths could sometimes be reduced based upon the unconfined compressive strengths of the natural soils. This information was used to produce adjustments to scour calculations now included in the Illinois DOT Bridge Manual (Illinois DOT 2012a). The reductions are only considered after a geotechnical analysis has been conducted and subsequently recommended on a case-by-case basis (Illinois DOT 2012a). Obvious benefits are the cost savings realized by the reduction in substructure costs (VanBebber 2019). Illinois DOT recently modified a revision to a scour countermeasure, standalone riprap instal- lation. These installations are used to protect slope walls that secure abutments, bridge piers, channel, or streambed revetment, etc. Recent research has eliminated the filter fabric wrap within the toe and now terminates at the lowest toe point. When failure of the riprap pro- tection occurs, the riprap can be easily repaired without the filter wrap (Illinois DOT 2012b). The revision is based upon the filter fabric guidance provided by the FHWA (Illinois DOT 2012b, VanBebber 2019). Maine Changes in precipitation and streamflow instituted the need for adaption policies that included scour-related goals of bridge inspections every 2 years and underwater scour Figure 2. Thunder Road fish barrier project (reprinted from U.S. Department of Agriculture [USDA] Natural Resource Conservation Service [NRCS] West Area Biologist Rachel Maggi, Thunder Road fish barrier project La Push, Washington, 2018).

Literature Review 13 inspections. Maine DOT evaluated over 350 bridges during 2007-08, and the scour-critical bridges were flagged as requiring flood monitoring or countermeasures (Maine DOT 2014). During the program, information such as water surface elevations, scour depths, and debris were collected. Maine DOT then completed several scour countermeasure projects in the 2013–2014 time frame. The projects in general were very challenging from an environmental and construct- ability perspective. During that time frame, Maine DOT was trying to rapidly reduce the number of scour-critical bridges in the inventory by funding many projects. The projects were on many single span bridges with concrete abutments which had exposed footings. Less scour-critical bridges now exist so there are currently fewer such projects (Myers 2019). Originally, some projects used precast concrete block mat applications, but they did not perform well so changes include designing a specific layout and final grade of the precast con- crete block mats. Additionally, larger blocks have been used for recent applications than had previously been used (Myers 2019). Pier scour countermeasures are usually riprap installation, and this approach continues to perform well. No changes to scour design practice have been made for new bridges (Myers 2019). Maryland Maryland SHA has a scour program for existing bridges, as outlined in Chapter 7 of the Manual for Hydrologic and Hydraulic Design, and also detailed instructions on scour calcula- tions, as outlined in Chapter 11 of the manual (Maryland SHA 2011, 2016). A key change in scour analysis was incorporation of the use of the two in-house programs: ABSCOUR 10 for scour evaluation and TIDEROUT (Kosicki 2019). The ABSCOUR 10 Pier Module incorporates HEC-18, but does not account for bed load material composition. Additionally, HEC-18 coarse bed pier scour is incorporated, but without spread footings and pile caps. Pier scour and debris, the pressure method, MDSHA abutment and contraction scour, bottomless culverts, erodible rock, design flood selection, and bridge foundation design criteria are also included (Maryland SHA 2016). Manual for Hydrologic and Hydraulic Design, Chapter 11, Appendix B: TIDEROUT2 specifies use of this software to perform scour calculations on bridges in tidal waterways. TIDEROUT2 can model scour in tidal basins up through upland watersheds and is particularly useful in scour estimation for older bridges that constrict waterways. TIDEROUT2 is user-friendly software that can model unsteady flow, potential storage benefits, and worst-case scour scenarios involving combinations of tidal flow and riverine hydrographs. Some limitations are that littoral drive or sediment transport and complex currents cannot be modeled. Also, local abutment and contrac- tion scour must be separately computed (Maryland SHA 2015). There is a learning curve for new users, but the method allows for quick and easy sensitivity studies. The inputs to the software should be based on full-scale hydraulic models, and coeffi- cients used should be input mindfully. Maryland DOT follows the FHWA guidance and does not account for any riprap protection while computing scour depths (Kosicki 2019). Minnesota Minnesota DOT is incorporating a new scour countermeasure called matrix riprap instal- lations for use at spill-through bridge abutments. Matrix riprap uses Class 5 riprap with grout placed in the riprap to glue the individual stones together, but not entirely filling the void space (Minnesota DOT 2015a). This option allows for smaller size riprap stones to be used as the grout provides increased shear resistance (Minnesota DOT 2015b). Minnesota DOT con- ducted laboratory experiments in a steep slope flume and also field experiments, with results

14 Highway Hydraulic Engineering State of Practice showing that matrix riprap can provide up to three times the shear strength of non-grouted riprap (Minnesota DOT 2015b). The grout penetrates the void spaces, connecting the stone contact points. Optimal fill for matrix riprap in the total void layer is 50%, where one-third of the bottom layer is filled and two-thirds remains in the top layer. Grout consistency is important and should be approximately 7 inches in a slump test so that it has the proper viscosity (i.e., it does not all remain in the top layer, nor does it all flow to the bottom). In lieu of the slump test and for more accurate results, Minnesota DOT uses the European Tapping Table spread before and after agitation. Grout place- ment is accomplished with a concrete pump truck with hose diameter of 2 to 3 inches; the hose operator will fill the riprap voids as specified above (Minnesota DOT 2015a). Use of geobags or geotextile containers are another countermeasure recently used by Minnesota DOT. Geobags consist of a geotextile filter bag that contains coarse aggregate (Minnesota DOT 2016). Geobags are placed next to bridge abutments with a clamshell. Geobags are approximately 4 ft wide x 4 ft long x 2 ft thick, weighing approximately 1900 lbs., and with 400 lbs. of fabric tensile strength as specified by Minnesota DOT 2511. Minnesota DOT 2511 specifies that a bag is filled with a coarse filter aggregate using a standing funnel and a loader, and then the open end is stapled (Minnesota DOT 2012). Additionally, Minnesota DOT is active in bridge monitoring technology, where 3D sonar is used to image underwater conditions to document defects for condition assessment. This technology is used in scour evaluation, construction pre-planning, flood damage assessment, and in diver safety applications (Minnesota DOT 2017c). New Mexico Previously, New Mexico DOT analyzed a limited number of bridges using the U.S. Geological Survey (USGS) 96-4310 Method for Rapid Estimate of Scour at Highway Bridges Based on Limited Site Data (USGS et al. 1997). New Mexico DOT recently addressed bridges where scour calculation and evaluations have not been made (FHWA 2001) and began evaluating them using the Rapid Scour Method (Level 1) (USGS et al. 1997). Based on the results of the Rapid Scour Method analysis, bridges were either reclassified as NBIS Item 113 ratings of 5, 7, or 8, where the bridge foundation was determined to be stable (USGS et al. 1997), or further analysis was performed. For those bridges requiring further analysis, more rigorous hydrology was devel- oped, and all bridges were analyzed using HEC-RAS and HEC-18 (Level 2). Some bridges were reclassified as scour-critical after the Level 2 analysis, and these bridges are being replaced or countermeasures designed and installed based on severity of risk, budget available, and schedule (Morgenstern 2019). This change was not a radical change in methodology used in engineering analysis but instead a change in management approach. New Mexico DOT is in compliance with the FHWA Metric 18 Inspection Procedures Scour Critical Bridges, has a much better understanding of the scour susceptibility or stability of an increased number of bridges, and has an understanding of where projects and funds are needed for scour mitigation (Morgenstern 2019). Texas The Texas DOT geotechnical manual update included updating the scour section with more detailed guidance for various bridge/culvert types. This helps improve consistency and qual- ity. This edition includes definitions of the following bridge/culvert types: new bridges with known foundations, existing bridges with known foundations, existing bridges with unknown foundations, bridge class culverts, and scour-critical bridges. Recommended methods for

Literature Review 15 performing scour analysis for the different bridge/culvert types are provided (Nuccitelli 2019, Texas DOT 2018a). HEC-18 is widely used, and scour calculations are performed using this guidance; however, calculations using this provide maximum depths and not allowable limits or critical depth thresholds for single scour components. The critical depth threshold is typically not the same as the maximum as provided by HEC-18 (FHWA 2012b). Texas Secondary Evaluation and Analysis for Scour (TSEAS) manual (Texas DOT 1993) provides maximum allowable depth for pier scour. A new study, Allowable Limit Contraction Scour and Abutment Scour at Bridges, is directed at determining maximum allowable scour depths for abutments related to potential abutment failure and contraction scour related to bridge pier failure (Texas DOT 2017, 2018b). Factors considered in these new criteria are soil slope, the support structure for the abutment deck, the supporting embankment for the access roadway, and embankment material geotechni- cal stability (Texas DOT 2017, 2018b). Virginia Virginia DOT has made three substantial changes to scour policy and practices, allowing inclusion of practical judgment and thus resulting in more shallow foundations for more cost- effective design practices. The changes include: 1. During the evaluation of the rock core barrel, material with a rock quality designation (RQD) of less than 50% can now be considered as granular material based on the contents to a max D50 of the core barrel. This allows Virginia DOT to account for the material resistance that is less than actual interlocked material but is greater than the overburden (Virginia DOT 2019c, USGS 2018, Matthews 2019). 2. When evaluating live bed scour conditions, Virginia DOT also uses clear water equations when available to assess the effects of the D50 of the material on the scour potential. This is especially helpful when the material is increasing in size with depth. If the clear water equa- tions indicate that a particle is too large to scour at a given depth, then this should be the limiting factor regardless of the results of the live bed equations. This approach works well when applying the change in material evaluations as stated above (Matthews 2019). 3. Based on guidance from FHWA HEC-18, Virginia DOT focuses on the NCHRP Project 24-20 “Abutment Scour Results Over HIRE and Froelich” (Matthews 2019). For scour counter- measures, Virginia DOT uses the riprap abutment guidance in HEC-23 but will adapt the apron as necessary to increase the thickness if the defined extents cannot be met (Matthews 2019). Virginia DOT is also participating in the FHWA field pilot study of the in-situ device for the determination of the clay shear resistance and hopes to adopt the shear force method soon (FHWA 2018c, Matthews 2019). West Virginia The West Virginia DOT drainage manual specifies that scour computations should be per- formed in accordance with HEC-18 (West Virginia DOT 2015). A change was made to adopt scour calculation methods presented in NCHRP Project 20-24. Using NCHRP 20-24 is more time consuming but provides more realistic scour predictions. The previous methods used by West Virginia DOT are those in HEC-RAS, so scour calculation time was short (Kirk 2019). Since West Virginia is primarily located on the backbone of the Appalachian Mountains, rock is relatively shallow, and scour is not a major problem unless the rock is erodible (Kirk 2019, West Virginia DOT 2019). In the case of erodible rock, a scour prediction technique has been developed based on foundation inspection, coring, measuring scour depths, rock-water interac- tion, and gage flow measurements to rate the rock on a scale developed for scour versus stream power (Keaton et al. 2012, FHWA/WVDOT 2010–2012).

16 Highway Hydraulic Engineering State of Practice Advanced Hydraulic Modeling State information available focused on 2D hydraulic modeling, as most states responding indi- cated little use of 1D unsteady, 3D steady/unsteady, or sediment transport modeling in policy or practice. New Mexico and Rhode Island indicated sediment transport capacity calculations were previously required, and California and Ohio indicated policy pending, but information regarding use of 2D or 3D modeling was not located. Well-known limitations of 1D models are that velocities are averaged over cross sections and flow is assumed normal to the cross section, which may lead to inaccurate representation of flow and velocity distributions (Pierce 2018). For example, Figure 3 shows velocity distributions depicted in a 2D model that are not normal to the cross section. 2D models are accepted by FEMA if the specified criteria are met (FEMA 2019). Decisions to change practice to 2D hydraulic modeling have involved evaluation of technical aspects, time requirements, available data, and experience (Gosselin et al. 2006). In 2D hydraulic models, the depth-averaged velocity and direction of flow are calculated at every point in the domain, thus avoiding many assumptions required by 1D models (Pierce 2018), specifically that variables (velocity, depth, etc.) change predominantly in one specified direction along the channel. Since channels are rarely straight, 2D hydraulic models allow for more accurate rep- resentation of water surface profiles (Wagner 2007), flow conditions (Gosselin et al. 2006), and contraction scour (Rossell and Ting 2013). Two-dimensional models may also provide more accurate hydraulic representation of bridge crossings, especially for natural channels, compound channels (well-defined channel with floodplains), bridges on a skew, and crossings with multiple openings in a wide floodplain. To overcome 2D modeling challenges, FHWA Everyday Counts initiatives include use of next-generation hydraulic applications to improve understanding of interactions between trans- portation assets and river environments (FHWA 2019). Many state DOTs have joined FHWA in this effort as FHWA has partnered in pilot projects and provided National Highway Institute Courses with free licenses for DOTs (Nguyen 2019). Many states are also acquiring statewide Light Detection and Ranging (LIDAR) data, which may help overcome challenges associated with limited or unreliable data (Delaware DOT 2017a, Carleson 2019). Colorado Colorado DOT uses 1D steady-state models, generally HEC-RAS, for most hydraulic studies due to lower complexity and less data requirements. The Colorado drainage manual states that 5.0 4.0 3.0 2.0 1.0 0.0 Vel_Mag_ft_p_s Figure 3. Velocity distributions depicted in a 2D model that are not normal to the cross section. Courtesy of Georgia DOT (2019).

Literature Review 17 steady and unsteady models may be applied to complicated hydraulic conditions such as large overbank bridge crossings, large variations in roughness, multiple channels, and islands, where 1D models may lead to costly or improper overdesigns (Colorado DOT 2009). Georgia Georgia DOT incorporated 2D modeling as policy because traditional 1D modeling would not have accurately represented hydraulic conditions in a number of important cases (Georgia DOT 2018b). These include locations with wide floodplains, skewed embankments, significant roughness, complex geometry and ineffective flow areas, and where more accurate flow patterns are necessary to design countermeasures. Accepted modeling software programs as outlined in the Georgia DOT Drainage Design for Highways Revision 3.4 are Surface Water Modeling System version 11.1.4 (SMS), Finite Element Surface Water Modeling System version 3.22 (FESWMS/FST2DH), and Surface Water Modeling System version 4.56 (RMA2), each with specific strengths. In general, these software programs provide a comprehensive hydrodynamic environment for 1D and 2D surface water models and finite-element hydrodynamic computa- tions of horizontal velocity and water surface elevations in open channels under subcritical flow regimes (Georgia DOT 2018a). Michigan To overcome difficulties and FEMA permit acceptance, Michigan DOT practices routinely include comparing water surface elevation results; for example, Michigan DOT is currently applying 1D HEC-RAS and SRH-2D via SMS in parallel for a culvert replacement project (Carleson 2019). This practice has gained enough traction in Michigan such that FEMA now accepts 2D hydraulic models in the state (Carleson 2019). New York New York State DOT generally uses HEC-RAS for most hydraulic studies, but the New York State Bridge Manual states that 2D hydraulic analysis can be performed using SMS at com- plicated locations, such as multiple-inlet tidal bays or condolences to more accurately model geometry (New York State DOT 2019). Texas Texas DOT completes a few models using 2D analysis. The current Texas drainage manual states that 2D modeling techniques are highly specialized and recommends contacting the Design Division’s Hydraulics Branch for consultation (Texas DOT 2016). However, Texas DOT has seen a significant increase in 2D modeling in the past few years and is developing new policy guidance for 2D modeling that is planned to be released in early 2020. Utah Utah DOT specifies that 1D or 2D models may be used for culverts conveying irrigation water or culverts located in a FEMA Special Flood Hazard Area (SFHA). Additionally, a 2D model must be used for bridge hydraulic analysis, or else a Deviation from Drainage Design Criteria form must be submitted when a 1D model is used instead of a 2D model. The request for devia- tion must specifically address how the channel will be accurately represented using 1D (Utah DOT 2018a). Vermont Vermont uses 1D steady-state models, generally HEC-RAS, for most hydraulic studies because of the lower data requirements and time constraints. Vermont recently conducted a

18 Highway Hydraulic Engineering State of Practice pooled fund study with the FHWA in which they identified more applicable models for unsteady tidal hydraulics and associated scour analysis: UNET, FESWMS-2D, and RMA2 (Vermont DOT 2017). Regulatory Requirements Regulatory requirements establish a process for consultation and compliance with one or more federal laws. These agreements allow for predictability for projects in addition to intro- ducing standards (Center for Environmental Excellence 2018). Federal laws include the NPDES and the NEPA. State highlights include memoranda of understanding for permits for projects disturbing more than 1 acre, along with required LID solutions for treatment of municipal separate storm sewer system discharge (MS4). LID minimizes negative environmental impacts when properties are developed and impervious surface area increases. LID is aimed at mimick- ing pre-development conditions through techniques such as bioretention, permeable pavement, and grass swales (Dietz 2007). LID typically focuses on controlling quantity not quality of water (Davis 2005); however, some Best Management Practices (BMPs) used in conjunction with LID treat both water quality and quantity while others target one specifically (Ohio DOT 2018). State Oversight for FHWA National Environmental Policy Act (NEPA) and Clean Water Act (CWA) Section 404 New York New York State DOT maintains erosion and sediment control under NEPA and CWA. All federally aided projects involving clearing, grubbing, grading, or excavation must have erosion control and require sediment control plans included in plans, specifications, and estimates. In New York State a memorandum of understanding for the State Pollutant Discharge Elimina- tion System (SPDES) General Permit covers development in which the excavation is more than 1 acre of non-Tribal Indian land. An EPA NPDES Construction General Permit is necessary if Tribal Indian lands of 1 acre or more are disturbed and involve a stormwater discharge to surface water. A project is not eligible for a SPDES general permit if 2 or more acres of land with no existing impervious cover are disturbed, the land is characterized as Soil Slope Phase E or F, or if the project is within watersheds of AA or AAs waterbodies; in these cases, the permitting is deferred to NPDES (New York State DOT 2018). Compliance with Local Floodplain Requirements Ohio Ohio DOT requires use of HEC-RAS to model conveyance showing artificial effects on flood- plain encroachments, with an allowable water surface surcharge (up to 1 ft) established as a means of floodplain management, in accordance with the National Flood Insurance Program (NFIP) rise allowance. Highways that encroach on floodplains require design to the 100-year flood event to minimize property damage. If construction occurs within a FEMA A Zone, an Ohio DOT self-permit process is required along with coordination with a Local Floodplain Coordinator who determines the allowable surcharge. The Local Floodplain Coordinator is the liaison between the NFIP and the Ohio Department of Natural Resources Floodplain Manage- ment Program (FMP) for work in a FEMA Special Flood Hazard Area (Ohio DOT 2018). Tennessee and Colorado Tennessee DOT includes an Executive Order for floodplain management to avoid long and short-term impacts of floodplain modification, as well as to restore and preserve values presented

Literature Review 19 by floodplains. This applies to federal buildings, structures, roads, or facilities impacting a flood- plain (Tennessee DOT 2012). A similar Floodplain Management Executive Order is in place in Colorado (Colorado DOT 2004). Utah FEMA requires a Floodplain Development Permit from local jurisdictions in Utah, along with obtaining a Conditional Letter of Map Revision causing any change to the base flood elevation (BFE) when a floodplain is encroached. This allows for documentation of a changing floodplain as well as any necessary revisions to a previous FEMA map. For channels, culverts, and bridges, coordination with a Floodplain Administrator is required for proper floodplain management (Utah DOT 2018b). Requirements for NPDES and Requirements for LID (e.g., bioretention, permeable pavement, etc.) Delaware Delaware DOT aims to maintain water quality and reduce non-point source pollution under NPDES using biofiltration swales, bioretention areas, infiltration ditches, and retention ponds. Examples of a biofiltration swale under construction and a completed biofiltration swale are shown in Figure 4. Delaware DOT uses multiple or redundant LID measures for increased reli- ability, and these measures include maximizing retention of native forest cover, use of natural topo- graphic features, creating a hydrologically rough landscape to slow stream flows, and increasing time of concentration (Delaware DOT 2008). Hawaii Effective as of 2013, Hawaii DOT incorporates LID solutions for treatment of MS4 per the EPA NPDES; updates in the DOT manual include and prioritize these practices. Swales, bio- retention, infiltration trenches, and engineered wetlands are required LID options to treat the water quality design volume (WQDV). Figure 5 shows an example of a bioretention pond under construction. Any new development or redevelopment is required to have LID stormwater man- agement if an acre or more of impervious surface is generated. However, circumstances arise when LID is impossible or unsafe to implement. These variances are based on hydrogeological, Figure 4. Biofiltration swale under construction is shown on the left and a completed biofiltration swale on the right. Reprinted from Wikimedia Commons. 2010.

20 Highway Hydraulic Engineering State of Practice physical, and operational constraints, and all solutions must be considered before a waiver is granted (Hawaii DOT 2015). Maryland Maryland DOT incorporates LID practices (Maryland DOT 2009). The 2009 Maryland Transportation Plan addresses environmental challenges through coordinating land use and transportation planning to promote “Smart Growth,” preserving natural resources, and sup- porting environmental initiatives to commit to environmental quality. These ideas are based on the Smart, Green and Growing initiative enacted by Maryland state legislation (Maryland DOT 2009). Virginia Virginia DOT also incorporates LID practices (Virginia DOT 2019c). In Virginia, LID solu- tions include preserving riparian buffers, wetlands, steep slopes, mature trees, floodplains, woodlands, and highly permeable soils (Virginia DOT 2019c). Requirements for National Pollutant Discharge Elimination System (NPDES) and Clean Water Act (CWA) Section 404 Alaska In October 2008, the Alaska Pollutant Discharge Elimination System program was imple- mented in a phased transition to assume primacy for NPDES permitting. The Alaska Depart- ment of Environmental Conservation became the stormwater permitting authority in Alaska; however, the EPA is still involved with permitting, compliance, and enforcement. Alaska DOT also requires use of BMPs such as detention ponds, treatment swales, temporary vegetative buffer strips, and vehicle tracking entrance/exit designs (Figure 6) to protect water quality. Alaska DOT regulates stormwater from MS4s, discharges from industrial activity, and stormwater from construction sites, with each permit type having different requirements under NPDES (Alaska DOT 2011). Figure 5. Leucaena leucocephala habitat and water retention basin, Waikapu-Maui. Reprinted from Wikimedia Commons. 2009.

Literature Review 21 California The Caltrans NPDES Statewide Stormwater Permit regulates stormwater and non-stormwater discharges from Caltrans properties. Construction must comply with a construction general permit, implement a year-round program to control discharges, and meet water quality stan- dards through BMPs such as biofiltration swales and strips, earthen berms, sand filters, and solids removal devices. Since 2013, projects in Caltrans’ right-of-way must also incorporate new post-construction stormwater treatment requirements. A separate NPDES permit is needed if discharge is anything other than stormwater (Caltrans 2017). Georgia The EPA administers CWA permitting in Georgia; however, individual and general NPDES permits are issued at the state level. A Notice of Intent (NOI) must be submitted to the Georgia Environmental Protection Division. There are three general permit types: Stand-alone Construc- tion Activity, Infrastructure Construction Sites, and Common Development Construction. These permits are reissued every 5 years, along with an NOI and an Erosion, Sedimentation, and Pol- lution Control plan. MS4 permitting requires consideration of LID and green infrastructure. Georgia considers water quality early in the design process and implements green infrastructure options that include grass channels in place of curb and gutter in rural sections, bioretention basins, open-graded friction course (substituted for conventional asphalt), and stormwater wetlands (Georgia DOT 2018a). Nebraska Nebraska Department of Roads (NDOR) is required by the FHWA to identify erosion and sediment-sensitive areas in addition to NPDES and erosion and sediment control plans for sites 1 acre or larger. MS4 permitting develops strategies to include a combination of structural and/or nonstructural treatment BMPs. A local public agency (LPA) is permitted as an MS4 and operates under its own NPDES, reviewed by the Nebraska Department of Environmental Quality and EPA. Any LPA stormwater program overrules a NDOR stormwater program except when an LPA project is being constructed on a state or federal highway in an MS4 area (NDOR 2013). Figure 6. Vehicle Tracking Schematic. Source: Created by Tetra Tech for the U.S. EPA and the Kentucky Division of Water.

22 Highway Hydraulic Engineering State of Practice As a part of the Clean Water Act, a Total Maximum Daily Load (TMDL) is created for pol- lutants impairing a given body of water. This includes a treatment standard to lower pollutant levels in all bodies of water entering the water of interest. Highways may have additional require- ments if stormwater discharges into waters of interest with a TMDL established. For LID solu- tions for excess stormwater, NDOR implements vegetated filter strips, grass swales, infiltration trenches, infiltration basins, bioretention, media filters, extended dry detention, wet detention ponds, stormwater wetlands, pervious pavements, proprietary structural treatment control, and other BMPs (NDOR 2013). Ohio Ohio DOT requires a one-page NOI application for NPDES projects. The application is a certification provided by the applicant that ensures compliance of NPDES. These projects are earth disturbances of >1.0 ft and are projects susceptible to stormwater erosion. A NOI or post- construction BMPs are not required for routine maintenance projects less than 5 acres. However, the Ohio EPA has designated certain watersheds that require NPDES permits post-construction BMPs and also groundwater recharge mitigation, riparian setback mitigation, and temporary sediment basin locations. Based on earth disturbing activities, post-construction BMPs for construction include extended detention, retention basin, bioretention cell, infiltration meth- ods, and constructed wetlands as necessary to provide stream protection. Additionally, post- construction BMPs for any non-routine maintenance activities are required to treat runoff when disturbed earth for a project is equal to or greater than 1 acre (Ohio DOT 2018). Pennsylvania NPDES Permitting, Monitoring, and Compliance discharges associated with industrial activity and discharge from storm sewers are included in the drainage manual under Commonwealth of Pennsylvania, Title 25, Chapter 92 (Pennsylvania DOT 2015). Tennessee and Colorado Both Tennessee DOT and Colorado DOT abide by a Nationwide General Permit granted by the USACE. This permit involves Section 404 of CWA, which states a federal permit must be obtained for any activity impacting water quality of U.S. waterways, along with a water quality certification from their respective environmental state departments (Colorado DOT 2004, Tennessee DOT 2012). Floodplain Impacts and Mitigation Floodplain impacts and mitigation pertain to the requirements an agency must meet in regard to improvements in the floodplain. Agencies compare existing conditions to FEMA recorded and modeled water surface elevations in FEMA regulatory floodways. Agencies must comply with published water surface elevations and/or zero rise. Several states have made changes to policy and practice regarding mitigation and no adverse impact/zero rise, as indicated by survey responses. Colorado In 2017 the Colorado DOT Region 4 Hydraulics Unit issued Standard Operating Procedures specifically for pavement treatment construction and maintenance activities in FEMA regulatory floodways. For crossing or contact with any regulatory floodway, Colorado DOT must apply for a floodplain permit from the local agency of governance. Conditions to be achieved with any development are “no rise” or no change in 100-year water surface elevation or BFE; otherwise,

Literature Review 23 a Letter of Map Change from Colorado DOT must be sent to the FEMA Flood Insurance Rate Map (FIRM) requesting a revision and certification (Colorado DOT 2017). Even though final certification for floodway mitigation of highway projects is ultimately at the discretion of FEMA, Colorado DOT Region 4 general procedures for impacts from pavement treatment construction and maintenance activities are aimed to simplify the required Colorado DOT certification for these less complex projects. Colorado DOT has Standards of Practice (SOPs) intended for the least complex activities for “no-rise” in Colorado DOT applications for FEMA certification. These activities include patching, pavement maintenance, and the majority of mill and overlay projects, and SOPs indicate the level of detail and time frame required for engineering and planning, thus leading to better understanding of the time frame for compli- ance. Additionally, Colorado DOT mentions that the SOPs are not an exhaustive list, and each project should be considered on a case-by-case basis (Colorado DOT 2017). Maryland The Maryland SHA recently revised procedures to combine separate MDE and FEMA hydrau- lics models so that only one application is necessary for acceptance. Previously, MDE would review and approve models separately from the FEMA model updates according to Letter of Map Revision (LOMR)/Conditional Letter of Map Revision (CLOMR) procedures. Addition- ally, the regulations were different—MDE allows water surface level increases of 0.10 ft on prop- erties that were insurable per FEMA, but FEMA regulations call for “no rise” or 0.00 ft increase in 100-year water surface level (Maryland SHA et al. 2017). To accomplish the joint effort, FEMA and the MDE first began creating a Digital FIRM (DFIRM) database using Geographic Informa- tion Systems (GIS) through a DFIRM Outreach Program, and now 75% of the FIRM maps are no more than 3 years old (Maryland SHA 2016; FEMA 2019). The new joint modeling process between the Maryland SHA and FEMA involved the Maryland SHA Office of Structures creating milestones and revising the Hydraulics and Hydrology Design Manual Chapter 5 in 2016 and again in 2017. Maryland SHA tested the updated procedure with a pilot project: Maryland Route 144 over Evitts Creek. For this project, Maryland SHA used a 2017 FEMA model and best available 2015 FEMA map, extended the model limits, and incorporated Maryland SHA cross sections. The Maryland SHA floodway was wider than the FEMA floodway, and the resulting proposed condi- tions model included the revisions from the MDE FEMA coordinator’s review (Maryland SHA et al. 2017). In summary, the Maryland SHA and FEMA integrated modeling follows the new FEMA/MDE CLOMR Process Flow Chart from the Office of Structures Hydraulics and Hydrology Design Manual Chapter 5, Figure 2, (Figure 7) and also revisions to the Maryland SHA Hydraulics Report Checklist (Maryland SHA 2017). Minnesota Minnesota DOT has a statewide flood mitigation program that requires completion of hydraulic flood analysis, floodplain assessment, and a risk assessment for encroachment design. The risk assessment for encroachment design collects and reports detailed information regard- ing flood damage to existing structures such as shopping centers, hospitals, chemical plants, power plants, and housing developments during 100-year (1% AE) and 500-year (0.2% AE) flood frequency events. Reporting seeks to find the value of structures and contents and aims to justify whether further analysis and possibly redesign is needed to minimize damage. Traffic- related losses in the form of annual capital costs are evaluated for the 100-year and 500-year flood frequency events. Embankment, roadway, and scour protection costs are weighed against additional culvert and bridge capacity costs (Minnesota DOT 2011a).

24 Highway Hydraulic Engineering State of Practice Figure 7. Integrated MDE/FEMA Submittal Process Flow Charts. Reprinted from Maryland SHA. 2017., Maryland Hydraulics Panel. 2017. http://gishydro.eng.umd.edu/hydraulics_panel/report_2018/Hydraulics_ Panel_Report_2018.pdf.

Literature Review 25 Additionally, Minnesota DOT has many ongoing flood mitigation projects across the state. One example is the Minnesota River Flood Mitigation Study initiated by Minnesota DOT to investigate feasible designs at river crossings in the Minnesota River Valley that minimize flood risk during seasonal flooding. This study involved stakeholder input, traffic analysis, historical floods and modeling, cost estimation, and alternatives analysis. Alternatives analysis was con- ducted and funding pursued by Minnesota DOT (2011b). Pennsylvania Pennsylvania DOT is conducting a multi-phase effort in order to prepare for the conse- quences of extreme weather for a resilient transportation system. Phase I Pennsylvania DOT Extreme Weather Vulnerability Study is the first step to prioritize funding needs. The study included historical risk assessment, vulnerability mapping and forecasting, development of a damage cost system, best practices, etc. The study also included analyzing impacts to the state road system during 1% annual chance or 100-year flooding impacts. In the last decade, $140 million emergency funds were allocated to repair damaged Pennsylvania DOT federal- aid infrastructure, roadways, and bridges, due to flooding from weather related events such as Hurricane Sandy. These events have severely affected the planned life cycle of the infrastructure (Pennsylvania DOT 2017). In the second part of the Phase I Pennsylvania DOT Extreme Weather Vulnerability Study, Pennsylvania DOT in conjunction with FHWA has selected projects to conduct adaption studies to evaluate impacts due to climate change. The projects will include hydrologic and hydraulic analysis with precipitation forecasts, design options for adaption, and economic analysis. Out- comes will include a detailed hydrologic and hydraulic analysis template for climate change impact inclusion in studies, case example review for strategies and cost effectiveness of adaption designs, and evaluation of precipitation data and flooding forecasts (Pennsylvania DOT 2017). Phase II Resilient Designs starts with two workgroups. One internal workgroup will manage design, maintenance, and construction, and the second will focus on traffic operations. Resilient designs will be multi-year, and implementation of some items will start 6 months or more after the initiation of Phase II and span for several years. Items will range from embankment loss pre- vention and pipe encapsulation using geotextiles to updating the Pennsylvania DOT hydraulics and hydrology manual. The manual updates will include USGS equations revised with current precipitation data and also any stream flow statistics updates. Modifications to bridge opening design will include provisions for no allowable backwater or 0.00 ft rise. Scour design modifica- tions will include evaluation of the 100-year and 500-year storm events. Foundations will be designed for the 100-year event, and 500-year stability will be checked. Culvert design is similar to bridge design, checked for the 100-year storm event and impacts downstream, with possible bridge opening increases of 20% suggested. Stabilization measures such as rock embankment slopes or interlocking block designs will be added (Pennsylvania DOT 2017). Coastal Hydraulics Shore protection along coastal zones that are subject to storm surge and wave attack can be a critical element in highway design, construction, and maintenance. State response to the survey was limited; however, several state hydraulic design and drainage manuals have sections dedicated to coastal shore protection. State coastal engineering practices include vulnerability assessments based on sea level rise projections, advanced coastal flood risk modeling, and assess- ment and design practices to elevate, harden, abandon, or apply nature-based solutions. Para- metric models have also been developed to determine design storm surge and wave parameters for computing surge/wave loads on bridge superstructures (e.g., OEI, Inc.). Solutions include practices to elevate, harden, or abandon, as well as nature-based solutions.

26 Highway Hydraulic Engineering State of Practice With the increased use of coastal hydrodynamic modeling by transportation engineering professionals, the FHWA issued a manual (Webb 2017) to provide guidance on the use of coastal models in the planning and design of coastal highways and bridges, as well as when to solicit the expertise of a coastal engineer. Specifically, the manual provides information needed to determine scopes of work, prepare requests for professional services, communicate with consultants, and evaluate modeling approaches and results. The manual also provides guidance on the use of hydraulic and hydrodynamic models, including how they are used to determine the dependence of bridge hydraulics on the riverine or coastal design flood events. Finally, the manual gives recommendations for the use of models in coastal vulnerability assessments (Webb 2017). Other emerging technologies promise to advance approaches for coastal flood protection. For example, Unmanned Aerial Vehicle-based remote sensing is being used to conduct high spatial and temporal resolution assessment of infrastructure conditions and mapping shoreline condi- tions. Shilling (2019) combined this approach with real-time kinematic geo-positioning systems (RTK-GPS) field measurements for multi-scale visualization of shoreline change due to sea level rise. Shilling (2019) also demonstrated high-resolution terrain-mapping methods for estimating coastal flood impacts on coastal infrastructure and adjacent ecosystems, and described how this information could be used to validate sea level rise models and inform adaptation planning for shoreline infrastructure. In addition to the recent FHWA manual on coastal hydrodynamic modeling, a revised ver- sion of FHWA’s HEC-25 Highways in the Coastal Environment is planned for publication later in 2019. California The Caltrans (2016) Hydraulic Design Manual has a chapter focusing on the procedures, methods, and materials commonly used to mitigate shoreline erosion and attendant damage to transportation facilities. The guidance focuses on quantifying exposure to sea level rise, storm surge, and wave action (i.e., design high water, wave height, and wave break). The manual pro- vides design guidance for various types of armor protection, including revetments (rock slope protection), bulkheads (concrete or masonry walls, crib walls, and sheet piling), sea walls, and groins. The manual also states, “The practice of coastal engineering is still much of an art . . . for a variety of reasons including that the physical processes are so complex, often too complex for adequate theoretical description, and the design level of risk is often high” (pp. 880–1), which indicates that additional research is needed (Caltrans 2016). Delaware Nature-based solutions include a range of natural vegetation and other landscape features that serve as alternatives to, or ecological enhancements of, traditional techniques for shoreline stabilization and infrastructure protection. Nature-based solutions have been used extensively in diverse coastal settings such as in SR1 rehabilitation in Delaware (FHWA 2018b). Under- standing green engineering tools and methods for designing nature-based solutions can achieve alternative solutions (FHWA 2018b, Webb et al. 2018) addresses these issues through examples of nature-based solutions, highlighting the best available science that describes their perfor- mance as solutions for coastal highway resilience. Further, an implementation guide is planned based on monitoring and peer exchanges and will address the issues outlined here for the use of nature-based solutions (Webb et al. 2018). Traditional construction includes existing gray coastal stabilization infrastructure and drainage systems subject to erosion, sedimentation, and clogging, such as those experienced along SR1. Adaptation projects or green infrastructure along SR1 included raising and reshaping the dunes,

Literature Review 27 preserving existing and creating additional marshes, and protecting the eroding shoreline using oyster reefs and oyster shell bags (Figure 8). Engineered features that mimic natural features include naturally occurring materials instead of concrete or rock revetments and riprap. These naturally occurring materials include coastal vegetation, dunes, wetlands, maritime forests, beaches, and reefs. Larger culverts, or additional culverts with tide gate systems at the outlets, may be installed to improve the drainage system capacity (FHWA 2018b). Implementing green stormwater infrastructure in coastal areas poses challenges. Property owner access during construction and planning for long-term access is a typical challenge. Lighter construction equipment and construction dewatering are often necessary due to work- ing with sensitive habitats (FHWA 2018b). In Glenville, Delaware, FEMA and the Delaware DOT came together with stakeholders to relocate 172 families whose homes were located in the 100-year floodplain and had been subject to repeated flooding events. This approach has multi-use benefits of restoring wetlands and habitat, providing recreational area, and expanding flood storage (FEMA 2011). Support pro- grams, such as shared development of the University of Delaware Institute for Public Adminis- tration (IPA)toolkit for community development and land use planning, can help stakeholders avoid building in flood prone areas. Additionally, Delaware DOT has completed projects with partners such as the Department of Natural Resources (DNR) to combat chronic flooding (Delaware DOT 2017b). Florida The US 98 project on Okaloosa Island in Florida focused on the vulnerability of the barrier island roadway to overwashing from sea level rise and storm surge. Specifically, the assessment evaluated a Florida DOT critical coastal roadway with a buried sheet pile wall and gabions installed along the shoulder as a countermeasure to reduce erosion from overwashing during storm events. Overwash and wave velocities that scour out the road shoulders and damage the pavement (referred to as the “weir-flow-damage mechanism”) is a common cause of storm damage to roadways that run parallel to the coast. The study concluded that the buried sheet pile and gabion mattress measure is an economic adaptation option (FHWA 2016b). Figure 8. Building oyster reef breakwater structures or castles. Reprinted from Wikimedia Commons. 2018.

28 Highway Hydraulic Engineering State of Practice Louisiana Sheppard et al. (2016) conducted a hindcasting analysis of 50 of the most severe storms that have impacted the Louisiana coast over the past 160 years, including hindcasting of alternative paths for a selection of storms. Extreme value analyses were then performed on water eleva- tions, wave heights and peak periods, and wind speeds to produce a GIS-based Wave and Surge Atlas for the design and protection of coastal bridges in South Louisiana. Other products of the study included developing an AASHTO Wave Load Calculation Program based on the AASHTO Guide Specifications (AASHTO 2008), providing a training session for Louisiana DOT and Development (DOTD) employees so that DOTD will be able to update or modify the program as needed. The study also allowed Louisiana DOTD to compute the forces and moments on the spans of bridges determined to be vulnerable. New York Tang et al. (2018) conducted a vulnerability study of coastal bridges in the New York City (NYC) metropolitan region to storm surges and waves. Predictions were made for potential surges, waves, and consequent hydrodynamic loading and scour at bridge piers under condi- tions of sea level rise and a change in hurricane patterns. Computer modeling of storm surges and waves in the greater NYC area during Hurricane Sandy compared reasonably with field observations. Study results indicated that both sea level rise and changing hurricane patterns could result in a significant increase in hydrodynamic load and scour at bridge piers, and such potential increases should be considered in the development of resilient coastlines. North Carolina The North Carolina DOT commissioned a study by Ocean Engineering International, PLLC (2013) to determine the design storm surge and wave parameters needed to compute the loads on bridge superstructures deemed vulnerable to coastal storms. To perform this analysis, 62 of the most severe tropical storms and hurricanes that have impacted North Carolina coastal waters over the past 160 years were hindcasted, and extreme value analyses were performed on water elevation, wave heights, and depth-averaged current velocities to obtain 1% annual exceed- ance design conditions. To increase the data set for the extreme value analyses, the hindcasted storm paths were shifted to the right and left of the actual path and the modified-path storms hindcasted, resulting in a total of 186 storm paths. The results from the extreme value analyses were presented in a GIS database for ease of access and use (Ocean Engineering International PLLC, 2013). Computation of the surge/wave loads on bridge superstructures requires knowledge of the superstructure type, dimensions, and span low chord elevation, as well as the design water eleva- tion and wave parameters. A proprietary computer model developed by OEA, Inc. produced the data for development of the parametric equations in the AASHTO code “Guide Specifications for Bridges Vulnerable to Coastal Storms,” which were used to compute the surge/wave loads on the bridge superstructures. Following a Level I screening analysis, 191 bridges were selected for evaluation, and 105 were classified as potentially vulnerable to these types of loads (Ocean Engineering International, PLLC 2013). Oregon The FHWA supported a study by Oregon DOT to analyze how green, or nature-based, infra- structure, could mitigate storm impacts and coastal bluff erosion along the Oregon Coast High- way. Building on prior Oregon DOT research on dynamic revetments (i.e., cobble beaches or berms), and incorporating lessons learned from similar projects, the study developed conceptual design plans for three high-risk sites. Further analysis of the preferred designs was conducted

Literature Review 29 and included protection against anticipated future coastal impacts, estimated construction and maintenance costs, and implementation benefits and challenges (Oregon DOT 2014, Oregon DOT 2016). Rhode Island Given uncertainties in long-term projections of sea level rise, a scenario-based approach is typically used. For example, the Rhode Island Statewide Planning Program (2015) used a GIS- based approach to analyze the exposure and vulnerability of transportation assets, including roads and bridges, under 1, 3, and 5 ft of sea level rise. The study, however, did not consider projections of erosion, storm surge, or precipitation. Also, it was noted that there may be inland areas affected by sea level rise that are not included in the projections, and high tide and sub- sequent sea level rise scenarios may be higher in inlets. At longer timescales, Knott et al. (2017) assessed the impacts of rising groundwater levels from sea level rise on the service life of coastal road pavements. Texas The current manual (Texas DOT 2016) does not provide specific design guidance for coastal flood protection. However, it directs the designer to determine the FEMA SFHA zone designa- tion with the current, correct FIRM. In Texas, zones V, VE (V1-30), A, and AE are coastal zones flooded by the Gulf of Mexico stormwaters rather than riverine flows. It is recognized that FEMA modeling for coastal zones is not the same as for riverine modeling. The designer is not required to acquire the model but is directed to consider tidal flows, wave action, and storm precipitation, as well as to ensure the project will not trap flood waters or block drainage (Texas DOT 2016). More extensive guidance for coastal hydraulics, including risk assessment, modeling, and design, is to be included in the next version of the manual. Washington Washington DOT conducts vulnerability assessments for infrastructure planning and miti- gation, accounting for sea level rise as a long-term trend that threatens coastal transportation infrastructure (Washington State DOT 2017a). Guidance addresses potential impacts of sea level rise, higher storm surge, and more frequent and extensive inundation of low-lying areas (both temporary and permanent), including the following: coastal erosion and landslides that weaken roadbed and bridge footings; damage to stormwater drainage and tide gates; saltwater corrosion of facilities; and detours around frequently flooded coastlines (Washington State DOT 2017a). Alternative Project Delivery Methods The aim of this section is to discover and document alternative delivery policy for the hydraulic engineering aspects of projects. Thirteen out of the 38 responding states have hydraulic policy in RFP documents to accommodate alternative project delivery methods, while three have policy pending. RFPs are the mechanism that states are using to specify hydraulics guidance via scope outline documents for alternative delivery method projects. Additionally, other alternative project delivery methods exist. For example, Maine DOT, Ohio DOT, Minnesota DOT, North Carolina DOT, and Tennessee DOT use design-build. Tennessee DOT, Utah DOT, and Washington State DOT also use Construction Manager/ General Contractor Projects. Ohio DOT uses Value-Based Design-Build and Least-Cost Design- Build, and Washington State DOT uses Design-Bid-Build. However, hydraulics policy for the Construction Manager/General Contractor was not located, and thus this section focuses on design-build hydraulics policy.

30 Highway Hydraulic Engineering State of Practice Maine Maine DOT uses the alternative project delivery methods of “Design-Build Projects” and “Construction Manager/General Contractor Projects.” The Design-Build Project flowchart that Maine DOT uses is shown in Figure 9. For the Design-Build Projects, the process starts with issuance of a Request for Statement of Interest (RFSOI) for selection of design-builders. Then the design-builders selected are invited to submit proposals via RFP for the project. The RFPs are expected to include specified design and construction parameters, which may change. For example, Maine DOT is currently in the RFP stage for the Hampden Bridge Bundle project, which includes design and construction of eight bridges. The Hampden Bridge Bundle project is illustrated here to gain an understand- ing of how hydraulic policy is imbedded into the design-build policy. Maine DOT issues the project documents, information available, and design criteria via the Internet site. The design criteria includes survey, plan and profile design, geotechnical, utilities, right-of-way, NEPA/ environmental permits, historical investigation, U.S. Coast Guard permits, Disadvantaged Business Enterprise goals, and on-the-job training requirements. The Hampden Bridge bundle criteria is as follows: • Eight durable bridges with minimal maintenance needs; • Cross sections consisting of two 12-ft travel lanes with a 6-ft inside shoulder and a 10-ft outside shoulder within the project limits, transitioning to the existing shoulders on the approaches; • Providing a minimum 0.5% longitudinal profile grade across the structures; Figure 9. Contract procurement process flow chart. Reprinted from George Macdougall, PE, Contracts and Specifications Engineer, Maine DOT. Feb. 26, 2019.

Literature Review 31 • Locating the new bridge in the same location as the existing bridge; • Demolition and removal of the existing steel girder bridges; • In-water pier construction; • Design speed of 70 mph; • On-site maintenance of two lanes of interstate traffic in each direction at all times; • Approaches to tie as quickly as possible into the existing interstate alignment; • Net zero impact to the FEMA 500-year and 100-year floodplains; • Design speed of 55 mph for temporary works during construction; and • Cross sections consisting of two 12-ft travel lanes and two 2-ft shoulders for each direction during construction (Maine DOT 2019a, 2019b, 2019c). In the Hampden Bridge Bundle project, three of the structures span waterways. A Draft Hydrology, Hydraulics, and Scour Report is issued in the RFP Project Documents containing information for proposed structures as modeled and hydraulic requirements. Among other criteria, key personnel must contain a hydraulics/scour engineer with demonstrated experience in Maine DOT hydrologic and hydraulic analysis policies and practices (Maine DOT 2019a, 2019b, 2019c). Additional RFPs are issued for sub-proposals for other separate contract items; a specific sub- contract in this project is for automatic traffic control. A meeting is held and the question period is extended with the invited design-builders and Maine DOT. The Technical and Price Proposal Packages are submitted, Maine DOT issues a Notification of Technical Proposal Responsive- ness, and then the guaranty package is submitted by the design-builders. Maine DOT awards the contract based on price (Maine DOT 2019a, 2019b). Additionally, Maine DOT issues an $80,000 stipend to each design-builder who submits a proposal, and if the particular design-build contractor is not selected and agrees, ownership of the proposal would be transferred to the department upon agreement with the design-builder for use in the project. The design-build contractor may refuse (Maine DOT 2019b). Minnesota Minnesota DOT uses low-bid design-build and design-build best value. Like Maine DOT and Ohio DOT, Minnesota DOT provides the scope using a scope book methodology, in which different parts of the scope are contained in each book. These scope books follow the Minnesota DOT Design-Build Manual Section 4 templated outline. The scope as defined in Section 4: RFP outlines that standards, manuals, technical memorandums, standard specifications, and special provisions must be used on the project (Minnesota DOT 2017a). The drainage scope will be completed specifically by Minnesota DOT following the Minnesota DOT RFP template Book 2 Section 12: Drainage, along with hydraulic and drainage policies and procedures, and then the Minnesota DOT Design Build Program Manager will issue the RFP. The drainage scope is outlined in the Minnesota DOT RFP template Book 2 Section 12 summarized as follows: This section identifies the design and construction requirements associated with temporary and per- manent drainage, including culverts, storm sewer systems, bridge hydraulics, roadway ditches, perma- nent and temporary erosion and sediment control, structural pollution control devices, stormwater ponds, and infiltration/filtration features (Minnesota DOT 2017b). Section 12 lists in detail the complete drainage scope outline, and Section 12.1 lists the specific Minnesota DOT policies, including hydraulic and drainage policies and procedures, that must be followed. The design-build contractors are responsible for addressing all project specific items in proposals and in the design and construction (Minnesota DOT 2017a).

32 Highway Hydraulic Engineering State of Practice North Carolina The North Carolina DOT design-build program and submittal process was developed for expedited review with emphasis on project safety; ensuring environmental compliance, national, and state manuals and code requirements are met; and that RFP requirements are fulfilled. North Carolina DOT hydraulics policy is clearly defined in North Carolina DOT Design-Build Submittal Guidelines (North Carolina DOT 2009). North Carolina DOT scopes the project and may provide a culvert or bridge report as neces- sary per the project, depending on the contract with the Design-Build Team. Unless transferred to the Design-Build Team in the scope documents, FEMA compliance documents are trans- ferred upon receipt and review by North Carolina DOT. The North Carolina DOT Hydraulics Unit reviews permit application packages and the Design Build Team is responsible for designs in compliance with permits (North Carolina DOT 2009). North Carolina DOT conducts several plan reviews, including review by the North Carolina DOT Hydraulics Unit. The Design-Build Team is required to provide Release for Construction Plans (RFC), which North Carolina DOT agrees is the final plan set, with the only exception being the erosion control plans. The RFC contains the final roadway drainage design that has been accepted as final by North Carolina DOT. The RFC Erosion Control plans are submitted after the drainage design is accepted in the RFC and modifications are acceptable in the Design- Build process (North Carolina DOT 2009). Ohio Similarly, Ohio DOT has been using a design-build process in which hydraulic design is spe- cific for the project and the delivery type (Ohio DOT 2009a). Ohio DOT outlines specific project selection considerations in which a project is eligible for the design-build process. Among other considerations, the project must have a defined scope by Ohio DOT according to the Ohio DOT Design-Build Manual and Instructions for Completing the Scope of Services Form (Ohio DOT 2009a). For example, hydraulic considerations that are necessary for waterway permits must be determined prior to scope approval. For example, if a USACE Nationwide Permit is required, the permit must be obtained prior to and included in the Scope of Services provided by Ohio DOT (Ohio DOT 2009a). The drainage requirements for any project are scoped by Ohio DOT Design-Build Manual and Instructions for Completing the Scope of Services Form (Ohio DOT 2009a) by providing information per the applicable items using the following outline: 14.5 Drainage a. Address mainline, side roads and ramps. b. Retain existing system? c. Specify clean out or repairs needed. d. New open or closed drainage system? e. Raise catch basins, inlets and manholes to accommodate resurfacing or feather pavement? f. Catch basins or sodded flumes for bridge drainage? g. County Engineer flow line approvals? h. FEMA approvals? i. COE/EPA approvals? j. Additional right of way required at culverts? k. Can existing drainage conduits be reused if there is not a conflict with new construction? l. Will new headwalls be required for existing drainage conduits? m. Is there any intent to address possible drainage problems outside the toe of the embankment? n. If new underdrain runs are required, are there any ROW restraints for ditches? Would a closed system be required for outlets? Note: indicate that additional ROW acquisition is not allowed.

Literature Review 33 o. Post-construction stormwater Best Management Practices (BMP) according to Ohio DOT Location and Design Manual (Ohio DOT 2009a). In addition, hydraulic requirements are scoped by providing information as follows: a. ODOT shall determine if a flood hazard evaluation is necessary. b. Be aware that a decrease in the waterway opening must be carefully considered. In a designated Flood Insurance Area, generally a decrease in the waterway opening is not acceptable (Ohio DOT 2019a). The project is scoped by Ohio DOT for value-based design-build reconstruction projects and within the scope permits, environmental commitments, and hydraulic design of storm sewers is scoped. The Ohio DOT Construction and Material Specification Book specifies that the design- build team (DBT) provide engineering, design, and detailed construction plans for the project (Ohio DOT 2009b). The final design is the responsibility of the DBT, but Ohio DOT will identify project limits and also transfer any hydraulic or FEMA study to the DBT in the scope documents (Kahlig 2019). Typically, Ohio DOT will acquire floodplain and other associated hydraulic permits, but the FEMA floodplain risk and permit risks are assigned to the DBT on some proj- ects, e.g., if “no rise” is allowed (Kahlig 2019). An example of such scoping is the Ohio DOT I-71/I-670 Interchange project, for which the DBT is required to obtain Ohio EPA NPDES permits and is responsible for compliance (Ohio DOT 2011). Additionally, Environmental Commitments are included in the scope and are the responsi- bility of the DBT, for example: FEMA Flood Insurance Program: DBT shall submit a letter identifying any temporary or permanent impacts to the floodplain to the Department. Existing hydrology reports and information pertaining to existing sewers are included in the scope and transferred to the design-build team (Ohio DOT 2011). Tennessee Tennessee DOT uses Design-Build and CM/GC alternative contracting methods. Tennessee DOT provides a Request for Qualifications (RFQ), scopes the project, then follows the same bridge and hydraulic design policy and practice regardless of whether traditional or alterna- tive project delivery methods are used. For example in the project US 64 over the Ocoee River, Tennessee DOT provided all existing bridge plans, HEC-RAS files, and preliminary bridge layout. Engineering design and construction for Design-Build drainage are paid for in one lump sum. For Design-Build, Tennessee DOT provides scope books similar to Minnesota DOT (Tennessee DOT 2019b). In 2013, Tennessee DOT used the Construction Manager/General Contractor (CM/GC) Services Pilot Program on three projects that included at least 12 bridges on interstates. The first project allowed Tennessee DOT to obtain feedback on all project aspects, reduce risk, and improve constructability (Tennessee DOT 2019a). The proposal contract for Interstate I-240 Overhead Bridges at Norfolk Southern R/R, SR 57 (Poplar Ave. EB and WB) and Park Ave. in Shelby County, Tennessee, states that proposed structures will follow drainage requirements as specified in detail in the proposal contract (Tennessee DOT 2019c). Additionally, the proposal contract includes a required list of solutions for the CM/GC to explain certain project aspects such as management of drainage from the bridge, roadway, and during temporary construction. The CM/GC is responsible for obtaining all permits unless provided in the proposal contract (Tennessee DOT 2019c). Drainage calculations and plans must be submitted during Preliminary Engineering for approval as specified in the proposal contract at various completion milestones: concept, 30%, 60%, right of way, and 100% (Tennessee DOT 2019c).

34 Highway Hydraulic Engineering State of Practice Utah Utah DOT uses Design-Build and CM/GC alternative contracting methods. The schedules for documents are set using a Microsoft Project template. The Utah DOT Best Value Design- Build Selection Manual of Instruction (Utah DOT 2017) is similar to the Maine DOT process. Utah DOT stores all design-build templates including LOI, RFQ, and scope documents on Google Drive; the scope document for drainage is in Chapter 4 of the manual and includes a 17-page outline of detailed drainage design (Utah DOT 2019). Utah DOT construction manager-general contractor templates appear to be located in the same website but the link is under construction (Utah DOT 2019). Washington State Washington State DOT uses Design-Build, Design-Bid-Build, and CM/GC alternative contracting methods. The Design-Build Manual specifies that supplemental hydraulics and stormwater requirements be defined in RFP Technical Requirements and that hydraulic analysis must be completed for a design-build project. The manual also specifies that the templated RFP Section 2.14: Hydraulics clearly defines design and construction requirements using the Washington State DOT Highway Runoff Manual, M 31-16 and Washington State DOT Hydraulics Manual, M 23-03. The RFP is project-specific (Washington State DOT 2018).