National Academies Press: OpenBook

Improving Pedestrian and Motorist Safety Along Light Rail Alignments (2009)

Chapter: Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD

« Previous: Chapter 2 - State of the Practice Methodology and Summary
Page 21
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 21
Page 22
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 22
Page 23
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 23
Page 24
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 24
Page 25
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 25
Page 26
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 26
Page 27
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 27
Page 28
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 28
Page 29
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 29
Page 30
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 30
Page 31
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 31
Page 32
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 32
Page 33
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 33
Page 34
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 34
Page 35
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 35
Page 36
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 36
Page 37
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 37
Page 38
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 38
Page 39
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 39
Page 40
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 40
Page 41
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 41
Page 42
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 42
Page 43
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 43
Page 44
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 44
Page 45
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 45
Page 46
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 46
Page 47
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 47
Page 48
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 48
Page 49
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 49
Page 50
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 50
Page 51
Suggested Citation:"Chapter 3 - LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD." National Academies of Sciences, Engineering, and Medicine. 2009. Improving Pedestrian and Motorist Safety Along Light Rail Alignments. Washington, DC: The National Academies Press. doi: 10.17226/14327.
×
Page 51

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.

21 This chapter presents a review of LRT collision data obtained from transit agencies across the United States. The purpose of this chapter is to identify the purpose of collecting and storing collision data at each level of government, identify the actual data collected at each level, assess the quality of the data col- lected and identify any potential issues or deficiencies, and, where possible, analyze the data to determine root causes and factors contributing to LRT collisions. The Data Collection and Transfer between FTA/NTD, SSO, and Local Transit Agencies section outlines the relationship between the three hierarchical levels of transit administration, and identifies the datasets requested and received by the proj- ect team from various transit agencies. The National Transit Database, SSO Agencies, and Local Transit Agencies sections examine the collision data available at the federal, state, and local transit agency levels, respectively. The Comparison of Databases section concludes with a comparison of the data available across the three levels of transit administration. Data Collection and Transfer between FTA/NTD, SSO, and Local Transit Agencies Incident report data are collected and stored at three distinct hierarchical levels of transit administration in the United States: the federal level, state level, and local level. This section briefly outlines how transit incident data are stored and transferred between the three levels of transit administration. The purpose and method of incident data collection at each level of transit administration are presented in Chapter 7 and Appendix E. The primary source of incident data is the information col- lected at the scene of the incident using collision report forms. According to Rule 49 CFR Part 659 of the Federal Transit Administration Act, it is the responsibility of the SSO agency to “investigate, or cause to be investigated . . . any incident involving a rail transit vehicle or taking place on rail transit- controlled property” meeting certain notification require- ments (see the Data Collected by NTD section) (4). Based on discussions with SSO and local transit agencies, it is usually the local transit agency that conducts the incident investiga- tion at the scene. When this is the case, the local transit agency is required to incorporate practices specified and approved by the oversight agency into the investigation procedure used by the local transit agency. In addition, the local transit agency is required to transmit a final investigation report that identifies any contributing or casual factors and outlines a corrective action plan. The SSO agency is required to review the findings of the investigation report, and either adopt it or formally transmit its dissent to the report findings. When the SSO agency disagrees with the findings of the investigation report, the SSO agency may conduct its own independent investigation of the incident. In the case where the SSO agency does not authorize the local transit agency to conduct incident investigations on its behalf, but instead chooses to investigate the incident directly, the SSO agency is responsible for compiling the final investigation report and transmitting it to the local transit agency. The exact method- ology employed to meet the above requirements is left largely to the discretion of the SSO and local transit agencies, and may differ between jurisdictions. The SSO agency is respon- sible to compile and submit to the FTA an annual report summarizing its oversight activities, including a description of the contributing/casual factors of investigated collisions, and the status of any corrective actions. In addition to meeting its obligation to the SSO agency, the local transit agency is also required to investigate all inci- dents meeting the NTD criteria of a reportable incident out- lined in the Data Collected by NTD section. The local transit agency is responsible for completing and submitting the S&S-40 Reportable Incident Report form (S&S-40 form) for each reportable incident within 30 days of incident occur- rence. In addition to this, local transit agencies are required to submit the S&S-50 Safety and Security Monthly Summary C H A P T E R 3 LRT Safety Data Available from Local Transit Agencies, SSOs, and the NTD

Report form (S&S-50) on a monthly basis. This form reports a summary of the non-reportable safety and security inci- dents that occurred within the previous month. Figure 3 illus- trates the transfer of incident data from the scene of the incident across the three levels of transit administration. As indicated in Figure 3, the incident investigation can be performed by the SSO agency, the local transit agency, or both. However, in practice, it is the local transit agency that conducts the incident investigation and transfers the final investigation report to the SSO agency. Collision Data Available, Requested, and Received In an effort to conduct a comprehensive review and accurate analysis of LRT safety data, the project team requested collision databases from a variety of transit agencies at each of the three levels of transit administration. The NTD provided two complete databases in Microsoft Excel format, which contained all reported safety and security incidents that occurred between the years 2002 and 2007. The first database contained all of the “non-reportable” (formerly “non-major”) incidents reported to the NTD through the S&S- 50 form. The second database contained all the “reportable” (formerly “major”) incidents reported to the NTD through the S&S-40 form. Since the non-major incident database did not contain sufficient detail to either identify individual incidents or be used in safety analysis, only the major incident database (henceforth referred to as the “NTD database”) was examined. The project team also requested databases from a number of SSO agencies. However, only the CPUC provided a database for analysis. The CPUC database was provided in MS Excel for- mat, and included a total of 22 data fields. Based on discussions with SSO agencies, it appears that many SSO agencies do not store electronic collision data suitable for conducting safety analysis on multiple incidents. The data collected by SSO agen- cies is primarily intended for the investigation of individual incidents with the goal of developing a suitable corrective action plan. However, some SSO agencies are moving toward electronic database systems that would be suitable for conduct- ing large-scale safety analysis. The project team conducted a survey of local transit agen- cies to determine the availability and quality of collision data at the local transit agency level. In total, 24 local transit agen- cies responded to the survey. Of these 24 agencies, 21 had col- lision data in either hardcopy or electronic format. The project team followed up by requesting collision data from most of the transit agencies who had indicated that they had collision data, but many transit agencies either did not respond to these requests or declined to provide collision data. In total, colli- sion data was obtained for eight local transit agencies. How- ever, the transit agencies that did decide to provide collision data were either unable or unwilling to provide it in database format. In addition, the time period covered by the data var- ied from agency to agency. National Transit Database The United States Congress created the NTD to be “the Nation’s primary source for information and statistics on the transit systems of the United States” (5). The mandate of the NTD is set forth in Title 49 U.S.C. 5335(a), which states: To help meet the needs of individual public transportation systems, the United States Government, State and local govern- ments, and the public for information on which to base public transportation service planning, the Secretary of Transportation shall maintain a reporting system, using uniform categories to accumulate public transportation financial and operating infor- mation and using a uniform system of accounts. The reporting and uniform systems shall contain appropriate information to help any level of government make a public sector investment decision. The Secretary may request and receive appropriate information from any source. Any recipient or beneficiary of the “Urbanized Area For- mula Program” and “Other Than Urbanized Area (Rural) 22 NTD SSO Local Incident Annual Submission S&S-40 Reportable Incident Form within 30 days of incident S&S-50 Safety and Security Summary Form monthly Incident Investigation Incident Investigation Transfer of Final Investigation Report Figure 3. Transit incident data transfer across levels of transit administration.

Formula Program” is required to submit data to the NTD; there are currently over 650 transit agencies that send reports to the NTD (5). Purpose of NTD Safety Data Collection The FTA receives and compiles data in the NTD from each agency with an LRT system in the United States through a standard electronic reporting form. The data col- lected by the NTD is used by a variety of organizations for diverse purposes, including formulation of National Tran- sit Policy, state and regional planning and investment, academic/industry research, special analyses, and applications relating to the private sector/general public (5). In order to meet the diverse demands of its users and provide summary level information to the public, the NTD produces publica- tions such as Transit Profiles, Data Tables, and National Transit Summaries and Trends. These publications summa- rize important service, financial, operational, and modal data for both individual agencies and the nation as a whole. The NTD provides these publications for download free of charge on the NTD website (http://www.ntdprogram.gov/ ntdprogram). The NTD is not directly involved in efforts to improve tran- sit safety at the local level: transit safety is the responsibility of the SSO and local transit agencies. In fact, the NTD was not primarily designed for the purpose of reporting of safety and security data. It is only “in recent years (that) the NTD has grown to include safety, security, and rural transportation data” (5). According to the NTD website, the stated objective of the collection of safety and security data by the NTD is the devel- opment of performance benchmarks based on the statistics obtained from transit systems nationwide (6). The data col- lected by the NTD from the SSO and local transit agencies are used to produce publications summarizing national transit service and safety data. Data Collected by NTD The data collected by the NTD can be divided into four gen- eral categories: operational characteristics, service characteris- tics, capital revenue and assets, and financial operating statistics. Data pertinent to safety analysis primarily falls into the cate- gories of service and operational characteristics. The actual reporting of the data is the responsibility of the transit reporters, who generally fall into the following three categories: transit agencies, providers of purchased trans- portation services, and voluntary reporters (5). Transit reporters provide the NTD with the required data through the submission of the following four standardized reporting forms: the Annual Report, the Monthly Report, the Safety and Security Report, and the Rural Report (if applicable). Each of these documents consists of a series of forms that must be completed and submitted to the FTA by the transit reporter: 1. The Annual Report provides “a summary of transit char- acteristics for the fiscal year, including financial and non- financial operating statistics” (7). 2. The Monthly Report provides the NTD with a summary of the ridership and service provided by the transit agency over the past month (7). 3. The Safety and Security Report provides a summary of transit-related security and safety incidents for the year (7). 4. The Rural Report is a summary of transit data for rural sys- tems receiving funding under the FTA Formula Program for Non-Urbanized Areas (8). The data provided in the Annual Report, Monthly Report, and Rural Report are not primarily intended for safety analy- sis, but contain some information (e.g., measures of exposure to risk, vehicle revenue miles, etc.) that would be useful for that purpose. The Safety and Security Report is the primary source of the data required for safety analysis. The two forms used to report safety and security data to the NTD are the S&S-40 Reportable Incident Report form and the S&S-50 Safety and Security Monthly Summary Report form. Both of these forms are described in greater detail in Appendix E. The collision data reported in the S&S-50 form is provided in a monthly summary format that is not conducive to safety analysis. Incidents are treated as aggregate datasets that provide no detailed information regarding the conditions or circumstances present at each incident. Therefore, the data used in safety analysis is drawn almost exclusively from the S&S-40 report form. The S&S-40 form is used only to report incidents that satisfy the criteria of a “reportable incident” (formerly “major incident”). According to the 2008 Safety and Security Reporting Manual: A reportable incident is an event that involves a transit vehicle or occurs on transit controlled property and meets one or more of the following conditions: • A fatality (including a suicide or deaths resulting from other safety occurrences not otherwise classified), • Injuries requiring immediate medical attention away from the scene for one or more persons, • Property damage equal to or exceeding $25,000, or • An evacuation for life safety (7). The above criteria have undergone numerous significant changes between the years 2002 and 2008. In general, the reporting requirements have become increasingly sensitive 23

to incidents resulting in injuries/fatalities, and less focused on reporting incidents that do not. For example, prior to the issue of the 2008 Safety and Security Reporting Manual, an incident could not be classified as a major incident if a fatality was the result of a suicide, or if there were less than two injuries requiring immediate medical attention away from the scene. Table 4 shows criteria previously considered indicative of a major incident and the years they were included. The implications of these changes in incident reporting should be considered when analyzing the data contained in the NTD. For example, prior to 2003, rail collisions at a grade crossing were subject to the same reporting criteria as all other incidents, while in 2003 the NTD made all rail collisions at a grade crossing reportable. From 2004 to 2007, collisions at grade crossings were given their own specific criteria (i.e., prop- erty damage exceeding $7,500, or one or more injuries) to be used in evaluating whether or not they were reportable. In the most recent edition of the NTD Safety and Security Reporting Manual, rail collisions at an at-grade crossing are once again subject to the same reporting criteria as all other collisions, but these have been changed since pre-2003. It should also be noted that the definition of a crossing can vary. For a while, some cities (such as San Francisco) only defined LRT crossings with gates as crossings. Frequent changes in reporting standards result in some confusion among local transit agencies as to which inci- dents are reportable from year to year, and lack of uniformity in collision data between years results in reduced ability to make inferences from available data. Despite these disadvan- tages, the NTD has introduced changes to NTD reporting criteria as part of its effort to balance the need to maximize the quality and relevance of safety data in the NTD while minimizing unnecessary data collection and reporting at the local level. The NTD Database The NTD contained 66 data fields for each major incident. Table 5 shows the data fields in the NTD summarized by category of data. As shown in Table 5, the data fields contained in the NTD provide comprehensive information for a variety of data cate- gories. The date and time of each incident is clearly specified, and includes a field to indicate the time zone of the location where the incident occurred. The location of the incident is also clearly identified, and users are even provided with the option of specifying the exact longitude and latitude of the incident location. All parties involved in the collision are classified (i.e., pedestrian, motor vehicle) and described. The key categories of data with regard to safety analysis are consequence of the incident, alignment/crossing controls, environmental factors, and exposure to risk. The consequence of the incident is quantitatively defined by the number of injuries, the number of fatalities, and the extent of property damage (in dollars) resulting from the incident. The alignment is identified based on the classification of the right-of-way as exclusive, semi-exclusive, non-exclusive, etc. The crossing control field indicates the presence or absence of passive or active control devices for road/rail grade crossings. If the inci- dent occurred at a street intersection (common for LRT), the control device at the intersection is also provided. Environ- mental factors include illumination, weather, ROW condition (i.e., dry, wet, slush, etc.), ROW configuration (straight, curve, uphill, downhill, etc.), and ROW type (i.e., divided highway, intersection/grade crossing, etc.). Although the NTD included data fields for measures of expo- sure, all five data fields in this category were blank for all of the recorded incidents. In addition, examination of the NTD Safety and Security Manuals between 2003 and 2008 indicated that the 24 Reportable Criteria Years Included in Major Incident Reporting Criteria (Inclusive) A collision at grade crossing 2003 A collision at grade crossing resulting in at least one injury requiring immediate medical attention away from the scene or property damage equal to or exceeding $7,500 2004–2007 A collision with person(s) on a rail right-of-way (ROW) resulting in at least one injury requiring immediate medical attention away from the scene for at least one person 2003–2007 A mainline derailment 2003–2007 A collision involving a rail transit vehicle resulting in at least one injury requiring immediate medical attention away from the scene for at least one person 2003–2007 Forcible rape 2006–2007 Confirmed terrorist events 2006–2007 Table 4. Reportable criteria.

S&S-40 form did not provide the user with the opportunity to enter this data on the Internet reporting form. The reason for including the data fields in the NTD but not collecting data for them is unknown. The inclusion of these data fields on the S&S-40 form would provide valuable information to analysts. Finally, the NTD database includes a number of records intended to track the status of the record itself, such as when it was submitted/edited, who submitted it, how many revi- sions it has undergone, etc. There were also six data fields that did not correspond to fields on the S&S-40 form whose pur- pose could not be determined. NTD Data Quality Issues This section outlines the deficiencies identified in the NTD database, the remedial measures employed to address them, and suggestions to avoid future data quality issues. Data Cleaning Process A preliminary examination of the NTD database revealed several significant issues with the quality of the data. In an effort to facilitate data analysis, the project team performed a series of “data cleaning” exercises aimed at remedying the most common data deficiencies. This was accomplished using a systematic approach involving two major steps. First, data records were examined to see if contradictions and omis- sions in key data fields could be eliminated using the available information. Second, records that were either not LRT colli- sions or duplicates of other collision records were removed from the database. Data Record Correction. The first step of the data clean- ing process was to identify and rectify any contradictory or omitted information in the collision records. Due to the number of observed errors in the records, it was critical to avoid including errors, as well as avoiding the exclusion of viable records from the analysis by addressing these deficiencies where possible. In general, most significant contradictions/omissions in the data records generally occurred in the following fields: event category, collision manner, lighting conditions, dates, injury counts, right-of- way type, and grade crossing control type. Event Category. The “event category” field classified an incident either as a “collision,” “evacuation,” “security,” “derailment,” “fire,” or “not otherwise classified (NOC).” Dur- ing the data cleaning process, it was observed that 73 of the 2,226 records identified as LRT-related were classified incor- rectly, either due to an error in data entry, or because the classifications were not mutually exclusive. For example, it was observed that some incidents classified as derailments were actually the result of a collision, as indicated in the description field for the record. There were also collisions that were cat- egorized as NOC, or simply had a blank category field. The classification of these collisions was updated to reflect the information provided in the “event description field.” Collision Manner. The “collision manner” field described what other vehicles, objects, or individuals were involved in the collision. In 117 records, an entry of “with object: other object (describe)” was changed to “with motor vehicle” when the description field clearly indicated that another vehicle was involved. 25 Data Category No. Data Fields Additional Details Mode/Service 2 Transit Mode, Service Date/Time 6 Date, Hour, Minute, AM/PM, Time Zone, Time Period Desc. Location 7 Transit Agency, City, Location Desc., Latitude, Longitude Description of Involved Parties 2 Involved Party Category/Desc. Incident Classification 6 Event Level/Category/Type, Collision Manner, Local Level Desc. Consequences of Incident 3 No. Injuries, No. Fatalities, Property Damage ($) Alignment/Crossing Controls 4 Alignment Type, Grade Crossing Control, Intersection Control General Descriptions (i.e., actions) 7 Incident/Passenger/Other Veh./Action/Other Action/Event Desc. Environmental Factors 8 Weather, Lighting, ROW Conditions/Configuration/Type Contact Info (User) 4 Name, Phone No., Title, E-mail Data Record ID 6 Incident No., Revision No., Begin/End/Submitted Date Exposure to Risk 5 Pass. Trips, Veh. Rev. Miles/Hrs, Weekly Trip Cnt., Volume Cnt. Unknown/Other 6 Total 66 Table 5. NTD data fields by category.

Lighting Conditions. The “lighting conditions” field described the natural and artificial illumination present at the time of the collision. For 49 records, the fields that described the prevailing lighting/weather conditions and time of day produced seemingly contradictory accounts. For example, one record indicated the time of the incident to be 3:00 a.m., while the lighting conditions were listed as “daylight, clear.” In such cases, it was usually not possible to determine whether the time of day or the lighting condition had been incorrectly entered. Therefore, it was not possible to reconcile some of the apparent contradictions relating to lighting condition. Dates. The date of the incident was omitted in 27 of the records contained in the database. In some cases, it was pos- sible to retrieve the date from records of the same incident in either the SSO or local transit agency database. For example, a comparison of the NTD database with the California PUC (SSO) database for the Sacramento Regional Transit District allowed the recovery of four missing dates in the NTD data. In total, eight NTD records had dates recovered from either the SSO or local data for incidents with the same reported year, time, and incident description. Injury Counts. The number of injuries provided in the “injury count” field conflicted with the number of injuries pro- vided in the “event description” field for 7 NTD records. For example, a record listed as describing one injury included a detailed description noting that six individuals were taken to the hospital to be treated for injuries. In these cases, the “injury count” field was revised to match the “event description” field. Alignment Type. The “alignment type” field classified the ROW as exclusive, semi-exclusive, non-exclusive, etc. The clas- sification of the right-of-way in the “alignment type” field was updated in 62 records where the text description provided a clear description. Unfortunately, the description fields pro- vided insufficient information to update the alignment type for 176 of the data records. Grade Crossing Control Type. The “grade crossing con- trol type” field classified the control devices present at road/rail grade crossings. This data field was updated based on informa- tion contained in the description fields for 104 records. In most of these cases, the field was originally empty, but the text description provided a sufficiently detailed description of the type of crossing control. Transit Vehicle. The type of transit vehicle was identified incorrectly in 19 of the data records. These records identified the transit vehicle involved in the collision as an LRT vehicle when the description indicated it was a bus. These records were later removed from the database (see the Non-LRT Vehicle section). Many of the records in the NTD database suffered from incomplete and/or inaccurate information in key data fields, and required significant data cleaning in order to be included in subsequent data analysis. In many cases, the fields containing detailed descriptions of the incidents contained the information required to make the necessary corrections. However, in many instances it was not possible to make necessary corrections to the records because the record lacked sufficient descriptive information. This problem was partly due to the truncation of the detailed “event description” fields in both the NTD and SSO datasets. Discussion with FTA staff indicated that this was a database problem and that a fix would be forthcoming for the 2008 reporting that will eliminate this truncated data problem (e-mail communications with FTA staff, Feb. 2008). Unfortunately, the problem was only identified after part of the description had already been lost. In current and past editions of the S&S-40 online report- ing form, the user is required to enter information in certain mandatory data fields (indicated with an asterisk) before the user can either save or submit the report. The NTD Safety and Security Manual (on the NTD web site at http://www. ntdprogram.gov/ntdprogram/safety.htm) has included a state- ment requesting that users input information into all data fields but not all data fields are designated as mandatory. To prevent users from omitting information critical to safety analysis in the future, the NTD could expand the number of mandatory fields that require information to be entered by the user before a report can be submitted. It is more difficult to solve the problem of inaccurate data. The NTD Safety and Security Manual already contains detailed explanation for each data field, including descriptions of avail- able options/answers, and instructions for completion. The NTD continues to improve the Safety and Security Manual to address identified data reporting issues. To further mitigate the problem of incorrect data reporting, the NTD could pro- duce an annual list of common mistakes made in data report- ing, including an explanation of frequently misunderstood data fields. The feasibility of this may be an issue: to identify specific problems, the NTD would have to devote time and manpower to a data cleaning procedure similar to the one employed by the project team. Data Record Removal. Following the data correction process outlined above, the database was examined to identify and remove records that were attributed to non-existent tran- sit agencies, were the result of duplication of records for the same incident, or that incorrectly indicated the transit vehicle was an LRT vehicle. The following deletions were carried out: Non-existent Transit Agency. Of the 2,226 records in the database, 2 were attributed to a non-existent “ABC Agency” and were deleted. 26

Duplicate Records. The NTD database allows users to access previously submitted incident records to make any necessary additions or changes subsequent to the submis- sion of the data record. Each revised record is assigned a “revision number” corresponding to the number of revi- sions made to the record since original submission. Upon inspection of the NTD database, it appeared that duplicate records created for the same incident did not have a unique revision number. In most cases, the incident report was per- fectly cloned for all but one or two characteristics which were likely added during the revision of the record. For example, two records would contain identical information, but one data record listed the number of fatalities and no injuries, while the second record listed the number of injuries and fatalities. It was observed that many of the Southeastern Pennsylvania Transportation Authority (SEPTA) records suspected of being revisions contained slightly different information in many of the data fields. This made it very difficult to determine conclu- sively whether the two records represented one incident with a revised report, or two separate incidents. In cases where it was clear that multiple records referred to the same incident, either the records were merged to create one record with all of the relevant information pertaining to the collision, or all of the records except the most recent update were deleted. In addition, the data records reported by SEPTA from the year 2005 onward contained many blank data fields. This further compounded the task of identifying duplicate records in the database. Thus, while some SEPTA records were repaired, the majority had to be left intact due to lack of available information. Non-LRT Vehicle. As discussed in the Transit Vehicle section, 19 of the data records incorrectly classified the transit vehicle as an LRT vehicle when the description indicated it was actually a bus. These records were removed from the database. Table 6 shows the number of records removed from the NTD dataset due to errors or omissions in data entry. 27 Transit Agency Total Records Deleted Records % of Total Records Remaining Records ABC Agency Bi-State Development Agency Central Puget Sound Regional Transit Authority Dallas Area Rapid Transit Denver Regional Transportation District Hillsborough Area Regional Transit Authority King County Department of Transportation – Metro Transit Division Los Angeles County Metropolitan Transportation Authority Maryland Transit Administration Massachusetts Bay Transportation Authority Memphis Area Transit Authority Metro Transit Metropolitan Transit Authority of Harris County, Texas New Jersey Transit Corporation New Orleans Regional Transit Authority Niagara Frontier Transportation Authority Port Authority of Allegheny County Sacramento Regional Transit District San Diego Trolley, Inc. San Francisco Municipal Railway Santa Clara Valley Transportation Authority Southeastern Pennsylvania Transportation Authority The Greater Cleveland Regional Transit Authority Tri-County Metropolitan Transportation District of Oregon Utah Transit Authority Total Total without SEPTA 2 2 100.0% 0 9 2 22.2% 7 1 0 0.0% 1 92 18 19.6% 74 23 6 26.1% 17 12 4 33.3% 8 32 0 0.0% 32 163 34 20.9% 129 36 11 30.6% 25 55 15 27.3% 40 3 0 0.0% 3 22 3 13.6% 19 104 14 13.5% 90 2 1 50.0% 1 13 4 30.8% 9 5 2 40.0% 3 14 0 0.0% 14 66 2 3.0% 64 41 7 17.1% 34 185 17 9.2% 168 21 6 28.6% 15 1130 108 9.6% 1022 58 6 10.3% 52 90 17 18.9% 73 47 5 10.6% 42 2226 284 12.8% 1942 1096 176 16.1% 920 Table 6. Summary of records deleted from NTD dataset due to errors/omissions in data entry.

Table 6 shows that of the original 2,226 records, 284 records (12.8%) were deleted, resulting in a total of 1,942 records remaining. It should be noted that in the SEPTA data from the year 2005 onward, there was often insufficient information to determine whether one or more records were actually dupli- cations of the same event. It is suspected that if more informa- tion had been available, the number of SEPTA records deleted due to duplication would have increased. Based on the data records examined, it appears that the cre- ation of duplicate records for the same incident is a problem that needs to be addressed. It is unknown whether this problem is the result of incorrect use of the NTD database by the user, or a malfunction in the database causing the creation of multiple records when revisions are made to an incident report. The next step in the data cleaning process was the removal of records that were identified as being incidents other than collisions. The database provided by the NTD initially con- tained all incidents identified as meeting the criteria of a reportable incident and involving an LRT vehicle. However, because these incidents were reported based on the “reportable incident” criteria outlined in the Data Collected by NTD sec- tion, the records included incidents that did not involve an LRT vehicle colliding with a motor vehicle or pedestrian, which were the only incidents relevant to the project. Therefore, the project team conducted an examination of the data to remove these non-collisions from the database. Prior to 2008, the online S&S-40 form required the user to specify a “primary event” and “secondary event(s)” under the “incident classification” heading. Both primary and second- ary events listed in the S&S-40 form satisfied the NTD crite- ria of a reportable incident. The primary event was defined as the first harmful occurrence of the incident, while the second- ary event(s) were event(s) resulting from the primary event. Based on this definition, the only incidents relevant to the analysis were those with the primary event classified as a col- lision. In total, 222 of the remaining 1,942 records were iden- tified as non-collisions, based on the “collision classification” field, and removed from the database. These records included: • 5 records that had a blank “collision classification” field, • 91 records identified as derailments with no evidence of a collision, • 15 records identified as evacuations with no evidence of collision, • 22 records identified as fires with no evidence of collision, • 77 records identified as “NOC,” • 3 records identified as security problems, and • 9 records stating that the transit vehicle had left the roadway, which was assumed to indicate either a non-rail vehicle or a derailment, but in either case could not be confirmed to be the result of a collision. In summary, the data cleaning process resulted in a total of 1,720 crash records remaining from the original 2,226 total records provided by the FTA. It is this remaining dataset that is analyzed in the following sections. Disparity in Local Transit Agency Reporting to the NTD Examination of the NTD database revealed a large dispar- ity both in the number of collisions reported by transit agen- cies, and the total number of collisions reported by year. Table 7 shows the total number of collisions by year for each transit agency. Some variation in the number of collisions is expected across transit agencies. Variation is inevitable because of differences in the size of LRT system, measures of exposure to risk (i.e., vehicle revenue miles), ROW classification, etc. However, it was suspected that differences in data reporting procedures across transit agencies also accounted for a significant portion of the variation observed. In particular, the number of colli- sions reported by SEPTA was very high from 2002 to 2005, and then dropped to levels the project team considered more in line with the approach and level of reporting expected. As demonstrated in Table 6 and Table 7, SEPTA accounted for over half of the total number of collisions in the NTD database both before and after the data cleaning process. In contrast, agencies such as the New Jersey Transit Corporation reported only one collision over the course of six years. These two agencies represented the polar extremes of the observed collision reporting, so the project team contacted them in an effort to understand the cause of the variation observed across transit agencies. SEPTA staff identified two primary causes that they believed contributed to the overrepresentation of SEPTA incidents in the NTD database. The first was the fact that incidents involv- ing LRT were often reported by multiple departments within SEPTA. For example, a single incident could be reported through both the vehicle maintenance system and worker’s compensation if it resulted in both damage to the LRV and injury to the operator. According to SEPTA, this over- reporting of incidents continued until 2005 when the prob- lem was identified and rectified by SEPTA staff. The second explanation offered by SEPTA was the nature of the transit system itself. The collisions classified as related to LRT included eight street trolley lines, five of which oper- ate in mixed traffic conditions with high exposure to auto- mobile traffic. A sample of the SEPTA records was checked using mapping software, and it was confirmed that a great pro- portion of the collisions actually occurred at locations of full streetcar operation, in mixed traffic for an extended section. These sections operate differently than separate or median operating alignments which do not share space with general traffic. While streetcar alignments in mixed-traffic are still 28

considered as LRT (Type c.1) according to the previously used classification system of TCRP Report 69, most new LRT sys- tems tend to avoid sustained operations in mixed traffic and so SEPTA’s mixed traffic operations are not typical, and are not the focus of this study. A third possible explanation was that SEPTA had reported collisions that do not meet the NTD reporting criteria. It appears that before 2006, the S&S-40 Internet reporting form did not filter incidents based on whether or not they satisfied the NTD reporting criteria. This feature has been added to the most recent installment of the S&S-40 form (7). The SEPTA collision database included two separate fields used to measure the severity of collisions reported. The first field was “NTD Reportable,” which included a response of either “Yes” or “No” for each collision, based on whether the collision met the NTD criteria of a reportable incident. The second field was “NTD non-major/major,” which classified collisions as “major,” “non-major,” or were left blank (unclas- sified). This field was used to classify each incident based on the NTD criteria for major incidents. Table 8 shows the number of collisions reported by SEPTA to the NTD based on their clas- sification in the above two categories. Since transit agencies are only required to report incidents meeting the criteria of a major incident to the NTD, it is expected that an incident SEPTA classified as “major” under the NTD classification would also have an entry of “yes” in 29 Agency 2002 2003 2004 2005 2006 2007 Total Bi-State Development Agency Dallas Area Rapid Transit Denver Regional Transportation District Hillsborough Area Regional Transit Authority King County Department of Transportation – Metro Transit Division Los Angeles County Metropolitan Transportation Authority Maryland Transit Administration Massachusetts Bay Transportation Authority Memphis Area Transit Authority Metro Transit Metropolitan Transit Authority of Harris County, Texas New Jersey Transit Corporation New Orleans Regional Transit Authority Niagara Frontier Transportation Authority Port Authority of Allegheny County Sacramento Regional Transit District San Diego Trolley, Inc. San Francisco Municipal Railway Santa Clara Valley Transportation Authority Southeastern Pennsylvania Transportation Authority 202 171 147 364 The Greater Cleveland Regional Transit Authority Tri-County Metropolitan Transportation District of Oregon 17 17 17 13 42 16 30 10 16 28 31 14 17 12 22 41 18 11 10 17 19 47 16 10 14 10 16 13 17 Utah Transit Authority 1 1 1 1 1 5 2 1 4 4 5 2 3 1 7 9 8 8 8 8 1 5 1 2 6 6 4 4 1 2 1 5 3 3 1 2 1 1 2 6 3 2 4 7 7 4 7 3 5 5 2 8 2 2 1 3 1 3 5 5 3 9 7 5 9 9 1 5 6 5 Grand Total (Count) 372 278 304 505 144 117 Grand Total (Percentage of Total Crashes) Total without SEPTA 97 170 107 157 141 101 Total without SEPTA (Percentage of Total Crashes) 21.6% 16.2% 17.7% 29.4% 8.4% 6.8% 22.0% 13.8% 20.3% 18.2% 12.5% 13.1% 5 6 3 1 4 2 71 14 32 14 23 12 90 11 56 30 12 47 67 35 122 116 947 1720 773 100% 100% Table 7. Total crashes per year by transit agency from NTD database (2002–2007).

the “NTD reportable” field. Similarly, if an incident were deemed “non-major,” it should have an entry of “no” in the “NTD reportable” field. However, Table 8 shows that 25 of the 238 incidents classified as “major” were also identified as not reportable to the NTD. In addition, 19 of the 459 incidents classified as “non-major” were identified as reportable to the NTD. This shows that although the entries in the “NTD non- major” and “NTD reportable” fields for each incident co- incided in most cases, they were not always consistent. In addition, Table 8 also shows that the classification of each collision based on the above categories had little impact on whether the collision was reported to the NTD. For exam- ple, only 114 of the 213 collisions classified as “major” and reportable to the NTD were actually reported to the NTD. In addition, of the 440 incidents identified as both “non- major” and non-reportable to the NTD, 227 of them were reported to the NTD. Finally, 392 of the 634 “unclassified” incidents, all of which were determined to be not reportable to the NTD, were actually reported to the NTD. These data suggest that the primary explanation for the high proportion of SEPTA incidents in the NTD database was the reporting of incidents not meeting the NTD criteria for a “major” or “reportable” incident. It is unknown why some of the incidents classified by SEPTA as “NTD major” were considered non-reportable, and why some of the NTD “non-major” incidents were marked as reportable in their data. Technically, the NTD “major” inci- dents should all be reportable, while NTD “non-major” inci- dents should not be reportable. Unfortunately SEPTA was unable to provide additional data or assistance for a more detailed review due to the staff time involved on their part. In contrast to SEPTA, there were numerous transit agen- cies who reported very few collisions during the six years examined. For example, The New Jersey Transit Corporation and Niagara Frontier Transportation Authority reported only one and two incidents, respectively, to the NTD over a period of six years. The limited number of collisions can be partially explained by the fact that both of these transit agencies oper- ate most of their light-rail network along exclusive rights-of- way. However, it seemed unlikely that this was the only factor involved in the low number of reported incidents for many of the transit agencies. Discussions with staff from New Jersey Transit during the study team’s site visit to the HBLR indicated that they had until recently only reported collisions to the SSO on the assumption that those reports would be routed to the NTD, although they still reported all other regulatory requirement data on operation etc. to the NTD; the roles of the NTD and SSO in this case were not clear to the agency. Examination of the partial collision database provided by New Jersey Transit during the site visit substantiates this explanation. The New Jersey database contained a total of 50 incidents that occurred between the years 2002–2008. Of these incidents, 19 resulted in at least one injury, and 49 resulted in property damage. Although the data provided contained no further detail as to the number of injuries or extent of property damage for each collision, it seems likely that at least some of these incidents would have met the NTD reporting criteria. Thus, it is a pos- sibility that many of the transit agencies under-represented in the NTD were not reporting many incidents that satisfied the NTD reporting criteria. The variation in the number of collisions reported across the years should relate in part to changes in NTD’s criteria for a reportable collision (see the Data Collected by NTD section). In 2003, all collisions at grade crossings were iden- tified as reportable, but from 2004 to 2007 collisions at grade crossings were only reportable if they resulted in at least one injury away from the scene, or property damage exceeding $7,500. This change should be reflected in a decline in the number of collisions reported from 2003 to 2004, but the number of reported collisions did not drop until 2006. In this case, it appears that the change in colli- sion reporting criteria did not have a discernable impact on the number of reported collisions. The significant drop in collision reporting between 2005 and 2006 can be explained by the steep decline in incidents reported by SEPTA. The decline in SEPTA reporting corresponds with the identifi- cation and elimination of the problem of multiple SEPTA divisions submitting collision reports to the NTD for the same incident. 30 NTD Classification in SEPTA Database Flagged as NTD Reportable Reported to NTD Not Reported to NTD Total Major Yes 114 99 213 No 13 12 25 Non-major Yes 10 9 19 No 227 213 440 Unclassified Yes 0 0 0 No 392 242 634 Total 756 575 1331 Table 8. SEPTA collision reporting by NTD classification (2002–2005).

Analysis of the NTD Database This section summarizes the main findings resulting from the analysis of the NTD database, which was conducted after the completion of the data cleaning procedure. Due to the potential impact of the large proportion of collisions reported by SEPTA, all findings are also presented with the SEPTA data excluded from the analysis. Location of Collisions Collisions by type of alignment (right-of-way classification) are summarized in Table 9. When the SEPTA collisions are eliminated from the analysis, it is clear that most of the colli- sions observed occurred on non-exclusive ROW, followed by semi-exclusive ROW. This table includes columns referring to the FRA, which does not normally have jurisdiction over LRT facilities, except in cases where the LRT system is connected to or shares track with the general railway network. (In these cases those local LRT agencies must also report to the FRA.) In contrast, the vast majority (approximately 81%) of the collisions occurring within the SEPTA system were reported as occurring on exclusive ROW. One possible explanation for this phenomenon was the inclusion of collisions involving commuter rail vehicles run by the SEPTA Regional Rail divi- sion. However, the large proportion of SEPTA collisions reported under this ROW classification appears to preclude this explanation. The majority of the light rail lines run by SEPTA operate on a combination of exclusive ROW and non-exclusive ROW. It seems likely that the SEPTA collisions were reported not based on the ROW classification of the spe- cific incident location, but instead on the ROW classification that characterized most or at least some of the rail route on which they occurred. This explanation is supported by inspection of the partial database provided to the project team by SEPTA. Table 10 shows the number of SEPTA collisions included in the NTD for the years 2002–2005 based on route classification. In Table 10, only the third-rail interurban RT route 100 runs totally on exclusive ROW. The other suburban trolley routes are run mostly on exclusive ROW, but include segments of rail that are not exclusive ROW. The subway-surface trolley routes each include a section run on exclusive ROW in the city center, but operate mainly on exclusive ROW in mixed traffic conditions. The data in Table 10 show that the vast majority of collisions reported on the SEPTA routes occurred on non-exclusive ROW operating LRT vehicles in mixed traffic conditions. Collisions by type of crossing control are shown in Table 11. The majority of collisions occurred where the type of grade crossing control was reported to be traffic signals. The pres- ence of traffic signals is often indicative of a semi- or non- exclusive ROW, as operations along exclusive ROW are often accompanied by flashing lights and/or crossing gates. It is interesting to note that approximately 81% of the SEPTA col- lisions were reported as occurring at intersections with traf- fic signals, despite the fact that 81% of collisions from the same dataset were reported as occurring along exclusive ROW in Table 9. This would seem to corroborate the theory that the majority of SEPTA collisions actually occurred along semi- or non-exclusive ROW. Collisions by Measures of Exposure to Risk Table 12 shows the number of collisions for each transit agency as a function of the number of road/rail crossings in the system. The numbers of crossings per system were obtained from APTA 2004 system summary reports. In general, the number of at-grade road/rail intersections present in a system is likely to be an indicator of the total num- ber of collisions that system will experience. A high number of crossings will expose rail vehicles to an increased risk of colli- sion, particularly when the majority of the track is run on exclusive or semi-exclusive ROW. The results of Table 12, however, do not show a generally consistent relationship between the number of collisions and the number of crossings. The ratio of annual collisions per crossing ranges from 0.008 to 0.636. It is likely that this wide range of values is due in large part to the differences in collision reporting across transit agencies, as discussed in the Disparity in Local Transit Agency Reporting to the NTD section, and to the amount of service provided. Another major difference between agencies is likely to be the expo- sure of the crossings, as both the road traffic volume and the frequency of LRT operations will have an effect on the num- ber of collisions. Table 13 shows the number of collisions per million vehicle revenue miles (VRM) for each agency. If all other characteristics of transit systems were equal, it would be expected that the number of collisions for a system would increase proportionately to the number of vehicle rev- enue miles, as revenue miles is a measure of exposure. The results of Table 13 suggest significant variation between tran- sit agencies in the number of reported collisions per million vehicle revenue miles. Values ranged from 0.2 for Bi-State Development Agency (largely a type b.1 alignment with sep- arate right-of-way and at-grade intersections) to 194.7 for the King County Department of Transportation (a street car operation). Some variation among the rates of collisions is expected based on the different characteristics of the transit systems. Another known source of error for this dataset was the dis- parity in the number of collisions reported across transit agencies discussed in the Disparity in Local Transit Agency 31

Agency Exclusive ROW: At Grade Exclusive ROW: Elevated Structure Exclusive ROW: Tunnel Non- exclusive ROW: LRT/ Pedestrian Mall Non-exclusive ROW: Mixed Traffic/ LRT Non-exclusive ROW: Tran sit Mall Other Non- exclusive ROW Sem i- exclusive ROW Shared Trac k (LRT/FRA): Non-temporal Separation Shared Trac k (LRT/FRA): Tem poral Separation Not Categorized Total Bi-S tate Deve lopm ent Ag en cy 3 2 Dallas Area Rapid Transit 10 1 21 14 25 Denver Regional Transportation District 1 2 11 Hillsborough Area Regional Transit Authority 2 4 King County Department of Transportation – Metr o Tr an sit Division 1 31 Los Angeles County Metropolitan Transportation Authority 1 105 16 Mary land Tr an sit Ad mi nistr ation 7 6 1 Massa ch usetts Bay Transportation Authority 4 2 3 1 13 Mem phis Area Transit Authorit y 2 1 Metr o Tr an sit 7 3 2 Metropolitan Transit Authority of Harris County, Texas 1 87 2 New Jersey Transit Corporation 1 New Orleans Regional Transit Authority 3 1 Niagara Frontier Transportation Authority 1 1 Port Authority of Allegheny County 2 5 4 Sacram ento Regi onal Transit District 2 26 1 1 24 2 San Di eg o Trolle y, Inc. 14 8 3 5 San Fra nc isco Mu ni ci pa l Ra ilway 3 3 1 51 1 16 1 40 Santa Clara Valley Transportation Au thority 1 8 1 2 Southeastern Pennsylvania Transportation Authority 766 4 1 12 3 3 3 155 The Greater Cleveland Regional Transit Authority 16 25 1 5 Tri-County Metropolitan Transportation District of Oregon 6 2 59 Utah Transit Authority 3 21 1 9 1 Grand To ta l (C ount ) 851 4 7 6 324 19 10 158 1 9 331 Grand To ta l (Pe rce nt of To tal Crashe s) 49.5% 0.2% 0.4% 0.3% 18.8% 1.1% 0.6% 9.2% 0.1% 0.5% 19.2% Total without SEPTA (Count) 85 0 6 6 312 16 7 155 1 9 176 Total without SEPTA (Percent of Total Crashes) 11.0% 0.0% 0.8% 0.8% 40.4% 2.1% 0.9% 20.1% 0.1% 1.2% 22.8% 5 71 14 6 32 122 14 23 3 12 90 1 4 2 11 56 30 116 12 947 47 67 35 1720 100% 773 100% Table 9. Collisions by alignment type (ROW classification) (2002–2007).

Reporting to the NTD section. An example of the impact of this error can be seen in the decrease in average number of collisions per million vehicle miles observed between the years 2005 and 2006. The drop from 7.49 to 1.98 collisions per million vehicle revenue miles clearly corresponds to the steep decline in the number of collisions reported by SEPTA during this period. Both the King County Department of Transportation and the Hillsborough Area Regional Transit Authority also reported a comparably high number of collisions per million vehicle miles. This can be explained by the comparatively low number of vehicle revenue miles travelled on both of these systems, which resulted in relatively few collisions signifi- cantly inflating the ratio. Due to the disproportionately high ratio of collisions to vehicle revenue miles observed for King County Department of Transportation, Hillsborough Area Regional Transit Authority, and SEPTA, the bottom row of Table 13 shows the results of the analysis with these agencies excluded. The number of collisions per million vehicle rev- enue mile is more consistent across the time period examined with the removal of these three agencies, although it is clear that significant variation between transit agencies still remains. Figure 4 graphically illustrates the ratio of collisions to vehicle revenue miles for all agencies excluding King County Department of Transportation, Hillsborough Area Regional Transit Authority, and SEPTA. 33 Number of Collisions Reported by SEPTA to the NTD by ROW Classification (2002–2005) Route Classification SEPTA Route Exclusive ROW Mixed Traffic Unknown Interurban RT Route 100 - - - Suburban Trolley Route 101 37 21 - Route 102 30 16 - Subway-Surface Trolley Route 10 1 198 - Route 11 - 97 1 Route 13 2 113 2 Route 34 1 54 - Route 36 18 97 1 Surface Trolley Route 15 - 67 - Total 89 663 4 Table 10. Collisions in SEPTA database reported to NTD by route classification (2002–2005).

Agency Active Devices: Flashing Lights Active Devices: Gates (Median Barrier) Active Devices: Gates (No Median Barrier) Active Devices: Quad Gates Active Devices: Traffic Signal Active Devices: Train Approaching Sign No Control Device Other Passive Devices: Cross bucks Passive Devices: Stop sign Unclassified Total Bi-State Development Agency 4 Dallas Area Rapid Transit 3 12 3 1 1 1 Denver Regional Transportation District 2 3 2 Hillsborough Area Regional Transit Authority 1 1 King County Department of Transportation – Metro Transit Division 1 2 12 6 Los Angeles County Metropolitan Transportation Authority 6 8 17 1 7 1 Maryland Transit Administration 1 1 1 Massachusetts Bay Transportation Authority 3 4 Memphis Area Transit Authority 2 1 Metro Transit 7 1 Metropolitan Transit Authority of Harris County, Texas 3 10 1 11 New Jersey Transit Corporation New Orleans Regional Transit Authority 4 Niagara Frontier Transportation Authority 1 Port Authority of Allegheny County 6 1 Sacramento Regional Transit District 4 10 3 3 1 San Diego Trolley, Inc. 4 2 3 9 1 San Francisco Municipal Railway 4 1 8 9 4 18 Santa Clara Valley Transportation Authority 7 Southeastern Pennsylvania Transportation Authority 2 13 7 11 The Greater Cleveland Regional Transit Authority 8 3 2 Tri-County Metropolitan Transportation District of Oregon 2 1 1 1 Utah Transit Authority 7 2 4 1 Grand Total (Count) 36 35 50 2 40 75 20 8 43 Grand Total (Percent of Total Crashes) Total without SEPTA (Count) 34 35 50 2 32 7 1 11 37 9 2 4 65 1 1 4 20 8 29 4 766 28 20 18 1067 301 40 62 20 1 32 1 18 3 45 2 14 15 3 43 1 148 6 42 3 344 196 Total Without SEPTA (Percent of Total Crashes) 2.1% 2.0% 2.9% 0.1% 62.0% 2.3% 4.4% 1.2% 0.5% 02.5% 20.0% 4.4% 4.5% 6.5% 0.3% 38.9% 5.2% 8.0% 2.6% 0.1% 4.1% 25.4% 5 71 14 6 32 122 14 23 3 12 90 1 4 2 11 56 30 116 12 947 47 67 35 1720 100% 773 100% Table 11. Collisions at road/rail crossing by type of grade crossing control (2002–2007).

35 2004 Average 2002–2007 Agency Crossings Collisions Ratio Collisions Ratio Bi-State Development Agency 24 1 0.042 1 0.042 Dallas Area Rapid Transit 98 17 0.173 12 0.121 Denver Regional Transportation District 39 1 0.026 4 0.090 Hillsborough Area Regional Transit Authority 21 2 0.095 King County Department of Transportation – Metro Transit Division 14 8 0.571 8 0.571 Los Angeles County Metropolitan Transportation Authority 104 30 0.288 20 0.196 Maryland Transit Administration 52 5 0.090 Massachusetts Bay Transportation Authority 65 6 0.092 4 0.059 Memphis Area Transit Authority 62 2 0.024 Metro Transit 45 1 0.022 3 0.067 Metropolitan Transit Authority of Harris County, Texas 68 28 0.412 23 0.331 New Jersey Transit Corporation 88 1 0.011 New Orleans Regional Transit Authority 238 2 0.008 1 0.006 Niagara Frontier Transportation Authority 8 2 0.250 Port Authority of Allegheny County 44 4 0.083 Sacramento Regional Transit District 104 22 0.212 9 0.090 San Diego Trolley, Inc. 96 5 0.052 5 0.052 San Francisco Municipal Railway 351 11 0.031 19 0.055 Santa Clara Valley Transportation Authority 119 1 0.008 2 0.017 Southeastern Pennsylvania Transportation Authority 1,702 147 0.086 158 0.093 The Greater Cleveland Regional Transit Authority 22 14 0.636 8 0.356 Tri-County Metropolitan Transportation District of Oregon 128 9 0.070 11 0.087 Utah Transit Authority 72 1 0.014 6 0.081 Grand Total 3,564 304 0.081 310 0.082 Total without SEPTA 1,862 157 0.077 152 0.073 Source: APTA 2004 data Table 12. Collisions per number of road/rail crossings.

2002 2003 2004 2005 2006 Average Agency No. VRM (106) Ratio No. VRM (106) Ratio No. VRM (106) Ratio No. VRM (106) Ratio No. VRM (106) Ratio No. VRM (106) Ratio Bi-State Development Agency 1 5.16 0.2 1 5.23 0.2 1 5.02 0.2 1 4.44 0.2 4.38 0.0 4.85 Dallas Area Rapid Transit 5 3.97 1.3 17 5.63 3.0 17 5.15 3.3 17 5.17 3.3 13 5.10 2.6 5.01 Denver Regional Transportation District 2.98 0.0 3.76 0.0 1 3.87 0.3 4 3.73 1.1 4 4.37 0.9 3.74 Hillsborough Area Regional Transit Authority 2 3 0.08 37.4 0.08 0.0 1 0.08 11.9 0.09 0.0 0.08 King County Department of Transportation – Metro Transit Division 7 0.04 175.8 9 0.04 210.0 8 0.04 186.6 8 0.04 206.4 0.04 194.7 Los Angeles County Metropolitan Transportation Authority 42 5.78 7.3 16 6.78 2.4 30 7.70 3.9 8 8.11 1.0 10 8.05 1.2 7.29 Maryland Transit Administration 8 2.63 3.0 2.78 0.0 2.06 0.0 1 1.49 0.7 2.05 0.0 2.20 Massachusetts Bay Transportation Authority 1 5.69 0.2 2 5.73 0.3 6 5.68 1.1 6 4.54 1.3 4 5.58 0.7 5.44 Memphis Area Transit Authority 0.31 0.0 0.50 0.0 0.32 0.0 1 0.37 2.7 2 0.39 5.1 0.38 Metro Transit 1 0.51 2.0 5 1.55 3.2 3 1.79 1.7 1.28 Metropolitan Transit Authority of Harris County, Texas 28 0.47 59.2 31 0.81 38.5 14 0.86 16.3 0.71 New Jersey Transit Corporation 1 0.52 1.9 1.30 0.0 1.63 0.0 2.66 0.0 3.39 0.0 1.90 New Orleans Regional Transit Authority 0.65 0.0 0.73 0.0 2 0.97 2.1 1 0.16 0.0 0.63 Niagara Frontier Transportation Authority 0.84 0.0 2 0.76 2.6 0.76 0.0 0.74 0.0 0.77 0.0 0.78 Port Authority of Allegheny County 6 1.61 3.7 3 1.47 2.0 1.46 0.0 1.86 0.0 2 1.98 1.0 1.67 Sacramento Regional Transit District 12 2.13 5.6 4 2.17 1.8 22 2.88 7.6 7 3.43 2.0 7 3.89 1.8 2.90 San Diego Trolley, Inc. 7 7.05 1.0 3 6.92 0.4 5 6.98 0.7 5 7.06 0.7 2 8.18 0.2 7.24 San Francisco Municipal Railway 41 5.46 7.5 18 5. 53 3.3 11 5.66 1.9 10 5.52 1.8 17 5.36 3.2 5.51 Santa Clara Valley Transportation Authority 2 2.47 0.8 2 1.84 1.1 1 1.90 0.5 3 2.46 1.2 1 2.81 0.4 2.30 Southeastern Pennsylvania Transportation Authority 206 3.03 68.0 177 3.13 56.6 151 3.32 45.5 367 3.32 110.5 47 3.55 13.2 3.27 The Greater Cleveland Regional Transit Authority 10 0.94 10.6 5 0.95 5.2 14 1.01 13.8 10 1.01 9.9 5 0.87 5.7 0.96 Tri-County Metropolitan Transportation District of Oregon 16 5.66 2.8 13 5.82 2.2 9 6.02 1.5 17 6.67 2.5 7 6.38 1.1 6.11 Utah Transit Authority 9 2.32 3.9 9 2.28 3.9 1 2.97 0.3 5 2.74 1.8 6 2.83 2.1 1 14 3 2 8 21 5 4 2 3 24 1 2 2 4 10 4 19 2 190 9 12 6 2.63 0.2 2.8 0.8 24.0 2.9 2.0 0.7 4.0 2.3 34.1 0.5 2.4 2.6 2.2 3.6 0.6 3.5 0.8 58.0 9.2 2.0 2.3 Grand Total 376 59.22 6.349 284 63.46 4.475 308 66.49 4.632 508 67.82 7.490 144 72.81 1.978 347.1 66.91 5.188 Total Without Hillsborough, King, and SEPTA 161 56.16 2.87 95 60.21 1.58 149 63.04 2.36 132 64.38 2.05 97 69.17 1.40 147.5 63.51 2.32 Table 13. Collisions per million vehicle revenue miles (VRM) (2002–2006).

Collisions by Type of Impact Table 14 shows the number of collisions classified by type of impact for each transit agency. The type of impact refers to the orientation of the transit vehicle at time of impact. For example, “back” means that the transit vehicle was struck in the rear by another vehicle, while “angle” means that the other vehicle approached from an angle and struck the side of the LRT vehicle. Table 14 shows that most of the collisions (56.4%) resulted in an impact to the front of the LRT vehicle. Review of the detailed description of the incidents indicated that most of these collisions were the result of a motor vehicle making a left-turn or U-turn in front of an oncoming LRT vehicle. The same pattern was reported in TCRP Report 17: Integration of Light Rail Transit into City Streets. Collisions by Severity Table 15 shows the number of collisions resulting in fatali- ties, injuries, and property damage only (PDO) by agency. The table also gives details of the manner in which the collisions occurred. Of the 63 collisions resulting in a fatality (including SEPTA), 56 (89%) were the result of an LRT collision with a pedestrian or cyclist. The Los Angeles County Metropolitan Transportation Authority (LACMTA) and San Diego Trolley systems experienced a higher proportion of fatalities that the remaining systems. The LACMTA and San Diego collisions were almost exclusively the result of collisions with pedestrians/ cyclists, and have the highest average vehicle revenue miles of all the transit agencies observed (Table 13). Detailed exposure information for the pedestrians and cyclists at each crossing on each system is required to make any more detailed mean- ingful comments on these collisions. Of the 535 total collisions resulting in an injury, 362 (67.7%) were the result of an LRT collision with a motor vehicle, while 136 (25.4%) were the result of a collision with a pedestrian. These data suggest that in collisions between LRT vehicles and motor vehicles, the risk of fatality is relatively low when com- pared to the risk of injury. Conversely, the risk of fatality compared to injury is much higher for collisions between 37 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 2002 2003 2004 2005 2006 Year Co lli si on s pe r M ill io n Ve hi cl e Re ve nu e M ile s Metro (St. Louis) DART RTD (Denver) LACMTA MTA (Maryland) MBTA MATA Metro Transit (Minnesota) NORTA NFTA Port Authority SRTD SDMTS MUNI SCVTA RTA (Cleveland) TriMet UTA AVERAGE AVERAGE_LESS Figure 4. Safety performance: collisions per million vehicle revenue miles (2002–2006).

LRT vehicles and pedestrians (79.4% of the 63 fatal collisions involved a pedestrian, and 24.8% of all LRT collisions involving a pedestrian were fatal). Environmental Factors Contributing to Collisions Table 16 shows the number of collisions for each transit agency and the lighting conditions at time of collision. Infor- mation on lighting conditions was available for 1,344 of the 1,720 collisions. Many collisions were unclassified (369 of 1,720, or 21.5%). The effect of this omission in terms of pos- sible distortion of the results is unknown. Most collisions (1,005 of 1,344, or 74.7%) occurred in day- light conditions. The proportion of collisions occurring dur- ing period of dawn or dusk (8.9%) may be significant because of the short duration of those time periods, but once again this could only be substantiated through the availability of expo- sure data for LRVs, road vehicles, pedestrians, and cyclists by hour and lighting conditions. Figure 5 shows the percentage of collisions by time of day, excluding the data from SEPTA. 38 Agency Angle Back Fixed Object Front- En d Side swipe NOC Other Total Bi-State Development Agency 1 2 2 Dallas Area Rapid Transit 25 1 38 6 1 Denver Regional Transportation District 11 2 1 Hillsborough Area Regional Transit Authority 4 1 1 King County Depart me nt of Transportation – Metro Transit Division 23 5 2 2 Los Angeles County Metropolitan Transportation Authority 9 97 7 9 Maryland Transit Administration 6 5 3 Massachusetts Bay Transportation Authority 2 1 7 3 1 9 Memphis Area Transit Authority 2 1 Metro Transit 1 10 1 Metropolitan Transit Authority of Harris County, Texas 8 76 4 2 New Jersey Transit Corporation 1 New Orleans Regional Transit Authority 2 2 Niagara Frontier Transportation Authority 2 Port Authority of Allegheny County 3 1 7 Sacram ento Regional Transit District 34 1 13 8 San Diego Trolley, Inc. 1 20 4 5 San Francisco Municipal Railway 36 5 53 16 6 Santa Clara Valley Transportation Authority 1 8 2 1 Southeastern Pennsylvania Transportation Authority 258 94 2 357 188 13 35 The Greater Cleveland Regional Transit Authority 8 27 11 1 Tri-County Metropolitan Transportation District of Oregon 27 31 4 2 3 Utah Transit Authority 2 1 23 4 5 Grand To tal (Count) 448 103 4 793 271 17 84 Grand Total (Percentage of Total Crashes) 26.0% 6.0% 0.2% 46.1% 15.8% 1.0% 4.9% Total without SEPTA (Count) 190 9 2 436 83 4 49 Total without SEPTA (Percentage of Total Crashes) 24.6% 1.2% 0.3% 56.4% 10.7% 0.5% 6.3% 5 71 14 6 32 122 14 23 3 12 90 1 4 2 11 56 30 116 12 947 47 67 35 1720 100% 773 100% Table 14. Collisions by type of impact (2002–2007).

Fatalities Fatalities & Injuries Injuries Property Dam age Only Agency W ith Vehicle: Motor Vehicle W ith Per son (Pedestrian) W ith Cy clist Tot al W ith Ve hi cl e: Motor Vehicle W ith Ve hi cl e: Motor Vehicle W ith Per son (Pedestrian) W ith Cy clist W ith Rail Vehicle W ith objec t: Other Tot al W ith Vehicle: Motor Vehicle W ith Per son (Pedestrian) W ith Cy clist W ith Rail Vehicle W ith Object: Other Tot al Total Bi-S tate Deve lopm ent Ag en cy 1 3 1 5 5 Dallas Area Rapid Transit 5 5 4 20 8 28 32 1 1 34 71 Denver Regional Transportation District 1 1 9 4 13 14 Hillsborough Area Regional Transit Authority 1 1 1 1 2 2 4 6 King County Department of Transportation – Metro Transit Division 1 1 2 30 30 32 Los Angeles County Metropolitan Transportation Authority 1 13 4 18 2 41 18 4 63 31 1 7 39 122 Mary land Tr an sit Ad mi nistr ation 3 3 9 1 1 11 14 Massa ch usetts Bay Transportation Authority 1 1 5 12 1 3 21 1 1 23 Mem phis Area Transit Authorit y 2 2 1 1 3 Metr o Tr an sit 2 2 4 5 2 7 1 1 12 Metropolitan Transit Authority of Harris County, Texas 1 54 11 65 24 24 90 New Jersey Transit Corporation 1 1 1 New Orleans Regional Transit Authority 1 2 1 3 4 Niagara Frontier Transportation Authority 1 1 1 1 2 Port Authority of Allegheny County 4 4 7 7 11 Sacram ento Regional Transit District 2 2 12 5 4 21 31 1 1 33 56 San Di eg o Trolle y, Inc. 10 10 11 7 1 19 1 1 30 San Fra nc isco Mu ni ci pa l Ra ilway 1 4 5 1 33 27 2 4 66 41 3 44 116 Santa Clara Valley Transportation Authority 2 1 1 4 1 3 2 5 2 2 12 Southeastern Pennsylvania Transportation Authority 1 3 4 101 19 2 7 2 131 769 12 3 13 15 812 947 The Greater Cleveland Regional Transit Authority 1 1 10 10 35 1 36 47 Tri-County Metropolitan Transportation District of Oregon 1 1 2 26 12 1 39 26 26 67 Utah Transit Authority 2 2 1 11 3 3 17 14 1 15 35 Grand To ta l (C ount ) 7 50 6 63 11 362 136 18 15 4 535 1047 16 5 19 24 1111 1720 Grand Total (Percent of Total Crashes) 0.4% 2.9% 0.3% 3.7% 0.6% 21.0% 7.9% 1.0% 0.9% 0.2% 31.1% 60.9% 0.9% 0.3% 1.1% 1.4% 64.6% 100% Total without SEPTA (Count) 6 47 6 59 11 261 117 16 8 2 404 278 4 2 6 9 299 773 Total Without SEPTA (Percent of Total Crashes) 0.8% 6.1% 0.8% 7.6% 1.4% 33.8% 15.1% 2.1% 1.0% 0.3% 52.3% 36.0% 0.5% 0.3% 0.8% 1.2% 38.7% 100% Table 15. Crashes by severity and type of collision by agency (2002–2007).

40 Agency Dark: No Street- lights Dark: Street- lights Dawn or Dusk Daylight Not Applicable Unclassified Total Bi-State Development Agency 4 1 Dallas Area Rapid Transit 4 10 2 40 15 Denver Regional Transportation District 3 1 1 9 Hillsborough Area Regional Transit Authority 1 2 3 King County Depart me nt of Transportation – Metro Transit Division 1 31 Los Angeles County Metropolitan Transportation Aut hority 8 10 46 58 Maryland Transit Administration 2 3 4 5 Massachusetts Bay Transportation Authority 7 3 13 Memphis Area Transit Authority 1 2 Metro Transit 1 5 6 Metropolitan Transit Authority of Harris County, Texas 7 1 51 31 New Jersey Transit Corporation 1 New Orleans Regional Transit Authority 1 2 1 Niagara Frontier Transportation Authority 1 1 Port Authority of Allegheny County 9 2 Sacram ento Regional Transit District 1 11 5 26 13 San Diego Trolley, Inc. 4 4 3 9 10 San Francisco Municipal Railway 8 6 27 75 Santa Clara Valley Transportation Authority 8 4 Southeastern Pennsylvania Transportation Authority 131 80 654 3 79 The Greater Cleveland Regional Transit Authority 9 6 22 10 Tri-County Metropolitan Transportation District of Oregon 7 2 34 1 23 Utah Transit Authority 3 2 19 11 Grand Total (Count) 12 207 120 1005 7 369 Grand Total (Percentage of Total Crashes) 0.7% 12.0% 7.0% 58.4% 0.4% 21.5% Total without SEPTA (Count) 12 76 40 351 4 290 Total without SEPTA (Percentage of Total Crashes) 1.6% 9.8% 5.2% 45.4% 0.5% 37.5% 5 71 14 6 32 122 14 23 3 12 90 1 4 2 11 56 30 116 12 947 47 67 35 1720 100% 773 100% Table 16. Collisions by transit agency and lighting conditions (2002–2007).

Figure 5 shows that the peak period for collisions occurred between 3:00 P.M. and 5:00 P.M. The peak period may be connected with the end of the school day, and with greater pedestrian activity in the afternoon. SSO Agencies In 1991, the FTA received a list of recommendations from the National Transportation Safety Board (NTSB) outlining the need for state governments to provide safety oversight to rail transit agencies. These recommendations resulted in the issuing of the Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA), which added Section 28 to the Federal Transit Administration Act (codified at 49 U.S.C. 5330). Sec- tion 28 required the FTA to issue rule 49 CFR Part 659 (referred to as the “SSO Rule”), which required state governments to “oversee the safety and security of rail fixed guideway systems through a designated oversight agency” (4). Each state with a fixed rail guideway system was required to create a SSO agency responsible for rail transit safety and security (9). As of 2006, this program consisted of agencies in 27 states and the District of Columbia responsible for the oversight of 37 local transit agencies. Purpose of SSO Agency Safety Data Collection The role of the SSO is to establish standards for rail safety and security practices and procedures to be utilized by the transit agencies under its jurisdiction, and to oversee the implementa- tion of these practices and procedures to ensure compliance with the regulations specified in the SSO rule. The SSO Rule specifies the criteria required to fulfill this mandate in detail. In general, the SSO is required to first develop and distribute to affected transit agencies a system safety program standard which both outlines the role of the SSO agency and provides the transit agency with guidance on meeting the requirements of the SSO Rule. The SSO agency must then require transit agencies to develop and implement both a system safety pro- gram plan and a system security plan that are in compliance with the SSO Rule and the specific system safety program stan- dard developed by the SSO agency. These reports are reviewed by the SSO on an annual basis to ensure any subsequent revi- sions do not compromise compliance with the above standards. In addition, the SSO agency is responsible for conducting trien- nial on-site investigations to assess the implementation of the system safety and security plans developed by the transit agency. In addition to this broad role, the SSO agency is responsible for ensuring that incidents meeting specified criteria are inves- tigated and documented. The SSO agency must ensure the development and implementation of corrective action plans in response to these documented incidents and/or deficiencies identified during the annual reviews or triennial investigations. The SSO agency is responsible for submitting an annual report to the FTA outlining its activities over the previous twelve months. The annual report must include, in addition to other requirements, descriptions of the causal factors for investigated 41 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% 8.0% 9.0% 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Time of Day (24 Hour Clock) Pe rc en t o f 2 4- Ho ur C ra sh es 0 10 20 30 40 50 60 70 N um be r o f C ol lis io ns Figure 5. Collisions by hour of day (2002–2007) (excluding SEPTA).

incidents and the status of recommended corrective actions. Thus, in addition to providing transit agencies with overarch- ing safety and security standards, the SSO agency is actively involved in overseeing the improvement of safety conditions at specific locations in response to specific incidents. Data Collected by SSO Agencies According to the SSO Rule, SSO agencies are required to investigate, or cause to be investigated, incidents that satisfy one or more of the following criteria: • A fatality at the scene, or where an individual is confirmed dead within thirty (30) days of a rail transit-related incident; • Injuries requiring immediate medical attention away from the scene for two or more individuals; • Property damage to rail transit vehicles, non-rail transit vehicles, other rail transit property, or facilities and non- transit property that equals or exceeds $25,000; • An evacuation due to life safety reasons; • A collision at a grade crossing; • A main-line derailment; • A collision with an individual on a rail right-of-way; • A collision between a rail transit vehicle and a second rail transit vehicle, or a rail transit non-revenue vehicle (4). Information on the data management practices of a range of SSOs was gathered though interviews with the key staff at each agency. Of the seven SSOs (Arizona, California, Col- orado, Florida, Louisiana, Oregon, and Utah) contacted, both Louisiana and Colorado rely on hardcopy collision records from the local LRT agency. Arizona was contacted before their LRT system opened, and said that they intended to start col- lecting collision records using hardcopy. The other states receive their reports in some electronic form from the local LRT agency either through e-mail, an online entry system, or electronic documents. A number of the SSOs are either pursuing increased elec- tronic data handling or are already using some sort of electronic data management: • Louisiana is working with the University of New Mexico to develop a web application capable of receiving all incident data electronically, and transferring the data to FTA elec- tronically. This setup would avoid the manual entry of data into the FTA spreadsheet forms. • Arizona would like to eventually move to a database system. • Colorado enters the non-confidential data from its hard- copy collision reports into the FTA spreadsheet format. • Florida has an online reporting system developed in-house based around the Oracle database management system. It handles electronic submission of collision notification, cor- rective action, and hazards forms from the local agencies, and groups the related files. Unfortunately the system han- dles each report file as a unit so it cannot support any sort of detailed analysis. They have a process by which reports are reviewed and approved by the SSO staff electronically, or sent back to the local agency for changes. When the reports are approved, the SSO simply provides a password and user- name to the FTA staff to allow access to the system and does not then need to send the hardcopy or digital reports to the FTA. Florida has shared this system with North Carolina, and is willing to share it with other states. • Utah maintains a searchable log of collision notifications, but this is a tracking log and does not include the entire collision report. They prepare the annual report for the FTA using the reports they receive and send the report to the FTA electronically. • California receives reports from the local LRT agencies and enters them into a database with an incident number. This is searchable by type and fatalities, but it is very limited. They are currently doing a feasibility study for a new data- base system that would contain more information. They use the FTA template to report information. • Oregon receives collision notification reports by e-mail or telephone, and they are logged and tracked by agency. Portland TriMet submits collision reports electronically, while Portland Streetcar submits in hardcopy. Oregon issues the annual report to the FTA with all the data, cod- ing it into the FTA templates. SSO Databases Of the seven SSO agencies contacted, only the CPUC pro- vided a copy of their collision records initially. Utah DOT later provided some summary data which are also included in tables below where applicable. Some SSOs do not have databases, but instead have hard or soft copies of incident reports, while others could not release their databases for unspecified reasons. The CPUC database was provided in MS Excel format, and contained a total of 22 data fields. Table 17 shows the data fields contained in the CPUC database by data category. The CPUC database contains basic description of the date, time, location, and classification of incidents. The consequence of the incident is quantified based on the number of injuries and fatalities sustained by both transit employees and other parties involved in the incident. The database also provides information regarding the crossing or intersection controls present. The “warning device” field indicated whether the cross- ing or intersection was controlled by lights, gates, traffic signals, passive control devices, etc. Driver action is recorded in the general “summary of incident” fields. The CPUC database also included a separate field indicating whether the collision was the result of a vehicle driving around a crossing gate. 42

43 Data Category Number of Data Fields in the NTD Additional Details Date/Time 2 Collision Date/Time Location 4 Carrier, County, City, Street Incident Classification 2 Collision Type Consequences of Incident 4 Employee Injury/Fatality, Other Injury/Fatality Alignm ent/Crossing Controls 4 Crossing (Y/N), Crossing No., Crossing Type, Warning Device General Descriptions (i.e., actions) 2 Summ ary of Incident, Gate Drive Around (Y/N) Data Record ID 4 Report Date/Time/No./Type Total 22 Table 17. CPUC data fields by data category. SSO Data Quality Issues The data provided in the CPUC database lacks significant information required for detailed safety analysis. For exam- ple, no information regarding environmental conditions is included in the database. It is not possible to assess the rela- tionship between lighting conditions, weather, or ROW con- dition and the collisions observed. In addition, there is no information provided regarding the classification of the ROW, geometry of the alignment, or measures of exposure to risk (i.e., traffic volumes). These data are critical for performing analysis of incidents. As discussed previously, part of the mandate of the SSO is to ensure that all incidents meeting certain criteria are investi- gated to ascertain any causal/contributing factors. Subsequent to this, the SSO is mandated to develop or approve a corrective action plan that addresses the factors contributing to the spe- cific incidents. The information provided in the CPUC data- base is insufficient for the type of analysis required to clearly identify all factors that may have contributed to each incident. It seems likely therefore that corrective action plans are devel- oped using details from specific incident investigation reports, which include information that is not entered into the data- base. In order for the database alone to be used for detailed causal analysis, it would be necessary for environmental con- ditions, ROW classification, geometric design, and measures of exposure to risk be added. Local Transit Agencies Purpose of Local Transit Agency Data Collection Local transit agencies collect incident data for a wide variety of purposes, such as loss prevention, crime reporting, safety analysis, and submitting incident reports to the NTD and SSO agencies. The incident data collected at the local transit agency level is the source of the data stored in the NTD and most SSO databases. For this reason, data collected at the local transit agency level must be the most detailed of the three levels of transit administration. Data Collected by Local Transit Agencies The project team conducted a survey of local transit agen- cies to determine the availability and quality of collision data at the local transit agency level. Of the 24 local transit agen- cies that responded to the survey, 21 had collision data stored in either hardcopy or electronic format. Table 18 provides a summary of the data available at each local transit agency based on the results of the survey and subsequent discussions with transit agencies. The project team followed up on the survey by requesting collision data from most of the transit agencies identified in Table 18. Only eight local transit agencies provided collision data. In some cases no reason for the refusal was provided, but several agencies had legal or privacy concerns, and several others responded that they did not have the staff time avail- able to respond to the request. Table 19 shows the format and time period for the collision data obtained from the eight local transit agencies. None of the collision datasets presented in Table 19 were provided in database format. The majority of these datasets were records extracted from agency databases and reports. For example, the collision records provided by the LACMTA were extracted from the appendix of a report provided in PDF file format. The collision records provided by New Jer- sey Transit were written text summaries of individual inci- dents which were not compiled into a database/spreadsheet format. Santa Clara Valley Transportation Authority and Metro Transit provided collision data in PDF document for- mat that were organized in the form of a table, but these data sets contained very few data fields. The databases provided in electronic format were all in MS Excel format.

44 Collision Data Provided Agency Format Time Period Los Angeles County Metropolitan Transportation Authority Hardcopy 1990 to 2006 Metro Transit (Minneapolis) Hardcopy 2004 to 2008 New Jersey Transit Hardcopy 2000 to 2008 Santa Clara Valley Transportation Authority Hardcopy 1987 to 2007 Southeastern Pennsylvania Transportation Authority Electronic 2000 to 2007 San Francisco MUNI Electronic 2006 to 2007 Tri-Met Portland Electronic 1994 to 2008 Table 19. Local agency collision data obtained. Data Availability Agency Format Time Period Bi-State Development Agency (St. Louis) Hardcopy 1984 to present Electronic 2000 to present Edmonton Transit Hardcopy 1980 to present Electronic 1980 to present Kenosha Transit Hardcopy 2000 to present Los Angeles County Metropolitan Transportation Authority Hardcopy 1990 to present Electronic 1990 to present Memphis Area Transit Authority Hardcopy 1999 to present Metropolitan Transit Authority of Harris County, Texas Hardcopy 2004 to present Electronic 2004 to present Metro Transit (Minneapolis) Hardcopy 2004 to present Maryland Transit Administration Electronic 1985 to present New Jersey Transit – River Line Hardcopy 2004 to present Electronic 2004 to present New Jersey Transit – Hudson-Bergen Light Rail Hardcopy 2000 to present New Jersey Transit – Newark City Subway Hardcopy 1991 to present Electronic 1991 to present Port Authority of Allegheny County Hardcopy 1998 to present Regional Transit District, Denver Hardcopy 1994 to present Santa Clara Valley Transportation Authority Hardcopy 1987 to present Electronic 1987 to present San Diego Trolley Inc. Hardcopy 1981 to present Electronic 1981 to present Southeastern Pennsylvania Transportation Authority Hardcopy 2000 to present Electronic 2002 to present San Francisco MUNI Electronic 1987 to present Sound Transit Link, Tacoma Hardcopy 2003 to present Tri-Met Portland Hardcopy 1986 to 1999 Electronic 2000 to present Toronto Transit Commission Hardcopy 2004 to present Electronic 1991 to present Utah Transit Authority Electronic 1999 to present Table 18. Summary of data available at local transit agencies.

For a number of reasons, the local transit agencies were either unwilling or unable to provide collision data in a data- base format. For example, the Santa Clara Valley Transporta- tion Authority declined to provide a database extract due to privacy concerns. San Francisco MUNI was initially willing to provide a database, but ultimately decided that they did not have the authority to release the information. SEPTA was also willing to provide a complete database, but was unable due to a lack of staff resources. Due to the inability of the local transit agencies to provide comprehensive collision databases, it was not possible to conduct a comprehensive, accurate assessment of the collision data available at the local agency level. Analysis of Local Transit Agency Collision Data This section summarizes the main findings of the analy- sis of the local transit agencies’ collision data. As only eight local transit agencies supplied data, and there were gaps in the data provided, the analysis was limited. The analysis of the NTD data in the Analysis of the NTD Database section is more comprehensive. Location of Collisions Only the SEPTA database provided any detail regarding ROW classification. The data provided by the other local transit agen- cies consistently excluded information on the ROW classifica- tion of the LRT alignment, and indicated only whether each collision occurred on an exclusive ROW or in mixed-traffic con- ditions. Table 20 shows the number of collisions by ROW clas- sification for SEPTA between the years 2002 and 2007. Table 20 shows that the vast majority of LRT collisions on the SEPTA transit system occurred in mixed traffic. This is consistent with the findings of the NTD analysis outlined in the Location of Collisions section. Collisions by Severity Of the eight datasets provided by the transit agencies, three contained information regarding fatalities and three con- tained information regarding injuries resulting from LRT collisions. Table 21 and Table 22 show injuries and fatalities respectively. Both tables show the manner of collision by local transit agency. The high vulnerability of pedestrians is very clear: pedes- trians account for 75% of fatalities. They also account for 33% of injuries. The results of Table 21 are similar to those of Table 15, which also showed the vulnerability of pedestrians. Comparison of Databases The purpose of this section is to assess the consistency of collision data across the three levels of transit administration. Despite the limited response to requests for collision data, the databases obtained provide useful insight into how informa- tion is transferred between the different levels of transit agen- cies. The data provided by the local transit agencies contained sufficient detail to identify individual incidents that were con- tained in multiple databases. The hierarchical structure of data reporting provides the opportunity to assess the consis- tency of data across the levels of transit administration and determine whether there are significant differences in the data recorded in each that might have implications on their suitability for analysis. The purpose of data collection differs for the NTD, the SSOs, and the local transit agencies. As local transit agencies may have their own specific reasons for collecting certain col- lision data, and are only required to report collisions meeting specific criteria to the NTD and SSO agencies, it is likely that some of their collisions are not included in the SSO of NTD database. All the collision data in the NTD and SSO agency database should, however, be found in the applicable local agency databases. 45 Route Classification SEPTA Route Exclusive ROW Mixed Traffic Unknown Interurban RT Route 100 14 Suburban Trolley Route 101 66 41 1 Route 102 72 33 1 Subway-Surface Trolley Route 10 3 486 1 Route 11 287 1 Route 13 4 277 2 Route 34 3 131 Route 36 36 236 1 Surface Trolley Route 15 406 Total 198 1897 7 Table 20. SEPTA collisions by alignment type (ROW classification), 2002–2007.

Comparison of Local Transit Agency and SSO Agency Databases Table 23 shows the number of incidents recorded for each local transit agency in the California SSO (CPUC) database compared to the number of incidents recorded in the local transit agency databases. As expected, the number of incidents contained in the SSO database is much lower than the number of incidents contained in the local transit agency database. It is interest- ing to note that the CPUC database contained between 27–31% of the number of incidents recorded in the LACMTA and SCVTA databases, but only 10% of the num- ber of incidents contained in the San Francisco MUNI database. Table 24 shows the number of incidents reported in three local transit agency databases in California, and the number of incidents reported in the CPUC (SSO) database for the time periods shown. Only 17.6% of collisions recorded in the three local transit agency databases were also recorded in the CPUC database. The percentage varied from 4.8% of San Francisco Municipal Railway incidents to 28.8% of Los Angeles County Metro- politan Transportation Authority incidents. This variation may indicate that some local transit agencies were either reporting incidents that did not meet the SSO criteria for a reportable incident, or were not reporting incidents that did meet the SSO criteria. It is also possible that the disparity reflects the relative severity of incidents that occurred across the local transit agencies. 46 Collisions with Injuries Agency Collisio n with Motor Vehicl e Collisio n with Pedestria n Collisio n with Cyclis t Othe r Tota l Collisions with Injuries Tota l Collisions in Local Database Percent of Collisions Resulting in Injury Data Period Available Portland Tri-Met 53 68 30 – 151 542 28% 1994 – 2007 San Francisco MUNI 36 29 6 12 83 387 21% 2006 – 2007 Santa Clara Valley TA 95 18 2 – 115 476 24% 1987 – 2007 Total (Count) 184 115 38 12 349 1405 Proportion of Total Collisions (Does not sum to 100%) 13.1% 8.2% 2.7% 0.9% 24.8% 100.0% Proportion of Injuries (Sums to 100%) 52.7% 33.0% 10.9% 3.4% 100.0% Table 21. Injuries by transit agency and collision type (local data). Collisions with Fatalities Agenc y Collisio n with Motor Vehicl e Collisio n with Pedestria n Collision with Cyclis t Tota l Fatalities Tota l Collisions in Database Percent of Collisions Resulting in Fatality Data Period Available LACMTA 15 57 - 72 775 9% 1990–2006 Portland Tri-M et 0 12 2 14 542 3% 1994–2007 Santa Clara Valley TA 4 5 3 12 476 3% 1987–2007 Total (Count) 19 74 5 98 1793 Proportion of Total Collisions (Does not sum to 100% ) 1.1% 4.1% 0.3% 5.5% 100.0% Proportion of Fatalities (Sums to 100% ) 19.4% 75.5% 5.1% 100.0% Table 22. Fatalities by transit agency and collision type (local data).

Table 25 shows the proportion of CPUC data records that also appear in the local transit agency database. Table 25 shows that approximately 82% of the records con- tained in the CPUC database were also contained in the respective local transit agency databases. This suggests that the majority of incident data was supplied to the CPUC via local transit agency incident reports. However, there were 30 inci- dents recorded in the CPUC database that were not found in the local transit agency databases. Upon further inspection it was determined that of these 30 incidents, 10 were confirmed collisions, 17 were confirmed non-collisions, and 3 had insuf- ficient data to determine whether they were a collision. 47 Agency Name Number of Incidents Reported in Agency-Level Database Number of Incidents Reported in SSO database Size of SSO Database Relative to Size of Local Agency Database Time Period Los Angeles County Metropolitan Transportation Authority 278 86 30.9% Jan. 2000– Dec. 2006 San Francisco Municipal Railway 292 28 9.6% Jan. 2006– Jul. 2007 Santa Clara Valley Transportation Authority 185 49 26.5% Jan. 2000– Jan. 2007 Total 755 163 21.5% Table 23. Number of records in SSO databases compared to local transit agencies. Agency Name Number of Incidents Reported in Agency-Level Database Number of Incidents Reported in California SSO Database % of Agency- Level Incidents Also Reported in SSO Database Time Period Los Angeles County Metropolitan Transportation Authority 278 80 28.8% Jan. 2000– Dec. 2006 San Francisco Municipal Railway 292 14 4.8% Jan. 2006– Jul. 2007 Santa Clara Valley Transportation Authority 185 39 21.1% Jan. 2000– Jan. 2007 Total 755 133 17.6% Table 24. California local transit agency collision data transferred to the SSO. Agency Name Number of Incidents Reported in SSO Database Number of Incidents also Reported in Local Agency Database % of Agency- Level Incidents Also Reported in SSO Database Years Los Angeles County Metropolitan Transportation Authority 86 80 93.0% Jan. 2000– Dec. 2006 San Francisco Municipal Railway 28 14 50.0% Jan. 2006– Jul. 2007 Santa Clara Valley Transportation Authority 49 39 79.6% Jan. 2000– Jan. 2007 Total 163 133 81.6% Table 25. Proportion of SSO data records appearing in local transit agency databases.

48 The fact that some incidents included in the CPUC data- base were not found in the local transit agency databases raises the question of how they were obtained by the SSO agency. One reason CPUC has a higher rate of reported accidents may be that CPUC has a designated representative for each transit agency that has a close working relationship with the local safety manager and usually participates with the transit agency in acci- dent investigations. Depending on the severity, the CPUC may perform its own independent investigation. Besides engineers on the staff, CPUC also has FRA certified railroad inspectors, i.e., signal and train control, motor power and equipment, and track inspectors to conduct independent investigations. All incident reports originate from the transit agency, and are reported as required by FTA rules. It is also possible that the datasets provided by the three local transit agencies simply did not contain all of the records for the years examined. Whatever the explanation for the differences found, the fact that each organization maintains their own database independent of each other is surely a significant contributor to the apparent discrepancy and the uncertainty that may arise in subsequent analysis. It is clearly important to have confidence that a dataset is complete in order to carry out sta- tistical examinations. Comparison of Local Transit Agency and NTD Databases Local transit agencies are obligated to report incidents meet- ing the criteria specified in the Data Collected by NTD section to the NTD on a monthly basis. The NTD relies entirely on incident data submitted by local transit agencies; unlike the SSO agencies, the NTD is incapable of conducting indepen- dent investigations of transit incidents. Table 26 shows the number of incidents recorded for each local transit agency in the NTD database compared to the number of incidents recorded in the local transit agency databases. Table 26 shows that in total, the size of the NTD database was approximately half the size of the local agency data- bases over the same period of time. However, there was substantial variation observed in the relative number of incident records in the transit agency databases compared to the NTD database. For example, when the SEPTA records were removed, the total number of records in the NTD data- base was one-quarter the number in the remaining local transit agency databases. Table 27 shows the proportion of incident records in the local transit agency databases also found in the NTD database. Table 27 shows that, in total, almost 40% of the incident records contained in the local agency databases were also con- tained in the NTD database. However, this statistic was cut in half with the removal of the SEPTA database from the cal- culation. This difference may have been partially due to the severity of the incidents experienced on different transit sys- tems. Some agencies may have experienced a high number of total collisions with relatively few meeting the NTD criteria. However, it is more likely that this difference reflects variation in collision reporting practice, either in what incidents local Agency Name Number of Incidents Reported in Local Agency Database Number of Incidents Reported in NTD Database Size of NTD Relative to the Size of Local Agency Database Years Los Angeles County Metropolitan Transportation Authority 175 111 63.4% 2002–2006 Minneapolis Metro Transit 22 19 86.4% 2004–2007 New Jersey Transit Corporation 50 1 2.0% 2002–2007 San Francisco Municipal Railway 387 62 16.0% 2006–2007 Santa Clara Valley Transportation Authority 153 15 9.8% 2002–2007 Southeastern Pennsylvania Transportation Authority 1335 954 71.5% 2002–2005 Tri-County Metropolitan Transportation District of Oregon 253 72 28.5% 2002–2007 Grand Total 2375 1234 52.0% Total (without SEPTA) 1040 280 26.9% Table 26. Number of records in NTD database compared to local transit agencies.

49 transit agencies were willing to investigate, or in what inci- dents local transit agencies chose to report to the NTD. This variation may have been the result of transit agencies failing to report incidents that satisfied the NTD reporting criteria, or reporting incidents which did not meet these criteria. Table 28 shows the proportion of NTD incident records that also appeared in the local agency database. Approximately 76% of the records contained in the NTD database were also identified in one of the local transit agency databases. As expected, this statistic revealed the reliance of the NTD on the incident records contained in the local agency databases, but it is difficult to identify the origin of the remain- ing records. The NTD does not conduct independent inves- tigations of transit incidents, and relies entirely on the reports Agency Name Number of Incidents Reported in Local Agency Database Number of Local Agency Incidents Also Reported in NTD Database % of Agency- Level Incidents Also Reported in NTD Database Years Los Angeles County Metropolitan Transportation Authority 175 95 54.3% 2002–2006 Minneapolis Metro Transit 22 9 40.9% 2004–2007 New Jersey Transit Corporation 50 0 0.0% 2002–2007 San Francisco Municipal Railway 387 36 9.3% 2006–2007 Santa Clara Valley Transportation Authority 153 9 5.9% 2002–2007 Southeastern Pennsylvania Transportation Authority 1335 726 54.4% 2002–2005 Tri-County Metropolitan Transportation District of Oregon 253 65 25.7% 2002–2007 Grand Total 2375 940 39.6% Total (without SEPTA) 1040 214 20.6% Table 27. Local transit agency data transferred to the NTD. Agency Name Number of Incidents Reported in NTD Database for Agency Number of Incidents Also Reported in Local Agency Database % of Agency- Level Incidents Also Reported in NTD database Years Los Angeles County Metropolitan Transportation Authority 111 95 85.6% 2002–2006 Minneapolis Metro Transit 19 9 47.4% 2004–2007 New Jersey Transit Corporation 1 0 0.0% 2002–2007 San Francisco Municipal Railway 62 36 58.1% 2006–2007 Santa Clara Valley Transportation Authority 15 9 60.0% 2002–2007 Southeastern Pennsylvania Transportation Authority 954 726 76.1% 2002–2005 Tri-County Metropolitan Transportation District of Oregon 72 65 90.3% 2002–2007 Grand Total 1234 940 76.2% Total (without SEPTA) 280 214 76.4% Table 28. Proportion of NTD data records appearing in local transit agency data.

50 of local transit agencies. This suggests that the databases pro- vided by the local transit agencies did not include all of the incidents that occurred over the time period examined. Since duplicate records were noted (and removed from) the NTD, there may also be some additional near-duplicate spurious records in the NTD data. Comparison of SSO and NTD Databases Although the SSO agencies are not responsible for report- ing data directly to the NTD, both rely on the local transit agencies for their incident reports. In addition, the reporting requirements for the NTD and SSO agencies had many sim- ilarities over much of the time period examined. Examina- tion of the differences between the data records available to each can shed light on how the different reporting criteria used by each administration level over the years has affected the amount of data made available to them by the local tran- sit agencies. Table 29 shows the number of data records con- tained in both the SSO and NTD databases. In general, more transit incidents were reported by the local transit agencies to the NTD than to the SSO. This was unexpected as one would expect the reverse. One significant exception was the Utah TRAX database, which contained substantially more incidents than the NTD. However, closer inspection revealed that the Utah TRAX database included a column indicating the NTD classification of each incident. Of the 110 incidents, only 14 were classified as “Major” while 4 were classified as “Suicide.” Therefore, it is clear that in this case, the disparity was a reflection of a difference in reporting requirements. Table 30 shows the proportion of SSO incident records that were also included in the NTD. Approximately 34% of the records contained in the SSO databases were also found in the NTD database; this percentage increased to 42% when the Utah TRAX records were omitted. Table 31 shows the proportion of NTD incident records that were also contained in the SSO databases. Approximately 31% of the collision records contained in the NTD database were also present in the SSO database. Although this statistic is similar to the proportion of SSO data records found in the NTD, there is considerably more vari- ability observed between transit agencies. Conclusion This chapter had several basic objectives. It describes and compares the NTD, SSO, and local transit agency data. It con- tains detailed analyses of the locations, types, and severity of accidents (crashes) reported in the NTD database. Several implications are apparent. Firstly, it is desirable to achieve better consistency among the three reporting systems. Reporting should be both consistent and useful for researchers, transit system administrators, and oversight organizations (SSO and FTA). Secondly, the highest number of crashes involves same-direction LRV–motor vehicle collisions. Thirdly, the largest number of fatalities involves LRV–pedestrian collisions. These are the two areas, same-direction LRV–motor vehicle collisions and LRV–pedestrian collisions, where right-of- way design, operating policies, and traffic controls should be focused to reduce the number of collisions in both new starts and retrofit situations. Agency Name Number of Incidents Reported in SSO Database Number of Incidents Reported in NTD Database Years California Public Utilities Commission (SSO) Los Angeles County Metropolitan Transportation Authority 74 128 2002–2007 Sacramento Regional Transit District 33 64 2002–2007 San Diego Trolley, Inc. 52 34 2002–2007 San Francisco Municipal Railway 85 168 2002–2007 Santa Clara Valley Transportation Authority 28 15 2002–2007 Utah TRAX 110 16 2004–2006 Total (SSO) 382 425 Table 29. Number of records in SSO database compared to NTD.

51 Agency Name Number of Incidents Reported in SSO Database Number of Incidents Also Reported in NTD Database % of SSO Incidents Also Reported in NTD Database Years California Public Utilities Commission (SSO) Los Angeles County Metropolitan Transportation Authority 74 38 51.4% 2002–2007 Sacramento Regional Transit District 33 14 42.4% 2002–2007 San Diego Trolley, Inc. 52 26 50.0% 2002–2007 San Francisco Municipal Railway 85 25 29.4% 2002–2007 Santa Clara Valley Transportation Authority 28 11 39.3% 2002–2007 Utah TRAX 110 16 14.6% 2004–2006 Total (SSO) 382 130 34.0% 2002–2007 Table 30. SSO data records included in the NTD. Agency Name Number of Incidents Reported in NTD Database for Agency Number of Incidents Also Reported in SSO Database % of Agency- Level Incidents Also Reported in SSO Database Years California Public Utilities Commission (SSO) Los Angeles County Metropolitan Transportation Authority 128 38 29.7% 2002–2007 Sacramento Regional Transit District 64 14 21.9% 2002–2007 San Diego Trolley, Inc. 34 26 76.5% 2002–2007 San Francisco Municipal Railway 168 25 14.9% 2002–2007 Santa Clara Valley Transportation Authority 15 11 73.3% 2002–2007 Utah TRAX 16 16 100.0% 2004–2006 Total (SSO Data in NTD) 425 130 30.6% 2002–2007 Table 31. Proportion of NTD data records appearing in SSO database.

Next: Chapter 4 - Safety Issues and Their Treatment »
Improving Pedestrian and Motorist Safety Along Light Rail Alignments Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Transit Cooperative Research Program (TCRP) Report 137: Improving Pedestrian and Motorist Safety Along Light Rail Alignments examines pedestrian and motorist behaviors contributing to light rail transit (LRT) safety and explores mitigating measures available designed to improve safety along LRT alignments. The report also includes suggestions to facilitate the compilation of accident data in a coordinated and homogeneous manner across LRT systems. Finally, the report provides a catalog of existing and innovative safety devices, safety treatments, and practices along LRT alignments. Appendices B through E of TCRP Report 137 were published as TCRP Web-Only Document 42.

  1. ×

    Welcome to OpenBook!

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

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

    No Thanks Take a Tour »
  2. ×

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

    « Back Next »
  3. ×

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

    « Back Next »
  4. ×

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

    « Back Next »
  5. ×

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

    « Back Next »
  6. ×

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

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

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

    « Back Next »
Stay Connected!