LED Traffic Signal Monitoring, Maintenance, and Replacement Issues (2008)

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.

4This chapter focuses on current practices related to LED traffic signal monitoring, maintenance, and replacement. However, it is useful to understand how we progressed to the current set of issues. EVOLUTION TO LEDs Historically traffic signals were illuminated by turning on an incandescent (60 W to 150 W) lamp behind either an 8 in. or a 12 in. lens tinted red, yellow, or green. Conventional pedestrian and arrow signal indications are similarly illumi- nated by incandescent lamps. Incandescent lamps produce light by passing electrical current through a (typically, tungsten) filament. The efficacy of light production depends on the temperature of the filament. Higher temperatures yield a greater portion of the radiated energy in the visible spectrum but may adversely affect fila- ment life. The electrical resistance to the flow of electric cur- rent in tungsten is 12 to 16 times greater when hot than at cold temperatures. The lower cold resistance produces an inrush current that lasts about a tenth of a second. Light out- put (measured in lumens) depreciates over the life of the lamp; typically less than a 20% to 25% reduction at the end of rated life. General purpose incandescent lamps typically have a rated life in the 750-h to 2,500-h range. Incandescent lamps sold specifically for traffic signals typically have a rated life of approximately 8,000 h. Typical initial lumen output for a 135 W traffic signal incandescent lamp is approximately 1,750 lumens. While incandescent lamps have been used in a variety of applications for many years, alternatives have been slowly replacing incandescents in many homes as well as outdoor and industrial locations. In most cases, the change from incandescent has been driven by more efficient light sources. However, until the 1990s, the incandescent lamp was the pri- mary light source in traffic signal control applications despite changes in other applications. An LED is a semiconductor device that creates light using solid-state electronics. A diode is composed of a layer of electron-rich material separated by a layer of electron deficient material that forms a junction. Power applied to this junction excites the electrons in the electron-rich material leading to photon emission and the creation of light. Depend- ing on the chemical composition of the semiconductor layers, the color of light emission will vary within the electromagnetic spectrum. The individual diodes are grouped together to form a traf- fic signal where, depending on the individual LED size, up to several hundred “lamps” are packaged into an array to form a traffic signal head. The “Vehicle Traffic Control Signal Heads: Light Emitting Diode (LED) Circular Signal Supple- ment,” July 27, 2005 (VTCSH LED 2005), the current ITE performance specification for circular signal indications, specifies a light display more consistent with traditional incandescent lamps and lens (1). A specification for arrows, “Vehicle Traffic Control Signal Heads: Light Emitting Diode (LED) Vehicle Arrow Traffic Signal Supplement,” has also been published (2). LEDs are much more energy efficient than their incan- descent counterparts for several reasons. LEDs are very energy efficient, producing light output with very little heat while incandescent lamps use a lot of energy generating heat. Incandescent lamps only produce white light, which must be filtered for traffic signal use, and this leads to an additional loss in energy. LEDs, on the other hand, produce colored light that quite often does not need to be filtered—all of the energy is concentrated around one color band and none is “wasted” on undesired colors. One significant difference in LEDs is that they rarely experience catastrophic failure, as do all incandescent lamps, although their light output continuously degrades over their significantly longer life. This has the potential to have a “dim” indication that will not be detected by electri- cal current monitoring methods that determine failure by a total lack of output resulting from a failed filament in an incandescent lamp. Also, because of the electronics power- ing the LEDs, the LED traffic signal design must account for a number of electrical issues including turn-on time, turn-off time, and failed impedance state in order for the safety monitoring device [conflict monitor (CM) or mainte- nance malfunction management unit (MMU)] to perform satisfactorily. CHAPTER TWO CURRENT PRACTICES RELATED TO LED TRAFFIC SIGNAL MONITORING, MAINTENANCE, AND REPLACEMENT ISSUES

5ENERGY POLICY ACT OF 2005 The Energy Policy Act of 2005 Title I, Subtitle C, Sec- tion 135 (z), applies to the manufacture and import of traffic signal and pedestrian modules (3). The sections of the EPACT 2005 relevant to traffic signals appear below. (a) DEFINITIONS— (43) The term “traffic signal module” means a standard 8-inch (200mm) or 12-inch (300mm) traffic signal indication that— (A) consists of a light source, a lens, and all other parts necessary for operation; and (B) communicates movement messages to drivers through red, amber, and green colors. (b) TEST PROCEDURES— (11) Test procedures for traffic signal modules and pedes- trian modules shall be based on the test method used under the Energy Star program of the Environmental Protection Agency for traffic signal modules, as in effect on the date of enactment of this paragraph. (c) STANDARD SETTING AUTHORITY— (z) TRAFFIC SIGNAL MODULES AND PEDESTRIAN MODULES— Any traffic signal module or pedestrian module manufac- tured on or after January 1, 2006, shall— (1) meet the performance requirements used under the Energy Star program of the Environmental Protection Agency for traffic signals, as in effect on the date of enactment of this subsection; and (2) be installed with compatible, electrically connected signal control interface devices and conflict monitoring systems. (d) GENERAL RULE OF PREEMPTION— (B) is an amendment to a regulation described in subpara- graph (A) that was developed to align California regulations to changes in the Institute for Transportation Engineers standards, entitled “Performance Specification: Pedestrian Traffic Control Signal Indications.” ENERGY STAR REQUIREMENTS The Environmental Protection Agency Energy Star Program Standard in place at the time of EPACT 2005 became the cri- teria for traffic signal modules. The energy efficiency crite- ria for Energy Star Qualified Traffic Signal Modules appear in Table 1. The net effect of the Energy Policy Act of 2005 is that it effectively eliminates the use of incandescent traffic signal modules on new installations and facilitates the transition of traffic signal modules to more energy efficient LED technol- ogy by placing energy consumption criteria on red and green LED traffic signal modules. The EPACT 2005, Title 1, Sub- title C, Section 135 (z) applies to Traffic Signal and Pedes- trian Modules manufactured on or after January 1, 2007. This act effectively overrides earlier Energy Star requirements for LED traffic signal modules. It should be noted that the focus on green and red indica- tions resulted from their longer duty cycle. The early cost of LEDs made it difficult to justify the electrical savings from the change-out of the yellow indication. As costs have come down and practical considerations of having to maintain two different types of technology have come into play, LEDs are now typically used for all colors. ITE LED SPECIFICATIONS ITE released the LED purchase specification, “Vehicle Traf- fic Control Signal Heads Part 2,” in 1998 (4). The VTCSH Part 2 was released as an interim purchase specification to meet the needs of public agencies in light of the rapid expan- sion of LEDs into traffic signal modules. The VTCSH Part 2 was intended to provide interim specifications while further human factors and photometric tests were completed on LED traffic signal modules. Studies on the effects of luminous intensity, chromatic variation, and degradation of light out- put needed to be fully understood before the ITE specifica- tion could be updated. Span wire-mounted LED traffic signal modules were implicitly excluded from the VTCSH Part 2 as luminous intensity was not addressed at an adequate variation of vertical and horizontal angles to encompass this mounting technique. ITE replaced the VTCSH Part 2 in June 2005 with a per- formance specification published under the name “Vehicle Traffic Control Signal Heads: Light Emitting Diode Circular Signal Supplement” (VTCSH-LED) (1). Full adoption of the new ITE 2005 VTCSH-LED occurred 1 year from the effec- tive date of the specification making the 1998 VTCSH Part 2 obsolete. The VTCSH-LED supplement states that agencies should use this specification as a minimum performance specification or document alternative requirements based on an engineering study. Arrow modules are addressed in an ITE-approved speci- fication entitled “Vehicle Traffic Control Signal Heads— Part 3: Light Emitting Diode (LED) Vehicle Arrow Signal Modules—A Purchase Specification.” ITE also adopted specifications on March 19, 2004, entitled “Pedestrian Traf- fic Control Signal Indications—Part 2: Light Emitting Diode (LED) Pedestrian Traffic Signal Modules.” While these pedestrian signal specifications are approved ITE standards, Module Type Nominal Wattage (at 25°C) 12-in. Red Ball 11 8-in. Red Ball 8 12-in. Red Arrow 9 12-in. Green Ball 15 8-in. Green Ball 12 12-in. Green Arrow 11 Combination Walking Man/Hand 13 Walking Man 9 Orange Hand Maximum Wattage (at 74°C) 17 13 12 15 12 11 16 12 16 13 TABLE 1 ENERGY EFFICIENCY CRITERIA FOR ENERGY STAR QUALIFIED TRAFFIC SIGNAL MODULES

it is the intent of ITE to further refine these specifications by harmonizing the language and content of these specifications with that of the new ITE 2005 VTCSH-LED. TECHNICAL ISSUES Many technical issues have been overcome since the initial implementation of LED traffic signals. The 2005 ITE speci- fication addresses many of the early problems with LEDs. The following discusses the issues that have been addressed, as well as some issues that are still outstanding. Traffic signal safety monitors continually look for potential problems in the operation of the traffic signal. The monitor- ing logic is designed around the field indication circuit and electrical characteristics of a simple incandescent lamp. An incandescent circuit is simply a filament that is connected across the hot and neutral leads of the field wiring to that particular indication. When power is applied, the filament quickly heats up and the lamp emits light. When power is removed, the filament quickly cools and the lamp no longer emits light. If the lamp fails, the circuit is open. It is very sim- ple and very predictable. An LED module is a system of trans- formers, electronic circuitry, and light-emitting diodes. Unlike a filament, the electrical characteristics of an LED module are component- and design-dependent. The slow voltage decay can be interpreted by the signal monitor as two conflicting indications being energized simultaneously causing the mon- itor to place the signal into conflict flash. This condition has been addressed and corrected in the new specification. A far worse condition than a signal monitor falsely plac- ing a signal in conflict flash is a scenario that would war- rant a conflict flash condition that is ignored or missed. There have been reports of malfunctioning LED modules that no longer emit light, which continue to present an electrical load to the conflict monitor, giving the monitor the appearance of normal operation. This potential prob- lem has been identified and addressed in the new ITE 2005 VTCSH-LED specification by requiring a failed state impedance circuit that will sense a problem and effectively appear as an open circuit to the signal monitor similar to an incandescent lamp. However, it must be realized that the only true test is observation of the presence or absence of adequate light output, which can only be inferred from LED electronics. Although the failure mode issue is addressed by the 2005 VTCSH-LED, it is not clear whether the requirements are sufficient to provide the most practically reliable systems. A different conflict occurs when the LED emits light when not energized. This can occur as a result of the design of the LED electronics. Some designs may discharge stored energy in a means that could allow the energy to be stored and dis- charged in an unintended manner. This issue, which is highly technical in nature, suggests an independent assessment of 6 current standards requirements and designs to ensure that only the most fault-tolerant approaches are used. This tech- nical review should consider all aspects of the electrical designs of LEDs with a focus on potential failure modes (including failing on and off). Another technical issue is the harmony between the ITE standard for LEDs and the NEMA standard for the signal monitor portion of the MMU in TS 2 or the Conflict Monitor (CM) in TS 1. The MMU or CM specifies operation down to 60 ± 10 volts, while LEDs are only required to operate at 80 volts. This creates an inconsistency in the expectation of two devices that should operate in harmony. A load switch is a solid state device, containing triacs, which is designed to use the traffic signal controller’s low voltage DC outputs to switch on high voltage AC outputs to the signal heads. The load switch requires a minimum current flow to trigger and hold the output in the “on” state. In some cases, extremely low wattage LED modules may not draw the necessary load current to satisfy the trigger current or hold current requirements of the load switch, resulting in flickering of the signal indication and/or permanent damage to the load switch. This issue is a practical constraint on the usage of low wattage LEDs. Current NEMA load switch specifications provide an operating range of 0.1 to 10 amps, effectively requiring a 12 W minimum load to operate. This is problematic in some applications such as single arrow application for left turns where the load could be as little as 5 W. This has lead to the practice by some of installing a load resistor in the output circuit, which can defeat the conflict monitoring process. This is an example of the lack of har- mony between load switches that were designed for incan- descent lamps that draw higher currents and LEDs whose purpose is to reduce load. While there are technical solutions such as designing a new load switch, there are issues of cost and interchangeability that make an interim solution difficult in practice. Transient voltage protection is also an issue in areas with lightning strikes. Although the 2005 ITE VTCSH-LED spec- ification makes reference to the NEMA TS 2-2003 voltage surge protection, the environment of the LED signal head electronics is significantly different from that with the road- side cabinet. These differences bring into question, as noted in the standard, the effectiveness of the current specification. Some users, most notably Texas DOT, have had concerns with LEDs and lightning strikes. Since LED modules are extremely energy efficient, they do not generate waste heat and do not melt snow and ice from their lenses like their incandescent counterparts. This can be a problem and may require some additional maintenance work to clear the faces after a driving snow storm. Some agencies have had success using products that are designed to repel rain from automobile windshields to mitigate this effect.

7Early LEDs were designed to the old ratio of red:yel- low:green of 1:4.6:2 based on circa 1933 standards developed based on glass lens. The new 2005 ITE LED performance specification changes the ratio to 1:2.5:1.3, which was based on human factors issues. While LED measurement issues are largely beyond the scope of this synthesis, some mention of the complexities is worth noting. LED clusters are unique light sources differing greatly from incandescent lamps in physical size, luminous flux (the measure of the perceived power of light), spectrum (colors), and spatial distribution (LED clusters are not point sources like incandescent lamps). Temperature of measure- ment also affects the result; therefore, LED measurement results differ considerably in various laboratories. LEDs’ unique characteristics therefore require new methods. The Commission Internationale De L’Eclairage (CIE), a standards body involved in the development of standards for measuring the optical properties of LEDs, has been recommending new definitions and new measurement conditions to reduce dis- crepancies. However, the standards have not kept up with LED technology as it continues to evolve. The net result is difficulty in assessing LED performance in the field. There is also an expectation that CIE will come up with a new edition of CIE 127-1997, which is the foundation of LED measure- ments (5). However, CIE 127-1997 did not cover sufficiently the measurement of total luminous flux and color of LEDs, which are very important in traffic signal applications. ITE SURVEY In 2006, the ITE conducted a survey of public agencies and vendors/manufacturers of LEDs. There were 75 responses from public agencies and 6 from vendors. The complete results of the survey are included in Appendices A and B. The fol- lowing is a summary of the principal findings: • 59% of respondents indicated that more than 50% of their signal modules are LEDs. • 82% use or plan to use the ITE LED specification. • The majority (73%) use a 5-year warranty period (10% do not specify a warranty). • Total failure rate (dark face) of LED modules is low (less than 5%) and is decreasing as product quality improves. • 33% do not use a qualified products list. • 85% do no compliance testing. • 60% have no monitoring/replacement procedure. • Half use the specification for minimum light output; half use no specification for minimum light output. • The number of responses dropped considerably on all questions related to agency practices/procedures for monitoring and replacement. This is possibly an indica- tion of the number of agencies with no replacement pro- gram and is consistent with survey results. • Replacement approach results: – No replacement program: 35%; – Complaint-driven: 35%; – Routine, scheduled replacement: 24%; – Replacement on vendor product life cycle: 3%; and – Based on in-service test results: 3%. • Results for scheduled replacement: – Greater than 6 years: 52%; – Five years: 38%; and – Six years: 10%. • Fifty-five percent prefer national guidelines (not stan- dards) for minimum light output with 60% preferring to adhere to the guidelines based on agency-established procedures. • Seventy-eight percent have inadequate or no funding for monitoring/replacement programs. The following is a summary of the main points ascer- tained from the survey: • Current usage of LED signal modules is prevalent and growing. • Many agencies are now approaching the life span of their initial installations. • Most use a 5-year warranty, but scheduled replace- ment tends to be on a greater than 6-year cycle; there- fore, there is a growing likelihood of old LED signal modules in the field with light output that is below specification. • Most have no routine replacement program or they are driven by complaints (complaints are less likely with LEDs as they gradually dim over time). • Although use of the 2005 ITE LED specification is strong, the minimum values for light output are of little use without routine monitoring/replacement programs. • Most do not have adequate funding for monitoring/ replacement of LED signal modules. RESULTS OF FOLLOW-UP DISCUSSIONS WITH AGENCIES FROM ITE Many of the follow-up discussions with agencies confirmed the basic issues discussed previously. The following adds some additional comments. • Indiana DOT has programmed replacement on a 6-year replacement schedule, subject to refinement as their experience base with the newer LEDs adds to their understanding. At this time, they do not plan to replace yellow LEDs on a 6-year cycle. They also plan aerial inspection and cleaning on a three-year cycle. At this time, they do not anticipate monitoring LED performance. • Texas DOT has had issues with transients. They have not been able to specifically pinpoint the cause. The problems appear to be less with new designs. Texas

8DOT reviews LED designs and does failure testing as part of their acceptance testing process. • Louisiana DOTD had historically required a special circuit in their LEDs that monitored light output. At 85% ITE minimum output, the circuit caused the LED to go dark and a breaker to create an open circuit. One manufacturer met this specification with a photo- cell on red, yellow, and green balls. The manufacturer has stopped making this LED due to Energy Star require- ments for wattage. Louisiana DOTD is concerned that the lack of output monitoring can lead to an unsafe condition.