Maintenance Planning for Rail Asset Management—Current Practices (2020)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 1.1 Project Background and Objectives As rail assets age and carry repeated loads, risk of failure increases, which in turn increases the risk of derailments and other catastrophic events. Rail failure often originates from internal fatigue defects that develop and grow to failure. The development of internal defects is expo- nential with age (accumulated traffic), and the rate of growth of a defect very much depends on axle load and support condition. Although many transit systems use track circuits that can aid in detection of broken rails, track circuits will not find internal rail defects that have yet to grow to failure, and thus provide little advance warning of potential failures. To detect internal defects, most transit agencies use ultrasonic testing (UT) as their primary method of inspecting rail. UT can be supplemented with eddy current testing, visual inspection, and video inspection, although those methods are mostly used for detecting external surface flaws and defects. Once an internal defect is found, the section of rail that contains the defect is usually replaced. Non-destructive testing (NDT) such as UT is combined with visual track inspection at various transit agencies. A number of factors, including the age of the rail, amount of traffic, environ- mental conditions, track support conditions, and location and characteristics of the rail, drive the frequency of inspection. It is important to consider all these factors to be able to identify and locate defects before they become full rail breaks. Many transit agencies use experience and regulatory guidelines for scheduling UT and other inspections. Railroads and the FRA have recently introduced risk-based UT scheduling, as currently incorporated into the FRA track safety standards (FRA, 2019). The transit industry does not currently have a uniform methodology for scheduling track inspection to detect rail flaws. The objective of this synthesis is to document the current state of transit system practices for rail inspection and maintenance as they relate to preventing rail breaks and derailments. 1.2 Technical Approach to the Project A broad range of both heavy- and light-rail transit systems in the United States received an industrywide survey. Working in conjunction with the TCRP Project J-7/Topic SE-07, “Main- tenance Planning for Rail Asset Management—Current Practices,” panel, the authors prepared a questionnaire that was sent to 28 transit agencies, as listed in Appendix A (List of Participating Transit Agencies). Appendix B contains the questionnaire. The survey questions included in the questionnaire addressed the following major topic areas: • System characteristics • Rail defect and broken rail history • Testing/inspection methodologies (i.e., UT) C H A P T E R 1 Introduction

4 Maintenance Planning for Rail Asset Management—Current Practices • Costs (labor and equipment) • Testing/inspection frequency • Challenges and constraints (i.e., maintenance windows) • Regulations, policies, and procedures • Measures of effectiveness Of the 28 agencies that received the survey, 16 responded. Those agencies appear in Table 1 and in the shaded boxes in Appendix A. It should be noted that all five of the largest heavy-rail transit systems responded (and seven of the top 10 heavy-rail transit systems responded), with the heavy-rail respondents accounting for approximately 80% of all heavy-rail transit miles. Likewise, seven of the top 10 light-rail transit systems responded, representing approximately 40% of light-rail transit system miles (and 64% of top 10 light-rail transit miles). In addition to the survey data, data on annual passenger miles were gathered from APTA’s (2019a) Public Transportation Factbook. Ridership trend information was obtained from APTA’s City Agency System System Size (Track Miles) Puerto Rico Puerto Rico Integrated Transit Authority Tren Urbano 21.4 Charlotte, NC Charlotte Area Transit System Lynx Light Rail 38.6 Minneapolis, MN Metro Transit Light Rail 44 Seattle, WA Central Puget Sound Regional Transit Authority (Sound Transit) Link Light Rail 46.9 Pittsburgh, PA Port Authority of Allegheny County Pittsburgh Light Rail 52.4 Phoenix, AZ Valley Metro Valley Metro Rail 56.4 San Jose, CA Santa Clara Valley Transportation Authority (VTA) Light Rail 82.2 San Diego, CA San Diego Metropolitan Transit System San Diego Trolley 107 Denver, CO Regional Transportation District Light Rail 120.2 Los Angeles, CA Los Angeles County Metropolitan Transportation Authority (Metro) Metro Rail 210 Oakland, CA Bay Area Rapid Transit Heavy Rail Rapid Transit 224 Chicago, IL Chicago Transit Authority Chicago “L” 224.1 Washington, DC Washington Metropolitan Area Transit Authority Metrorail 233 Philadelphia, PA Southeastern Pennsylvania Transportation Authority Subway-Elevated, Subway, Heavy Rail Rapid Transit, Trolley 254.6 Boston, MA Massachusetts Bay Transportation Authority Heavy Rail Subway, Light Rail Transit 293 New York, NY New York City Transit Authority New York City Subway 665 Table 1. Questionnaire recipients and respondents.

Introduction 5 (2018, 2019b) Q4 2018 and Q4 2019 Public Transportation Ridership Reports, for each individual agency (the Factbook has passenger mile information only up to 2017). FTA’s National Transit Database (NTD) was also used to obtain passenger car-mile data for each agency. Finally, a literature review was conducted with a specific focus on scheduling of ultrasonic rail testing. This review covered traditional and risk-based scheduling and their use worldwide, both on transit agencies and on other railway systems, including heavy axle load freight and passenger systems. 1.3 Rail Inspection Techniques For nearly a century, railroads have evaluated the condition of the rail and its susceptibility to fracture. Elmer Ambrose Sperry introduced non-destructive testing in the early 1900s and changed the complexion of how rail integrity is evaluated. Today, railroads employ several techniques for evaluating rail integrity, and their application is very much a function of rail condition and density of loading as well as economics of inspection. The primary methods used are as follows: • Ultrasonic testing • Eddy current or other non-destructive testing • Track circuits • Visual inspection by a track inspector Table 2 presents a summary of the strengths and weaknesses of these inspection techniques. 1.3.1 Ultrasonic Testing Ultrasonic testing is the primary method for evaluating rail integrity; it has the ability to look “inside” the rail for internal rail flaws. Thus, cracks emanating from the interior of the rail, which cannot be seen by the naked eye, can be identified and repaired or replaced before a Strengths Weaknesses Comments UT Internal assessment of rail Widely accepted Economic Can miss defects because of rail surface conditions or surface defects Does not see entire rail, can miss parts of base Industry standard Equipment availability for small rail systems limited Eddy current or NDT Identifies surface cracking or rolling contact fatigue Production testing method Can be expensive Limited equipment availability Focuses only on rail surface Good complement to UT Track circuits Fail-safe to minimize broken rail derailments Does not find defects Not an inspection technique Visual inspection by track inspector Performed frequently as part of inspection program Can detect rail breaks and obvious surface conditions Cannot see internal rail flaws Does not provide objective measure of defects Component of any good inspection program Table 2. Comparison of alternate testing techniques.

6 Maintenance Planning for Rail Asset Management—Current Practices rail fracture and the consequent risk of derailment develop. However, not all of the rail’s cross section is effectively evaluated by current UT technology. Typically, ultrasonic transducers send a sound wave at several angles through the rail (with respect to the rail surface contact) and at several locations transversely across the rail to identify many common defects. The time of flight through the material is interrupted when a defect exists; this interruption enables the identification of a defect. Defects in the base of the rail, such as those near the fastener contact area, are difficult to detect because the transducer signals initiate at the head of the rail. The rate of defect initiation and rate of crack growth are critical for setting effective ultra- sonic testing criteria, such as the interval between inspections. If the inspection interval is too long, a defect can initiate and grow to a size commensurate with incipient failure under a passing wheel, before the next inspection. In addition, a reliable reading from UT requires a good connection, referred to as a “couple,” between the measurement probe and the rail surface. A couplant such as water or soapy water ensures proper sound transmission. If the rail surface is very rough—as with oxidized rail—or has heavy surface defects—such as engine burns or rolling contact fatigue (spalling or shelling)—internal defects can be missed. 1.3.2 Eddy Current or Other Non-destructive Testing Eddy current and other NDT techniques focus on particular rail failure mechanisms such as surface cracking and rolling contact fatigue (RCF). These methods are often employed to address specific conditions and are limited to use only where such problems exist. These techniques do not see deep inside the rail (i.e., less than two or three millimeters) and thus do not find many internal defects, or do not find such defects until they are of a critical and dangerous size. 1.3.3 Track Circuits Railroads often use track circuits, which are usually installed for signaling purposes, as a fail-safe detection technique for rail breaks. Track circuits are not an inspection technique; rather, when a rail breaks and no longer conducts a signal, the interruption attributable to a rail break appears as a signal interruption (a train in the block) when no train actually exists. Track circuits can be an effective measure for identifying broken rails, such as those attributable to rail pull-aparts, which occur because of extreme temperature variation from installation or “neutral” rail temperature. Thus, track circuits help to prevent or minimize the occurrence of a derailment after the rail breaks. Track circuits provide no advance warning that a rail will break. 1.3.4 Visual Inspection by a Track Inspector Visual inspection, which is still used extensively by railroads, encompasses a wide range of conditions that are evaluated during an inspection, including rail condition. It can be used to identify rails that have already broken or “cracked out”—that is, when a crack grows and penetrates the external “skin” of the rail. Visual inspection can also, through careful observation, identify the onset of such rail defects as bolt hole cracks, surface fatigue, surface cracking, and RCF. Discoloration in the rail may be evidence of subsurface fatigue (such as that associated with squat defects) and can be identified visually, allowing for maintenance interventions to be performed before a rail break occurs. As noted, a rail break can be located visually after it occurs; however, the inspection has to be such that the broken rail is seen before a train passes over it. If a train passes over a rail

Introduction 7 break, the break could disrupt the traveling path of the wheel and result in a derailment. Visual inspection provides many benefits, but it is no substitute for UT because an inspector cannot see inside the rail. 1.3.5 Emerging Techniques Given the technological advancements over the past several decades, several inspection techniques have shown promise. Some of these advancements either have not progressed suf- ficiently or have not proven to be economically viable to warrant full-scale implementation. These include the following: • Fiber-optic cable adhered to the rail to identify excessive strain and the potential for a rail break • Phased array ultrasonic sensors that provide more angles and broader coverage of the rail • Contactless ultrasonic lasers that impart energy to the rail and measure the response