Prioritization Procedure for Proposed Road–Rail Grade Separation Projects Along Specific Rail Corridors (2019)

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.

17 3.1 Grade Separation Project Investment Modules Factors for the Safety Evaluation Module The objective of the safety module is to compute a safety score for each crossing and quantitatively rank them. A quan- titative method was developed to prioritize grade crossings along a corridor for grade separation. The safety score is composed of two components: safety score accident prediction value site-related adjustments 1 2 k k ( ) ( )= +   The first component of the safety module, accident pre- diction value, considers the predicted accident frequency for a location. The second component, site-related adjustments, is dependent on factors related to each road–railroad cross- ing site. Each of these components can be assigned a weight by users of the railroad crossing assessment tool (RCAT) (multipliers k1 and k2) to reflect the relative importance of each component. The default value for k1 and for k2 is one. The widely accepted USDOT accident prediction formula was used to estimate the expected frequency of accidents. Also, a zero-inflated negative binomial accident prediction model was tried as an alternative model (Mathew et al. 2015). The USDOT model was selected because of its acceptance by practitioners and familiarity of the users with the model. The results from the accident prediction formula can be used to rank the crossings, but this formula merely predicts the like- lihood of a collision occurring over a given period (FHWA, USDOT 2007). This formula does not consider accident severity. The research team explored whether the site-related vari- ables identified through the literature review could contribute to the safety score. These variables were used to apply adjust- ments to the base USDOT accident prediction value for each crossing. The site-related adjustments are calculated with the follow- ing equation: w ii i i n site-related adjustment normalized value of variable 1 ∑= = where wi is the weight for each of the n variables used in the adjustment. Variables of Interest The highway–rail grade crossing variables explored in this study are shown in Table 3-1. Researchers also identified variables such as maximum train length, anticipated nearby traffic generator, and queue length, although these variables were not used in the study because of data availability issues. The research team employed an exhaustive trial approach by trying different normalization schemes and different weights for each of the variables considered. Normalization schemes (scaled values for different categories within a vari- able) were considered to identify the relative adjustments within a variable. Different weights were tried to identify the relative adjustments between variables. Tables 3-2 through 3-7 list the normalization schemes applied to the variables that were adopted from the research to the safety element of the prioritization tool. Crossing Angle The database allows three categories for the crossing variable: a. Less than 30 degrees (<30), b. Between 30 and 60 degrees (30–60), and c. Greater than 60 degrees (>60). C H A P T E R 3 Developing the Railroad Corridor Crossing Assessment Tool

18 Because of the limited number of crossings with an angle less than 30 degrees, the researchers decided to use only two values for this variable. Various normalization schemes were applied to this variable, as shown in Table 3-2. Distance to Nearby Highway Intersection This variable gives the distance of the crossing to the nearby highway intersection in feet (null if the value is less than 500 feet). This variable was categorized into four distance bands. The categorization and normalization schemes applied are listed in Table 3-3. Total Number of Tracks The variable for total number of tracks is the summation of main tracks, siding tracks, yard tracks, and industry tracks at each location. The various normalization schemes tried on this variable are listed in Table 3-4. Maximum Timetable Train Speed The variable for maximum timetable train speed gives the maximum timetable train speed at the crossing. The normal- ization schemes tried on this variable are listed in Table 3-5. Posted Highway Speed Posted highway speed was categorized and normalized with the schemes shown in Table 3-6. Crossing Surface The crossing surface variable was grouped into five catego- ries. The categories and normalization schemes tested by the research team are listed in Table 3-7. Applying Weights to Safety-Related Variables Weights were tested on each of the site-related variables to understand the impact on safety scores. The researchers tried Variable Name Description Comments Xangle Crossing angle HwynDist Distance to nearby highway intersection (feet) MainTrk Number of main tracks The sum of these four variables is used to create a new variable: total number of tracks. SidingTrk Number of siding tracks TransitTrk Number of transit tracks YardTrk Number of yard tracks MaxTtSpd Maximum timetable train speed HwySpeed Posted highway speed limit XSurfaceIDs Crossing surface Table 3-1. Variables explored in the research. Scheme Angle Schemes Angle <60 >60 <60 >60 1 0.25 0 10 1 –0.25 2 0 –0.25 11 0.9 –0.2 3 0.25 –0.25 12 0.9 –0.3 4 0.5 –0.25 13 0.9 –0.4 5 0.25 –0.5 14 0.9 –0.6 6 0.5 0.25 15 0.7 –0.2 7 0.25 0.5 16 0.7 –0.3 8 1 0.25 17 0.7 –0.4 9 1 0.1 18 0.7 –0.6 Table 3-2. Normalization schemes used for variable angle. Distance to Nearby Highway Intersection (feet) Schemes 1 2 3 4 5 6 7 8 9 10 <75 1 0 0 1 –1 1 1 1 –1 –1 75–200 0.5 –0.25 –0.5 0.5 –0.5 0.75 0.75 0.25 –0.75 –0.25 200–500 0.25 –0.5 –0.75 –0.5 0.5 0.5 –0.75 –0.25 0.75 0.25 >500 0 –1 –1 –1 1 0 –1 –1 1 1 Table 3-3. Normalization schemes used for variable distance to intersecting roadway.

19 weighted values from 0.001 to 0.1 at intervals of 0.001. The range of the weights was limited to ensure that the site-related adjustment was not exceedingly influenced by a single variable. Relative Severity of Accidents The research team calculated a severity value for each grade crossing based on the accident history and the severity of accident at each crossing. This was calculated as severity value for Xing number of fatal accidents at Xing number of injury accidents at Xing number of property damage-only accidents Xing F I P    ( ) ( ) ( ) = + + where F = relative weight for fatal accident, I = relative weight for injury accident, and P = relative weight for PDO accidents. Scale options that were tested are shown in Table 3-8 for various levels of accident severity. Three weighting scale options were explored by the research team. The method used to select weights and the normaliza- tion scheme for each site-related variable are described in Appendix B. Using Scale Option A, the research team concluded the following: • For crossings with gates, maximum timetable train speed, distance to nearby highway intersection, and crossing sur- face improved the USDOT accident prediction values; • For crossings with flashing lights, the variables maximum timetable train speed, posted highway speed, and cross- ing surface improved the USDOT accident prediction values; and • For crossings with crossbucks, the variables maximum timetable train speed, crossing angle, and crossing surface improved accident prediction values. Using Scale Option B, the research team concluded the following: • For crossings with gates, maximum timetable train speed and distance to nearby highway intersection improved the USDOT accident prediction values; Number of Tracks Schemes 1 2 3 4 5 1 0.5 0 0.3 0 0.5 2 0.75 0.5 0.6 0.75 0.85 >2 1 1 1 0.75 0.85 Table 3-4. Normalization schemes used for variable number of track. Maximum Timetable Train Speed Schemes 1 2 3 10 0.1 –4 –3 10–20 0.2 –3 –2 20–30 0.3 –2 –1 30–40 0.4 –1 0 40–50 0.5 0 1 50–60 0.6 1 2 60–70 0.7 2 3 >70 0.8 3 4 Table 3-5. Normalization schemes used for variable maximum timetable train speed. Schemes Posted Highway Speed Limit 1 2 3 ≤20 –2 –3 –1 20–30 –1 –2 0 30–40 0 –1 1 40–50 1 0 2 50–60 2 1 3 ≥60 3 2 4 Table 3-6. Normalization schemes used for variable posted highway speed limit. Crossing Surface Schemes 1 2 Unconsolidated 1 –1 Timber 0.5 –0.5 Asphalt 0 0 Rubber –0.5 0.5 Concrete –1 1 Table 3-7. Normalization schemes used for variable crossing surface. Scale Option F I P A 5 3 1 B 10 4 1 C 46.5 1.7 0.1 Table 3-8. Relative weights of accident severity used in study.

20 • For crossings with flashing lights, maximum timetable train speed and posted highway speed were the improve- ment variables identified; and • For crossings with crossbucks, maximum timetable train speed and crossing angle improved accident prediction. Using Scale Option C, the research team concluded the following: • For crossings with gates or crossbucks, maximum time- table train speed improved accident prediction and • For crossings with flashing lights, maximum timetable train speed and number of tracks improved accident prediction The research team decided to apply the weights from Scale Option A to the safety module in RCAT. Scale Option A was chosen because it identified more variables for each category of crossing warning device that could improve the USDOT accident prediction value. On the basis of Scale Option A, the variable weights and normalization scheme applied to improve accident prediction values are given in Tables 3-9 to 3-11. Safety Score Equation Used in RCAT Safety score equations used in the RCAT are shown below, with the following assumptions: 11 2k k= = Safety Score for Crossings with Gates, k1 = k2 = 1    = + + + safety score USDOT accident prediction value 0.017 normalized value for maximum timetable train speed 0.017 normalized value for distance to nearby highway intersection 0.011 normalized value for crossing surface Variable Name Categories Scheme Used Weight Applied Maximum timetable train speed ≤10 0.1 0.017 10–20 0.2 20–30 0.3 30–40 0.4 40–50 0.5 50–60 0.6 60–70 0.7 >70 0.8 Distance to nearby highway intersection <75 1 0.017 75–200 0.5 200–500 –0.5 >500 –1 Crossing surface Unconsolidated 1 0.011 Timber 0.5 Asphalt 0 Rubber –0.5 Concrete –1 Table 3-9. Corrective variables/weights for gated crossings with severity Scale A. Variable Name Categories Scheme Used Weight Applied Maximum timetable train speed ≤10 0.1 0.047 10–20 0.2 20–30 0.3 30–40 0.4 40–50 0.5 50–60 0.6 60–70 0.7 >70 0.8 Posted highway speed ≤20 –2 0.005 20–30 –1 30–40 0 40–50 1 50–60 2 Crossing surface Unconsolidated 1 0.005 Timber 0.5 Asphalt 0 Rubber –0.5 Concrete –1 Table 3-10. Corrective variables/weights for crossings with flashing lights and severity Scale A. Variable Name Categories Scheme Used Weight Applied Maximum timetable train speed ≤10 0.1 0.047 10–20 0.2 20–30 0.3 30–40 0.4 40–50 0.5 50–60 0.6 60–70 0.7 >70 0.8 Crossing angle < 60 –0.25 0.005 ≥60 0.5 Crossing surface Unconsolidated 1 0.001 Timber 0.5 Asphalt 0 Rubber –0.5 Concrete –1 Table 3-11. Corrective variables/weights for crossings with crossbucks and severity Scale A.

21 Safety Score for Crossings with Flashing Lights, k1 = k2 = 1 safety score USDOT accident prediction value 0.047 normalized value for maximum timetable train speed 0.005 normalized value for posted highway speed limit 0.005 normalized value for crossing surface    = + + + Safety Score for Crossings with Crossbucks, k1 = k2 = 1 safety score USDOT accident prediction value 0.047 normalized value for maximum timetable train speed 0.005 normalized value for crossing angle 0.005 normalized value for crossing surface    = + + + Factors for the Economic Evaluation Module The research team evaluated a broad array of quantitative and qualitative factors in developing the analysis and priori- tization method for the economic module. The economic module incorporates key factors in crossing improvement decisions based on previous research and stakeholder input. The evaluated factors included the value of time savings to commuters and commercial vehicle drivers, supply chain savings, and impacts to land use and economic development opportunities. The economic evaluation factors were split into two categories: quantitative and qualitative. Factors calculated directly from existing data sets were considered quantitative; those requiring conversion, assigned values, or derivative cal- culations were considered qualitative. Scoring Quantitative Economic Factors Quantitative economic factors were developed from BCA guidance that USDOT had given for recent competi- tive funding opportunities, namely, TIGER and Fostering Advancements in Shipping and Transportation for the Long-Term Achievement of National Efficiencies (FAST- LANE)/Infrastructure for Rebuilding America (INFRA). USDOT guidance from the Benefit–Cost Analysis Guidance for Discretionary Grant Programs provides standard values for many public benefits related to long-term outcomes from an infrastructure investment (USDOT, Office of the Secretary 2017). These standard factors include • Value of statistical life, • Value of injuries, • Value of property damage-only crashes, • Value of travel time, • Value of emissions, and • Social cost of carbon. Most standard factors, along with five additional factors listed in Table 3-12, were used in developing the economic module. Originally the research team had included crossing operat- ing cost/life-cycle costs (EC3) and construction costs (EC4) as factors with a set of standard monetized values for default values. The default values could be selected by the user if local data were unavailable. After beta testing and additional review, it was determined that these factors added unin- tended influence into the prioritization process. As a result, the research team elected to remove EC3 as an evaluation factor, and EC4 was replaced with density near the crossing as an indicator that crossing improvement would improve the mobility near the crossing. EC6, impacts to land use, was added, as well as EC-7, supply chain savings, which was used to identify the crossings impact on freight moved through the crossing on trucks. Economic Factor Determination of Factor EC1. Vehicle operating cost, passenger vehicles Calculated EC2. Vehicle operating cost, commercial vehicles Calculated EC3. Crossing operating cost/life-cycle cost Engineering estimate EC4. Construction cost Engineering estimate EC5. Economic loss: positive or negative User observations based on location of crossing Table 3-12. Economic factor determination method.

22 Economic Factor Calculation EC1.Vehicle operating cost, passenger vehicles Total fuel savings = total gals saved * average cost/gal 1. Vehicle delay cost: ((AADT w/o trucks) * (averagedelay (2.5 min)/60 min)) * average hourly traveltime cost 1 2. Operating cost: idling fuel costs: • Fuel savings from reduced idle time in minutes * AADT/60 min to convert to total hours of idling per day. • Total gal saved from idle reduction = idle hours * 1.5 gal/hr. EC2. Vehicle operating cost, commercial vehicles 3. Vehicle delay cost in travel time savings: number of commercial vehicles (AADT * % of trucks) * (average delay (2.5 min)/60 min) * average hourly truck drivercost 2 4. Operating costs defined as idling fuel costs: • Fuel savings from reduction of idling time in minutes * AADT/60 minutes to convert to total hours of idling per day. • Total gals saved by idle reduction = idle hours * 1.5 gal/hour • Total fuel savings = total gallons saved * average$/gallon EC5. Economic losses • Yes or no as observed on basis of adjacent land use. 1Average annual travel time cost was $13.00/hr. in TIGER FY 2016 guidance 2Average hourly truck driver cost was $26.68/hr. for truck drivers in TIGER FY 2016 guidance Table 3-13. Calculation of economic factors. Land Use Density Action EC4. Density near project 1 = low: rural/industrial 3 = medium suburban/medium residential density 5 = high: urban (city center, high population density) Review adjacent land use. Rank 1 low–5 high. Table 3-14. Intensity of land use density. EC6. Impacts on land use 1 = low (in industrial area) 3 = medium (suburban residential density) 5 = urban (city center, high population density) Look at adjacent land use. Rank 1 low–5 high. EC7. Supply chain savings 1 = low (less than 5% truck traffic) 2 = (6%–10% truck traffic) 3 = medium (11–15% truck traffic) 4 = (16%–25% truck traffic) 5 = high (more than 25% truck traffic) Crossing inventory trucks as percentage of AADT Table 3-15. Impacts on land use. The economic factors methodology provides a basis to compare competing projects. On the basis of feedback dur- ing testing, it was determined this model should not include cost–benefit calculations that were more fully addressed in a formal BCA. Thus, construction costs, operating and main- tenance costs, and accident costs were removed from the model. Table 3-13 shows how these factors are calculated. Scoring Qualitative Economic Factors Qualitative factors such as impact on land use, economic development opportunities, and supply chain savings must be derived from data collection activities in combination cross- ing inventory data. For example, the visual amenity factor in the environmental module estimates the intensity of devel- opment around the crossing. Typically, scoring this factor will require practitioners to use online mapping tools and a site visit. The scoring process to analyze the intensity of population density adjacent to the crossing location factor is shown in Table 3-14. These data can be used to derive inputs for esti- mating the value to the local area, with improved mobility at the crossing as a result. In this case, a value ranking can be used to indicate the value added to adjacent lands based on visual observations of the adjacent uses. Impacts are rated according to the factors shown in Table 3-15.

23 The premise in scoring the impacts to land use factor is that improvements to a crossing in an industrial area may not add much value to adjacent property, but should increase the mobility and safety of all stakeholders. However, in an urban setting the addition of a grade separation should add value to the nearby property and opportunities for economic development because of the increased mobility and safety provided by such an improvement. Supply chain savings can be evaluated qualitatively on the basis of the percentage of trucks through a crossing. Factors for the Environmental Module Environmental stewardship and consideration of environ- mental impacts in the decision-making process have a long history in transportation planning. In January 1970, the enactment of the National Environmental Policy Act required all projects utilizing federal-aid funding to undergo an envi- ronmental review to document and understand the impacts of a potential project on the environment. For the purpose of this project, the environment has been categorized into three areas: natural, built, and social. The natural environment encompasses the living flora and fauna, as well as the naturally occurring geologic features that constitute the complex ecosystem aside from what has been molded by humans. The built environment includes the communities, neighborhoods, historical artifacts, and other features of the human-shaped world. Finally, the social environment constitutes the factors that describe the living conditions and characteristics of the population that resides in the built environment. Identified factors for consideration in each subgroup follow: • Natural environment – Coastal management areas – Critical habitat for threatened and endangered species – Wetlands – Wild and scenic rivers – Air quality nonattainment areas – Superfund sites • Built environment – Tribal lands – Federal or state-owned lands – Military installations – Historical properties – Parks and recreation areas • Social environment – Low-income populations – Minority populations – Limited English-proficiency populations – Community severance Environmental Factors Scoring Methodology Qualifying the value of an environmental factor can be difficult and contentious, in part because differing elements of our society place different levels of importance upon the natural, built, and social environments. For example, the value that an individual agency places on the critical habitat of an endangered species may differ from the societal value of a historical property or the potential severance of a community. To remove the ambiguity of the social value of an environ- mental good, it is proposed that a simple toggle approach be utilized for each of the above factors, that is, a 0 or 1. Before environmental scoring for grade separations in a corridor is begun, it is suggested that the corridor be mapped and the location of known environmental variables adjacent to the project locations be identified. A ½-mile buffer area is used for identifying adjacent environmental variables to pro- vide a local connection to the potential project. The presence of a sensitive environmental factor inside the project buffer will be scored as a 1 in the spreadsheet model. If no environ- mental factor is located inside the project buffer, a 0 score is assigned. On the basis of the 15 environmental factors, the following scores may be assigned: • 0 to 3 factors affected: 1 (low) ranking, • 4 to 6 factors affected: 2 ranking, • 7 to 9 factors affected: 3 (moderate) ranking, • 10 to 12 factors affected: 4 ranking, and • 13 factors affected: 5 (high) ranking. Each of these factors is discussed in more detail in the following sections. Natural Environment Coastal management areas. Coastal management area (CMA) locations were established under the Coastal Zone Management Act of 1972 (CZMA). These locations have been established to preserve, protect, develop, and, where possible, to restore or to enhance the resources of the nation’s coastal zone. Zones begin at the shoreline and extend inland to locations that have direct and significant impact on coastal waters. Additional lands held by the federal government are included in these zones unless otherwise noted. CMA loca- tion information can be found here: https://data.noaa.gov/ dataset/coastal-zone-management-act-boundary-for-the- united-states-and-us-territories-as-of-decemb-2013. Corridors

24 that may impact a CMA are assigned a score of 1. Corridors outside a CMA are assigned a score of 0. Critical habitat for threatened and endangered species. Threatened and endangered species vary from location to location. Reference lists for each project location should be developed with natural resource agencies for each state/ location. Locations of critical habitat can be identified and be mapped for use in the buffer analysis. The presence of critical habitat is assigned a score of 1. Areas lacking critical habitat are assigned a score of 0. Wetlands. Typically, a project map or buffer map will show the project location in relation to known wetlands. In states/locations where wetland banking is allowed, an assessment of the impact/cost can be compared to the project cost to develop an impact/cost ratio. Another, less quanti- tative assessment can assign a no impact or impact scoring system. For this example, it is proposed that the presence of wetlands inside the buffer area would yield a score of 1 and the absence of a wetland inside the buffer yield a 0. Wild and scenic rivers. Wild and scenic rivers are federally designated pristine habitat designated by the Wild and Scenic Rivers Act of 1968. These waterways are oriented to preserve the natural, cultural, and recreation value of the riverine system for future generations. The National Park Service can be consulted to identify wild and scenic rivers via https://www.nps.gov/ncrc/programs/rtca/nri/index.html. The project buffer will be used to identify locations that might impact wild and scenic rivers. For this criterion, the presence of a wild, scenic, or recreational river inside the buffer is assigned a score of 1, and the absence is assigned a 0. Air quality nonattainment areas. Corridors located inside areas that are not in attainment of the National Ambient Air Quality Standards may require additional analysis and/or other project activities or amenities to be included during the development of a grade separation project. The U.S. Environ- mental Protection Agency (EPA) maintains the listing of areas currently designated as nonattainment for air quality standards on their website: https://www3.epa.gov/airquality/greenbook/. Project locations should be identified within these areas and scored a 1 for a location inside a non attainment area and a 0 for areas in attainment of national air quality standards. Superfund sites. Areas with substantial pollution, requir- ing large-scale cleanup activities, have been identified by EPA as focus areas for a visible and lasting difference in commu- nities being made in response to environmental emergencies, oil spills, and other natural disasters. The location of superfund sites can be found here: https://www.epa.gov/superfund/ search-superfund-sites-where-you-live. Project locations should be identified within these areas and assigned a 1 for a location inside a superfund cleanup site and a 0 for areas that are not inside a superfund site. Built Environment Tribal lands. The lands controlled by the indigenous tribes inside the United States require additional care and coordination for projects located within their boundaries. The U.S. Department of the Interior’s Bureau of Indian Affairs works with the recognized tribes to administer tribal property. Additional information on tribal lands can be obtained here: http://www.bia.gov/yourland/index.htm. Project locations identified within these areas and impacting tribal lands are scored 1 and are scored 0 for separation loca- tions that do not impact tribal lands. Federal or state-owned lands. The relevant state DOTs, in partnership with the local Federal Highway Division Office, will be able to determine the location of state or federally owned land within the project area. As these locations require additional consultation and may be protected, their presence should be noted during the initial screening process. Locations should be identified within the ½-mile project buffer and scored 1 for locations with federal or state-owned lands and 0 for project locations without federal or state-owned lands. Military installations. Similar to federal or state-owned lands, military installations provide an additional layer of com- plexity. In addition to being federally or state-owned, the site may be a historical property or contain locations with unexploded ordnance. Military installations can be found here: https:// www.nps.gov/nagpra/DOCUMENTS/BasesMilitary MAP.htm. Locations should be identified within the ½-mile project buffer and scored 1 for locations adjacent to military installations and 0 for project locations without military installations. Historical properties. The National Park Service main- tains the National Register of Historic Places. Historical prop- erties provide societal benefits that are difficult to quantify and cannot be replaced. Project corridors should be overlaid with locations of historic places from the register (available here: https://www.nps.gov/nr/research/index.htm). Records have not been digitized for Arkansas, Illinois, Massachusetts, Michigan, Missouri, North Carolina, New York, Pennsylvania, Ohio, Texas, and Virginia. It is necessary to consult with a state’s historic preservation officer to iden- tify locations that may not be available through the national register because of their restricted status. The presence of historical properties inside the corridor buffer should be scored 1, whereas the absence should be scored as 0. Parks and recreation areas. Other locations that can be used by the public as parks or recreation areas should

25 also be identified during the mapping and corridor buffer process. These locations may be protected under Section 4f of the Department of Transportation Act of 1966. As spe- cific impacts are not readily available to be assessed in the preliminary screening for projects, the location of parks and recreation areas should be noted as a 1 in the Excel portal. The absence of parks and/or recreation areas within the project buffer should be noted as 0. Social Environment Low-income populations. Executive Order 12898 protects low-income populations from disproportionately high and adverse human health and environmental effects during the development of federal-aid transportation projects and pro- grams. Low-income populations can be mapped according to census geography. For each location, areas with high low-income population should be identified within the project buffer. Pres- ence of a high ratio of low-income population within the project buffer are assigned a 1, whereas the absence is assigned a 0. Minority populations. Executive Order 12898 also pro- tects minority populations from disproportionately high and adverse human health and environmental effects during the development of federal-aid transportation projects and pro- grams. Minority populations can be mapped according to cen- sus geography. For each location, areas with a high minority population should be identified within the project buffer. Pres- ence of a high ratio of minority population within the project buffer are assigned a 1, whereas the absence is assigned a 0. Limited English proficiency. Executive Order 13166 pro- tects populations with limited English proficiency (LEP) by requiring federal agencies to ensure that their services are avail- able to those with LEP. For transportation projects this means projects impacting LEP populations must undergo an effort to ensure programs and activities provided in English are acces- sible to LEP persons. LEP population locations can be identified via the U.S. Census geography and mapped for review within the project buffer for county and state using the American Community Survey. Additionally, DOT density maps are available via the 2011 Language Mapper (https://www.census. gov/hhes/socdemo/language/data/language_map.html). The presence of LEP populations is assigned a score of 1; the absence of LEP populations is assigned a 0. Community severance. An assessment of the cultural communities that exist within a jurisdiction should be under- taken to identify when potential grade separation decisions may close existing crossings that provide access to essential goods, services, or community locations such as churches and other gathering places. A visual assessment may be conducted on the basis of demographic and social information available through the U.S. Census Bureau and online mapping services such as Google Earth. The potential for community severance should be scored as a 1. If a community is not severed because of the proposed project action, it is assigned a score of 0. Factors for the Community Livability Module Livability is a concept that while not new, is increasingly being integrated into the transportation planning process. In recent years, FHWA has started a Livability Initiative— http://www.fhwa.dot.gov/livability/. In partnership with the U.S. Department of Housing and Urban Development (HUD), DOT, and EPA, FHWA also sponsors an interagency Partner- ship for Sustainable Communities: https://www.sustainable communities.gov/. FHWA defines livability as tying the quality and location of transportation facilities to broader opportunities such as access to good jobs, affordable housing, quality schools, and safer streets and roads. FHWA supports livable communities through funding transportation-related projects and spon- soring activities such as context-sensitive solutions and public involvement that help enable people to live closer to jobs, save households time and money, and reduce pollution(FHWA: https://www.fhwa.dot.gov/livability/). Under this definition, livability factors to consider for grade separations would be those factors related to safety, time sav- ings, access to work, and emissions. In a research paper pro- duced for FHWA about the role of FHWA programs in livability (The Role of FHWA Programs in Livability: State of the Practice Summary), other definitions of livability were also offered (FHWA, USDOT 2011a). Some livability definitions focused on the ability to access goods and services without having to use motorized transport. More broadly, several organizations define livability as part of a community’s overall character. The research team reviewed livability definitions from AASHTO (2010b), the American Institute of Architects (2005), and the American Association for Retired Persons (AARP 2005). Potential Community Livability Variables One thing seems certain from the review of literature: the term livable community is evolving. Generally, the phrase livable community includes the principles of safety and secu- rity, communities that encourage citizens to be engaged, and, from a strictly transportation standpoint, communities that do not require motorized transport to access goods and services.

26 The term livable community is also often used as an ancillary means for describing quality of life. Within the other three evaluation categories addressed by the research—safety, economics and environment—many of the elements that contribute to livability—for example, safety and access to jobs—are already covered. For scoring grade crossings, the livability evaluation criteria were viewed as factors that affected the security of citizenry, without being directly involved in a train accident/collision. The following criteria became the focus of potential scoring factors. Risk of Derailment/Release of Hazardous Materials “Since 1980 the U.S. railroad derailment rate has declined from 8.98 derailments per million train miles, to 1.63 in 2014, an 82% reduction” (Liu, Saat, and Barkan 2017, p. 2). While derailment statistics have dramatically improved over the past several decades, a single derailment involving the release of hazardous materials (hazmat) can have tragic consequences, as was the case in Lac-Mégantic, Quebec, in 2013. While the Lac-Mégantic tragedy was not related to an at-grade road– railroad crossing, recent research conducted by the Univer- sity of Illinois at Urbana–Champaign Rail Transportation and Engineering Center (RailTEC) found “that 22.6% of all grade crossing incidents involved a train carrying hazardous materials” (Chadwick et al. 2013, p. 13). The research also noted that grade crossing incidents involving a hazmat release are rare. Thirty release events involving at-grade crossing incidents occurred over a 20-year period. Statistics suggest that the likelihood of a hazmat release is highly dependent on a derailment event: the likelihood of release also appears dependent on derailment, as 0.641 percent of all derailments result in release and 0.063 percent of all nonderailments result in release (Chadwick et al. 2013, p. 14). The safety evaluation category addresses primarily crashes and incidents between a train and an automobile, bike, or pedestrian. “Traditionally, highway departments prioritize upgrading grade crossings with the highest risk of an accident, but do not account for the likelihood of a train derailment” (Chadwick et al. 2013, p. 3). The safety evaluation category does not consider the impact on citizens resulting from train- related incidents at crossings that result in explosions or the release of hazardous substances. While a predictive model was still being formulated, the RailTEC researchers identified several conditions that con- tributed to train derailments: (1) track class, (2) method of operation (signaled or non-signaled) and (3) traffic density. The research conducted by RailTEC has the following con- clusions: (Liu, Saat, and Barkan 2015, p. 9): • The higher the FRA track class is, the lower the train derail- ment rate. • Signaled track has a lower derailment rate than nonsignaled track. • Track with higher density has a lower derailment rate. The authors noted that the analysis focused only on derail- ments, excluding other types of train accidents such as grade crossing incidents (Liu, Saat, and Barkan 2017, p. 3). The research by Chadwick, Saat, Dick, and Barkan, which focused specifically on derailment occurrence at grade crossings, identified three variables with predictive value for identifying grade crossings and derailment risk, as well as easily obtain- able proxy variables (Chadwick et al. 2013). The variables are shown in Table 3-16. While the research suggests conditions that contribute to train derailments at grade crossings, other research has examined the potential impact of a derailment on the safety and security of resident populations near a crossing if hazmat is present: a. Population density. In 2014, the Minnesota Legislature directed the Minnesota DOT to study the issue of at-grade crossings and the movement of crude by rail through Minnesota: The study focuses on prioritizing risks, while also reducing potential collisions by improving the overall safety of each grade crossing. . . . This study is different because it expands the conventional evaluation scope to include the risk to adjacent residents and workers. The study shifts the focus to an area and population-based risk assessment, rather than just an accident prediction assessment” (Minnesota DOT 2014, p. 9). The Minnesota DOT used satellite imagery and a geographic information system (GIS) to delineate buffer zones around at-grade crossings to examine potential populations endangered by a hazmat release or explosion. b. Vulnerable populations. The MnDOT effort also made an attempt to identify “fixed vulnerable populations such as hospitals, nursing homes, prisons, and transient vulnerable populations, such as schools.” (Minnesota DOT 2014, p. 10). The presence of hospitals, senior care facilities, schools, and prisons can also be identified from GIS applications. Emergency Response Delays Safety evaluations of at-grade crossings typically do not address the impact of trains blocking a crossing and a timely response by emergency vehicles responding to a crisis. Recently, the City of Chicago identified a series of grade Incident-Specific Variable Crossing-Specific Variable Vehicle speed Highway speed limit Train speed Posted timetable speed Large vehicle involvement Percentage of truck traffic AADT Table 3-16. Derailment/crossing variable.

27 crossings frequently used by emergency vehicles, which have been identified as 911 critical grade crossings. In 2005, the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU) directed FRA to examine and report the impacts of blocked highways due to at-grade crossing on emergency response providers. The study found no national data collection efforts to identify or quantify blocked crossings or emergency response delays. However, the study identified several factors that contributed to traffic delays at-grade crossings: (1) moving trains and train length, (2) stopped trains, and (3) operational problems. The FRA report concluded: The impacts on communities from delayed response due to blocked crossings, while sometimes severe, are less than the impacts of traffic delays and congestion caused by blocked crossings. Another way to look at it would be to say that in places where blocked crossings are seen as a problem—to traffic, to safety and to emergency response—emergency response delays may help to justify a grade separation or other major expenditure, but such delays are unlikely, by themselves, to justify major remediation measures except in special cases. (FRA 2006, p. 15) Since 2001, the National Fire Protection Association has published NFPA Standard 1710 (standard for the organization and deployment of fire suppression operations, emergency medical operations, and special operations to the public by career fire departments), which establishes a baseline of emer- gency response units across the United States and measures their performance. In general, the gold standard for emergency response is that the first unit should arrive within 8 minutes 90 percent of the time (NFPA 2010). Scoring Community Livability Factors Based on the review of literature related to livability, Table 3-17 presents the preliminary scoring methodology for inclusion in the community livability module. 3.2 Defining Rail Corridors One of the questions to be answered for completion of the proposed corridor-based tool was, What constitutes a rail corridor? Data Source Comments 5– 65+ mph State DOT or local road authority 65+ mph = Score 10; 55–60 mph = Score 8; 45–50 mph = Score 6; 35–40 mph = Score 4; <35 mph = Score 2 Train speed 10– 80+ mph FRA Train speed, which is a FRA inventory data field. The following maximum speeds are for freight trains: 80 mph = Score 10 60 mph = Score 8 40 mph = Score 6 25 mph = Score 4 10 mph = Score 2 Large vehicle exposure Percentage of truck traffic State DOT or local road authority 20% or more = Score 10 15–19% = Score 8 10–14% = Score 6 5–9% = Score 4 <5% = Score 2 Exposure to Hazmat from Incident Presence of hazardous train cars Yes = 1 No = 0 Population density Low =1 Medium = 3 High = 5 GIS Population per square mile, measured within 1/2 square mile of crossing Vulnerable populations Yes = 1 No = 0 GIS Hospitals, senior living, schools, prisons; measured within 1/2 square mile of crossing Emergency response delays Yes = 1 No = 0 Emergency medical service Presence of police station, fire station, or hospital within 1/2 mile of crossing without access to a grade- separated crossing; measured within 1/2 mile of the crossing. Risk of Derailment Highway vehicle speed speed limit) (proxy variable = posted Factor Data Value Railroads report hazmat train volumes exceeding 1 million gal. to state emergency management officials. Hazmat cars/tankers are placarded to identify hazardous contents. Observation may be required to collect and analyze hazmat volumes. Railroads/ observation Table 3-17. Scoring community livability factors.

28 Class I railroads often refer to major mainline corridors that stretch for thousands of miles between coastal gateway ports and major inland cities. For the application of a grade separation evaluation tool, a more disaggregated definition of corridor that could be applied at the state or local level was desired. Previous TRB research defines a corridor: Broadly defined, a corridor generally refers to a geographic area that accommodates travel or potential travel. Normally, a cor- ridor is considered to be a “travel shed,” an area where trips tend to cluster in a general linear pattern, with feeder routes linking to trunk lines that carry longer distance trips in a metropolitan area. (Smith 1999) While the definition above discusses a corridor in a highway context as opposed to a rail corridor, this definition remains applicable to this research in many ways. Other research reviewed from FRA and Europe defined corridors as rail line connections between major populations centers, but again these distances are often relatively long. From an operations standpoint, a corridor is frequently defined as the segment of rail tracks between two rail junc- tions. A rail junction is the place where two or more rail routes converge or diverge. From a planning standpoint, previous research suggests that corridors are most often defined by geopolitical boundaries of state or local governments. On the basis of the online research, the research team’s initial suggestion is to allow users to define a corridor as having a minimum number of crossings and, on the basis of junctions or geopolitical boundaries, offering the most flex- ibility and likely the most utility for the target audience. This project suggests the minimum number of crossings within a corridor will be three. All crossings should be within a single railroad corridor, as defined by primary line ownership, for example, mainline, subdivision, or corridor. Users wishing to prioritize grade crossing within a corridor may place end- points on the corridor for analytical purposes on the basis of geopolitical boundaries. This definition discourages broad corridor definitions that might include tracks operated by multiple rail carriers in parallel alignments as approaching network analyses. 3.3 Identifying Candidate Corridors/ Corridor Profiles To ensure that the tool developed from this research would be robust and meet the expectations of potential users, the spreadsheet application was tested across a variety of grade crossing types in both rural and urban settings. The first step was to select a sample of potential case studies across urban and rural rail corridors faced with a variety of rail issues and needs. The solicitation for beta test corridors was conducted through several means: • The research team prepared a one-page flyer about the research project that was distributed during the 2016 Annual TRB Meeting at the Standing Committee on Highway/Rail Grade Crossings and the winter meeting of the AASHTO Standing Committee on Rail. • As part of Task 2, a link to a web-based survey was dis- tributed to 240 stakeholders at state DOTs, metropoli- tan planning organizations, cities, counties, and other agencies with responsibilities for road/rail grade sepa- ration decisions. Forty-seven stakeholders at least par- tially completed the survey. Once the survey application closed, respondents were sent an e-mail thanking them for their participation and input on potential beta test corridors. • In addition, FRA was asked for input regarding the beta test corridor. • The research team examined opportunities from other rail planning studies being conducted concurrently to identify viable corridor candidates. Each of the outreach efforts above resulted in at least one suggestion for beta test corridors, accounting for a preliminary list of 10 potential rail corridors: 1. Staples Subdivision, Minnesota; 2. Knowledge Corridor, Connecticut and Massachusetts; 3. All Aboard Florida High-Speed Rail Project; 4. California High-Speed Rail Project; 5. Great Northern Corridor, Seattle Subdivision in Kent, Washington; 6. Great Northern Corridor, Glasston Subdivision, North Dakota; 7. Terminal Railroad Association of Saint Louis (TRRA), Illinois Transfer Subdivision; 8. TRRA, Merchants Subdivision; 9. North Carolina Corridor from Raleigh to Charlotte; and 10. Chicago-to-Saint Louis Corridor. The research team eliminated five corridors from priori- tization procedure development because of a cursory review that focused on applicability to the project objectives. In several cases, nominated corridors had undergone signifi- cant investments, including grade separations to upgrade the corridor for higher-speed passenger service during the past several years. In other cases, corridors were eliminated from further consideration because of potential political implica- tions or overt complexity.

29 After the culling process, the following five corridors were advanced to the project panel as potentials for case study beta test candidates: 1. Great Northern Corridor, Seattle Subdivision in Kent; 2. Great Northern Corridor, Glasston Subdivision, North Dakota; 3. TRRA, Illinois Transfer Subdivision; 4. TRRA, Merchants Subdivision; and 5. North Carolina Corridor from Raleigh to Charlotte. Profiles of the five candidate corridors were presented to the NCHRP panel during a midproject meeting. The panel suggested using the Seattle Subdivision in the State of Washington, but extending the corridor as proposed from Seattle, Washington, to Olympia, Washington, to allow both rural and urban environments to be considered during the beta test. The panel also suggested using one of the TRR subdivisions in the Saint Louis area, with the final selection left to the research team. After officials in the Saint Louis area were consulted, it was decided the Merchants Subdivision would be used as a beta test candidate because it offered the opportunity to examine a corridor that crossed state lines (Illinois and Missouri). Corridor Profile: Seattle Subdivision in Kent The Kent Section of the Seattle Subdivision corridor is approximately 4.5 miles long. The section contains 12 rail crossings within the City of Kent. Of the 12 rail crossings, four have already been grade separated. This corridor can- didate is on the northern end of BNSF’s Seattle Subdivision on the Seattle–Vancouver, Washington, Branch of the BNSF. The northern end point of this corridor is located at Milepost 13.19 of the Seattle Subdivision. The City of Kent is to the east of the I-5 Corridor sitting about halfway between Seattle and Tacoma. Figure 3-1 shows the location of the City of Kent in relation to Seattle. As the map indicates, the rail lines and the highway corridor (I-5 and SR 167) all run parallel to each other. This factor poses a mobility challenge as both com- muters and freight move east–west across the rail corridor to access I-5 and SR 167. Many distribution centers supporting the Ports of Seattle and Tacoma are located in the Kent Valley. The volume of freight traffic generated by the distribution centers combined with this suburban city’s residential commuters present an east–west mobility challenge across the rail corridor. In recent years, three grade crossings (James, SR 516 Willis, and South 212th) have been considered for grade separation Kent Figure 3-1. Map of Kent, Washington. Source: Google.

30 to help mitigate this mobility challenge. These three crossv- ings have some of the highest exposure rates within the BNSF Seattle Subdivision. • The corridor is defined as the BNSF Mainline, Seattle Subdivision, Seattle–Vancouver Branch within the City of Kent. • There are 12 crossings; endpoints: Milepost 13.19 to Mile- post 17.55. • Uses of the corridor are – Freight, – Passenger trains: Amtrak long-distance and Amtrak Cascades, and – Commuter rail: Sound Transit from Lakewood, Washington, to Seattle. Twelve public crossings are located on this portion of the BNSF branch within the City of Kent, as shown in Table 3-18. Four of the crossings are already separated as noted in the position column. Jurisdictional Responsibility for Crossing Improvements The City of Kent led the grade separation discussions. Because one of the crossings is State Highway SR-516, Washington DOT would have to be involved in the final improvement decision and construction activities. Ownership and Physical Description • Rail line ownership and trackage agreements – This mainline is owned by BNSF Railway, which has trackage agreements with Amtrak, Amtrak Cascades (Washington DOT), and Sound Transit’s Commuter Rail. • Physical description of rail assets – Line capacity. 100 trains per day until passenger improvements are completed in 2017–2018, then the capacity jumps to 120 trains per day (WPPA 2011 Cargo Forecast). – Weight capacity. Mainline at 286,000. – Terminals. There are no rail terminals on this section of the branch line. There is a Sound Transit Rail Commuter Station located at 301 Railroad Avenue North, Kent. – Quiet zones. There are no quiet zones within this portion of the branch line. Corridor Activity • Total activity. Currently, average 76 trains per day. Since 2006, the FRA has shown train traffic varying from 57 to 66 trains per day on the Seattle Subdivision running from Vancouver, Washington, to Seattle, Washington, which does not include the Sound Transit trains running only along the subdivision from Seattle to Tacoma. • Freight activity. 52 trains per day. • Passenger activity: – Amtrak long-distance. One round-trip train per day (total of 2 trains per day). – Amtrak Cascades has four round-trip trains per day through this corridor (total of eight trains per day). – Sound Transit runs seven round-trip commuter trains (total of 14 trains per weekday) through this corridor during the morning and evening commuter windows each weekday. • Safety history – Two of the 12 crossings within the Kent BNSF rail corridor ranked 4th and 5th most dangerous out of 915 BNSF crossings in Washington State. James Street (ranked 4th) has reported accidents in 2011, 2013, and 2014. Crossing ID Position Status Milepost County City Street 1 077831A RR under Open 0013.190 King Kent South 196th Street 2 085625H At grade Open 0014.185 King Kent South 212th Street 3 085627W RR under Open 0015.187 King Kent South 228th Street 4 085628D RR under Open 0015.375 King Kent SR 167 5 085629K At grade Open 0015.942 King Kent James Street 6 085633A At grade Open 0016.179 King Kent Smith Street 7 085636V At grade Open 0016.289 King Kent Meeker Street 8 085637C At grade Open 0016.339 King Kent Gowe Street 9 085639R At grade Open 0016.436 King Kent Titus Street 10 085640K At grade Open 0016.557 King Kent SR516 Willis Street 11 085642Y At grade Open 0017.090 King Kent South 259th Street 12 085643F RR over Open 0017.557 King Kent South 266th Street Note: RR = railroad. Table 3-18. Public crossings in city of Kent on BNSF Seattle subdivision.

31 Special Considerations/Other • What makes this corridor a good candidate (e.g., recent increases in hazardous cargo movements or total train volumes)? Increased train volumes included oil trains and coal trains to northwest Washington and British Columbia. • New development conflicts in the corridor. The land use/ density of development of industrial, commercial, and resi- dential continues to increase along and near this corridor; this change causes increased AADTs at these crossings. In 2017–2018, Amtrak Cascades will be adding additional intercity trains to this corridor. Corridor Profile: Merchants Subdivision from Saint Louis to Granite City, Illinois The Merchants Subdivision is approximately 9.7 miles, starting from Saint Louis to Granite City, Illinois. There are 49 crossings along this corridor. Of the 49 grade crossings, 22 are at-grade crossings, while the rest are grade-separated crossings. This corridor crosses the Mississippi River at the Merchants Bridge, which is a ½-mile-long railroad-only bridge. TRRA owns all lines on this corridor. TRRA is owned by several Class I railroad companies. Currently no grade cross- ings on this corridor are considered for grade separation. The location of the Merchants Subdivision is shown in Figure 3-2. • The corridor is defined as the Merchants Subdivision of TRRA. • Endpoints: Milepost 0.22 to Milepost 9.95. • Uses of the corridor: – Freight and – Passenger (temporarily because of renovations on the MacArthur Bridge). • Number of crossings (49): – 22 crossings at grade, – 17 railroad over crossings, – 9 railroad under crossings, and – 1 crossing on Cerre Street marked private. – Jurisdictional responsibility for crossing improvements ISS and the Missouri DOT have the jurisdictional respon- sibility for crossing improvements along these lines. Ownership and Physical Description • The Merchants Subdivision line is owned by TRRA. • Physical description of rail assets: – Line capacity: not determined. – Weight capacity: mainline, 286,000 lbs. Figure 3-2. Merchants Subdivision location map.

32 Crossing ID Position of Crossing Highway Milepost Total Trains 803243W Railroad under Compton Avenue 0.22 18 803244D Railroad under Jefferson Avenue 0.82 18 803248F Railroad under 18th Street 1.37 41 803250G Railroad under 14th Street 1.65 41 424810P Railroad under Tucker Boulevard 1.8 0 Unknown Private Cerre Street 1.89 0 803264P Railroad over 7th Street 2.16 1 803265W Railroad over 6th Street 2.22 1 803266D Railroad over South Broadway 2.29 1 803267K Railroad over 4th Street and Cedar 2.35 1 803268S Railroad over 3rd Street 2.41 1 803269Y Railroad over 2nd Street 2.5 1 803270T Railroad over Plum and 1st 2.6 1 803271A Railroad under Poplar Street Bridge 2.61 1 803272G Railroad over Poplar Street 2.64 1 803273N Railroad over Washington Avenue 3.33 1 803274V Railroad under Eads Bridge 3.34 1 803276J Railroad over Lucas Avenue 3.38 1 803277R Railroad over Morgan Street 3.43 1 803280Y Railroad over Laclede's Landing Boulevard 4.49 1 803279E Railroad under Martin Luther King Bridge 3.49 1 803282M Railroad over Carr Street 3.66 0 803368W At grade Biddle Street 4.13 62 803366H At grade O'Fallon Street 3.83 62 803365B At grade Florida Street 4.13 62 803364U At grade Mullanphy Street 4.19 62 803362F At grade Brooklyn Street 4.39 62 803360S At grade Tyler Street 4.6 62 803359X At grade Chambers Street 4.65 62 803356C At grade Clinton Street 4.77 74 803354N At grade North Market Street 4.9 62 803353G At grade Branch Street 5.31 57 803078N At grade Hall and Dock 5.41 0 803077G At grade Hall and Buchanan 5.51 0 803350L At grade Angelrodt Street 5.58 84 803349S At grade Destrehan Street 5.66 94 803074L Railroad under Salisbury and Hall 5.8 14 803348K At grade Bremen Avenue 5.96 34 803347D At grade Angelica Street 6.23 50 803073E Railroad over Angelica Street 6.23 14 803346W At grade Ferry Street 6.45 38 803072X Railroad over Ferry S Street 6.45 1 803193V Railroad over Merchant's Bridge 7.61 0 803192N Railroad over Klein Street 7.82 0 803091C At grade Washington Avenue 7.9 10 803092J At grade Edwardsville Road 1.32 25 803090V At grade Market Street 8.32 72 803189F At grade Commonwealth Drive 9.77 0 803085Y At grade Niedringhaus Avenue 9.95 52 at Route 3 Table 3-19. Total trains using each crossing along Merchants Subdivision, TRRA. – Terminals: � Saint Louis passenger terminal (located at Milepost 1.77), � TRRA Madison Yard (located at Madison, Illinois), and � TRRA Carworks Yard (located at Madison). – Quiet zones. A 24-hour whistle ban exists on the cross- ings in Missouri, with no whistle ban on the crossings in Illinois. Corridor Activity Passenger trains are using this corridor because of the current renovations on the MacArthur Bridge. Approximately three passenger trains use the Merchants Corridor daily. If not for the renovation, Merchants Subdivision would have no Amtrak or other passenger train usage. The total number of trains using each crossing along this corridor is listed in Table 3-19.

33 Safety History Table 3-20 lists crossings that have experienced a highway/ rail accident at the grade crossings along the Merchants Sub- division in the past 10 years between the years 2006 and 2015. Special Considerations/Other The Merchants Subdivision covers the Merchants Bridge, which spans the Mississippi River. This bridge currently has clearance and weight restrictions. Two trains cannot cross the bridge simultaneously, and hence freight trains frequently come to a stop on or near an approach grade. The proposed Merchants Bridge replacement project includes removal and replacement of the three span trusses, seismic retrofit- ting of the existing river piers, and improvement to the east approach. An estimated 185,676 truckloads could be diverted from the highway to rail; thus, longer/more frequent freight trains could be expected along this corridor (Saint Louis Regional Freightway 2016). Crossing ID Highway–Rail Grade Crossing Accidents Position of Crossing 803353G 1 At grade 803090V 1 At grade Table 3-20. Grade crossing accident history 2006–2015 on Merchants Subdivision.