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.

32 6.1 Conclusions The results of this synthesis indicate that the occurrence of rail defects, broken rails, and broken rail derailments on transit systems is consistent with the rate of development found in other studies of U.S. railroads that look at larger populations of defects. Likewise, the larger and more heavily used transit systems develop increased levels of defects, a finding that is again consistent with the railroad industry at large. What is not consistent, however, appears to be the use of ultrasonic and other rail testing techniques. The larger and older transit systems seem to conduct UT testing at higher frequencies, whereas smaller, younger transit agencies seem to rely more heavily on visual inspection and track circuits, with a minimum level of UT testing. Although track inspectors are often skilled at finding surface defects, they cannot spot internal defects, which have the highest level of risk. These internal defects can grow to failure with no visible indication until they fracture (break), with the potential to cause a derailment. As these systems age, and as the rail accumulates more traffic, they will likely develop greater numbers of internal rail defects. This in turn will require a more aggres- sive UT program to help minimize the occurrence of broken rails. Because simply scheduling more tests is expensive, risk-based scheduling is a potential solution to determining how much additional testing should be performed—specifically, that which could benefit transit systems through both increased safety and minimized costs. However, risk-based scheduling is not well known or well understood by many transits—despite the fact that the recent change in FRA standards for rail testing incorporated the use of risk-based testing, as discussed previ- ously and presented in Appendix D. Some transit agencies lack a database of defects, which inhibits the implementation of risk-based scheduling. As noted, not all rail transit systems have thorough and easily acces- sible defect and maintenance records. Several rail transit systems have indicated that they are in the process of introducing more robust data management systems. Going forward, it is important for rail transit systems to have accurate defect history data, particularly as more sophisticated inspection scheduling techniques, such as risk-based UT scheduling, are introduced. Also observed was the fact that some transit systems, particularly those in cold weather regions, experience a significant number of broken rails. In most cases, such broken rails occur with few or no corresponding internal rail defects, such as those detectable by UT. These broken rails have been attributed to a number of causes; among them are initiation of defects where current UT procedures are not effective, such as on the base of the rail (e.g., because of rail base corrosion—a problem in subway tunnels) or on the field side of the rail head (because of the presence of bond wire welds). Another cause of these broken rails—particularly in locations that experience significant temperature variations—is cold weather pull-aparts. C H A P T E R 6 Conclusions and Suggestions for Further Research

Conclusions and Suggestions for Further Research 33 Track inspectors or track circuits are primarily responsible for finding these types of broken rails, which raises a concern; reliance on track circuits to find broken rails is not a practical means for managing risk. In addition, as transits (and railroads) move toward posi- tive train control, track circuits may be eliminated. Thus, a suitable method for managing risk of broken rail–related derailments—one that does not rely on track circuits—is required. Furthermore, although track inspectors have found many of these broken rails, concerns exist about the level of training of such track inspectors, particularly in the area of rail defects. In addition, there does not seem to be among the agencies surveyed any direct relationship between system size or passenger miles carried and the frequency of visual inspections. Freight and passenger railroads report similar broken rail problems (Zarembski and Palese, 2005a), but they represent a small percentage of defects detected by UT—usually less than 10%. Although this percentage is consistent with the broken rail problems that some of the larger transit systems experience, smaller transits report such defects as a more significant issue. Moreover, it is clear that transits continue to rely on UT contractors to conduct UT testing, despite occasional difficulty finding contractors in a timely and cost-effective manner. The fact that a few transits are looking to acquire their own equipment is of interest—although, on the basis of the survey results, those transits appear to be in the minority. Finally, most transit system rail maintenance plans seem to focus on replacement of worn rails, primarily in curves. The procedures for defect repair or replacement appear to be more reactive, and vary with the number and type of defect. In some cases, simply applying joint bars can be a good short- or intermediate-term fix. For more severe defects, removal of the defect and replacement with a rail plug (a short length of rail) are required. 6.2 Further Research Of particular interest is the fact that, although a large number of transits indicate that they follow FRA track safety standards, the vast majority do not follow risk-based scheduling as defined by the current version of the standards. Such a discrepancy suggests that those transits may be using an earlier version of the standards that was, in fact, time based; indeed, this is what the majority of the transits surveyed appear to use. The recent changes to the FRA standards introduced risk-based scheduling only in 2019. NYCT has already begun to use this risk-based scheduling approach, and several others are looking into it further; however, many transit systems are still not aware of this approach, as reported in this survey. Of real potential value, the risk-based scheduling approach needs to be validated for smaller transit systems with lighter axle loading. The current approach focuses on a class of defects associated with heavier axle loads that are predominant in freight railroads, and that are used to define the defect growth-rate curves inherent to the risk-based approach. In addition, large railroads will define different testing intervals and frequencies for different lines or routes that allow for focused testing. This method would not be efficient for smaller transits, so the approach needs to account for variation in system size and other relevant factors. Such accom- modation can be achieved either by inspecting the entire system the maximum number of times per year or by inspecting at the average frequency and supplementing with walking stick inspections. (Walking sticks are UT devices that inspect one or both rails. Unlike a system mounted on a hi-rail vehicle, walking sticks use a trolley pushed by an inspector to continuously inspect the rail at walking speed.) A case example of a transit system currently using the risk-based approach would help further demonstrate its benefits. Similarly, a pilot study at one or more of the small to medium-sized transit agencies would be of value in demonstrating the approach’s potential. Such research

34 Maintenance Planning for Rail Asset Management—Current Practices could also help transit systems identify the potential benefits of risk-based scheduling— particularly the smaller systems that do not currently detect significant numbers of internal rail defects. Because this type of research for some transit agencies would require improved record keeping, the introduction of rail defect and maintenance data systems would be of real value. However, research is still necessary to determine the types of data that can support both improved UT scheduling and early detection of rail breaks. Related to record keeping and database systems is the need for understanding of the tem- perature fluctuations experienced by different transit systems and, by extension, determina- tion of the proper rail-laying temperature to minimize the risk of both rail breaks and track buckles. This rail-laying temperature is directly related to the temperature range and corre- sponding temperature fluctuations experienced by an individual transit system as a function of its location and its geographic and climatic conditions. Research into the optimum rail- laying temperature for metro systems that experience significant fluctuations in temperature would be of value, especially because most existing guidelines focus on freight and large passenger railroads. Another area of interest is transits that are acquiring their own UT equipment, and its effec- tiveness as a supplement to or replacement for contractor-based UT testing. Also of interest is the introduction of new and emerging rail testing technologies, such as eddy current testing for detection of RCF and, potentially, internal defects that develop under the surface of RCF. Automated video inspection also falls into this area, particularly with the development of artificial intelligence to help analyze visual images and provide more timely and accurate defect detection. This capability would greatly complement and augment current visual inspection practices. Introduction of alternate remote monitoring techniques may also be effective. The current reliance on track circuits for broken rail protection is of concern, as noted previously. As next-generation positive train control technology is implemented, the ability to detect broken rails may be compromised or simply disappear. Thus, a suitable method for managing risk of broken rail–related derailments—one that does not rely on track circuits—is required. That method could include alternate remote monitoring techniques such as fiber optics, drones, or other new innovative approaches. The final area of research must address the required qualifications of the next generation of transit workers. Such research needs to cover the training and skills that are critical to (a) track inspectors’ ability to detect incipient rail breaks and assess rail conditions, and (b) transit engineers’ ability to safely and efficiently maintain their rail assets.