|
|
|
|
Traffic Signal LED Module Specification Workshop and Informational Report for Snow ConditionsDownload the Printable Version [PDF, 5MB]
NoticeThis document is disseminated under the sponsorship of the U.S. Department of Transportation in the interest of information exchange. The U.S. Government assumes no liability for the use of the information contained in this document. This report does not constitute a standard, specification, or regulation. The U.S. Government does not endorse products or manufacturers. Trademarks or manufacturers' names appear in this report only because they are considered essential to the objective of the document. Quality Assurance StatementThe Federal Highway Administration (FHWA) provides high-quality information to serve Government, industry, and the public in a manner that promotes public understanding. Standards and policies are used to ensure and maximize the quality, objectivity, utility, and integrity of its information. FHWA periodically reviews quality issues and adjusts its programs and processes to ensure continuous quality improvement. Source, Cover: ©Dave Hansen/City of Green Bay Technical Report Documentation Page
Form DOT F 1700.7 (8-72) AcknowledgmentsThe development of this Guide greatly benefited from the contributions of practitioners from State and local departments of transportation, metropolitan planning organizations, and universities. The Federal Highway Administration and the authors acknowledge the individuals who provided input through review and feedback from several workshops that were conducted: David A. Noyce, Ph.D., P.E. served as the Institute of Transportation Engineers (ITE) project consultant and was responsible for the technical content for this report. ITE Project Managers Edward R. Stollof, AICP and Siva R. K. Narla provided oversight and technical coordination for the project. Michelle Birdsall was the ITE technical editor for the report. ITE wishes to thank Robert Arnold, Director, Office of Transportation Management, Federal Highway Administration, for his support in the completion of this Information Report. Technical Review and First BallotingITE would like to acknowledge the following individuals who provided a technical review and first ballot of this informational report: Carl K. Andersen Joanna Bush, P.E. David J. A. Hansen, P.E., PTOE Heath Hoftiezer, P.E., PTOE Jerry Kotzenmacher Charlton Kent Moorman, P.E. Mark Taylor Lonnie Tebow Paul Nause, P.Eng. John Vines Final Review and Second BallotingITE would like to acknowledge the following individuals who provided the final review and second balloting of this informational report: James Helmer, P.E., T.E. Christopher K. Bauer, P.E., PTOE Richard J. Montanez, P.E. LeAnn Parmenter, P.E. Randy L. Boice, P.E. Table of Contents
Table of Figures
Table of TablesIntroductionBackground on LED Traffic SignalsBasic traffic signal requirements including the illumination of traffic signal lenses/indications are set forth in the Manual on Uniform Traffic Control Devices (MUTCD). The intensity and distribution of light from each illuminated signal lens should comply with the Institute of Transportation Engineers (ITE) standards publications entitled "Vehicle Traffic Control Signal Heads" and "Traffic Signal Lamps" (see: www.ite.org/standards/index.asp). References to signal lenses are no longer limited to incandescent lamps within optical assemblies that include lenses. Research has resulted in signal optical units that are light emitting diode (LED) traffic signal modules. ITE addressed the move to LED signal lenses by creating the "Vehicle Traffic Control Signal Heads: Light Emitting Diode (LED) Circular Signal Supplement" specification in 2005, followed by a supplement to include LED arrow indications in 2007. LEDs are the new standard in traffic signals, largely due to the Energy Policy Act of 2005 which set minimum standards for energy efficiency in traffic signal and pedestrian modules. The "Vehicle Traffic Control Signal Heads: Light Emitting Diode (LED) Circular Signal Supplement" specification effectively addresses the physical, mechanical, photometric, and electrical requirements for LEDs, including environmental requirements for protecting LEDs from dust and moisture intrusion, including blowing rain. One of the unintended consequences of the transition to LED traffic signals lenses/indications is the fact that LEDs do not produce heat to the traffic signal lens or face of the traffic signal display. Therefore, LEDs do not provide a "melting" element to traffic signals in the northern climates that experience snow and ice conditions. Incandescent lenses typically provide sufficient heat from the light source such that snow and ice melt on contact with signal lenses. Nevertheless, incandescent technology is not foolproof, as some traffic engineers have experienced significant snow events where lenses associated with incandescent traffic signals become partially snow covered. Additionally, in situations where traffic signals rest in phase for long periods of time, snow and ice buildup on dark incandescent lenses has been observed. However, these lenses tend to be self-cleaning by the heat generated when illuminated. Current LED traffic signal specifications do not address a requirement to prevent snow and ice from accumulating on traffic signal lenses. Traffic engineers in northern locations that experience winter weather have reported instances in which traffic signal lenses have become snow and/or ice covered during storm events, partially or completely obstructing the view of the signal indication from drivers. Numerous property damage-only crashes and at least one fatality have been attributed to snow-covered LED traffic signal lenses. Therefore, a potential safety issue exists with LED traffic signals in locations that experience winter storm events. The U.S. Department of Transportation (USDOT) has worked with ITE in determining the issues related to the snow and ice conditions on LED traffic signal lenses. ITE and USDOT conducted a stakeholder workshop December 14-15, 2010 that created a set of proceedings documenting the issues, findings, and potential next steps in addressing the snow and ice-covered LED traffic signal lens issue. One of the outcomes of the workshop was the initiation of a small research task to further explore the environmental conditions (i.e., weather), which cause the snow and ice problem to occur. This report summarizes the research effort completed to fulfill that outcome. The primary objective of this research was to identify significant environmental (i.e, weather) variables that may contribute to snow and ice buildup on LED traffic signal lenses. If a subset of variables can be identified that may predict the potential occurrence of snow and/or ice buildup on LED traffic signal lenses, then it may be possible for traffic engineers to take a more proactive approach to eliminating or minimizing the potential negative effects. The research study design included several tasks, focused on the objective of identifying winter storm variables that lead to snow or ice-covered LED traffic signal lenses. Through a "request for information" process, researchers identified a number of winter storm events in both the United States and Canada in which traffic engineers reported snow and/or ice-covered LED traffic signal lenses. Weather data were located and tabulated with each winter storm event, both during the storm and at least one day prior, to help identify weather attributes that are common to the LED traffic signal lens covering storm events. In some locations, winter storm events in which snow-covered LED traffic signal lenses were compared to winter storm events in which snow did not cover LED traffic signal lenses to again analyze and search for weather attributes associated with the LED lens-covering storms. This report summarizes the research activities completed within each research task, along with comments pertaining to technologies, specifications, and future research. Appendix A provides information on each study site as well as the weather data information for the reader's reference. Background Information on Responding AgenciesThe first task in the research was to identify sites in which a specific date and time of a winter storm event that led to snow and/or ice buildup on LED traffic signal lenses could be identified. A "Request for Information" document was created and circulated among ITE members in the traffic engineering community, along with the participants in the initial workshop, to identify these storms and some of the basic attributes of the locations. A copy of the "Request for Information" document is included in Appendix B. The objective of this data collection exercise was to identify locations which have experienced problems with snow and/or ice buildup on LED traffic signal lenses and transportation agencies that were willing to participate in this research. Nine complete responses to the "Request for Information" were received in November 2011. The responding agencies included:
To be included in the research analysis, responding agencies were required to provide specific dates and times of the referenced winter storm events that led to snow and/or ice buildup on LED traffic signal lenses, and they had to be able to identify the intersections affected. Researchers were most interested in storm events in which at least 50 percent of the LED traffic signal lenses were covered by snow and/or ice and occluded from the traveling public. While having experienced snow and/or ice-covered LED traffic signal winter storm events, the Virginia Department of Transportation, City of Winnipeg, Michigan Department of Transportation, and City of Helena were unable to provide specific storm event date data. Each of these locations were not included in this analysis as a result. Therefore, the "Request for Information" produced five agencies with the necessary data to be included in this research. Since it was desirable to include as many agencies as possible, researchers contacted several additional agencies who did not respond to the initial data request but were known to have experienced snow and/or ice on LED traffic signal lenses. This process led to three additional sites that had the necessary data and were subsequently included in this research. They were:
Therefore, a total of eight agencies were included in the final analysis, with several providing more than one study site or storm event. The geographical location of each agency location is highlighted in Figure 1. Note that the sites generally represent the upper Midwest of the United States along with one southeastern Canadian site.
To complete the preliminary data collection process, additional information was obtained and included with each of the study sites and storm events considered. A review of Appendix A shows that each site description generally contains:
Table 1. Information on the 12 Storm Events Evaluated
1Provided by Utah DOT. 2Estimate for Saratoga Springs. Utah DOT maintains 1,121 signalized intersections. 3Provided by Wisconsin DOT. 4Total Wisconsin DOT signalized intersections statewide. Weather Data CollectionThe second task involved an extensive data collection effort to capture all of the pertinent and available weather data for the storm events identified. Weather data collection was completed for the 8 agencies, 9 sites, and 12 storm events presented in Table 1. Four Internet-based public sources were considered to obtain the necessary weather data. Each is listed below with their corresponding Internet link:
Fortunately, weather data for all U.S. study sites were available at the NCDC and Weather Underground sites; hence, these two sites were used exclusively. Weather data for the RMOY, Newmarket, Ontario site was obtained from a Canadian National Weather Data source: www.climate.weatheroffice.gc.ca/climateData/canada_e.html. A review of Appendix A will show numerous data tables containing weather data. In all National Climatic Weather Center (NCWC) data tables presented in Appendix A, and described hereafter, a number of codes are used to describe weather data that may not be common to the transportation engineering community. The following list of weather attributes and codes (presented by typical NCWC column number) is intended to provide an initial description of the data terms used. Some of these attributes are further defined later in this report. A similar description is presented at the beginning of Appendix A. Commonly used codes include:
Appendix A also contains a description of each study site along with photos and relevant information provided. Weather and Location Data AnalysisThe primary objective of this analysis task is to analyze the large amount of weather data obtained in an effort to identify weather attributes that may be empirically and/or statistically correlated to the weather events leading to snow and ice-covered LED traffic signal lenses. Three approaches were taken to complete this analysis: (1) an anecdotal approach; (2) a statistical approach; and (3) a meteorological evaluation. Anecdotal ApproachAn anecdotal approach was used based on a detailed review of all of the weather data obtained. Weather attributes that were consistently recorded with each of the 12 storm events were identified and compared across data sets included in Appendix A. The objective was to identify weather variables and their associated values that were consistent in each of the LED traffic signal lens-covering weather events considered. Weather variables considered included:
These weather variables with quantitative or qualitative commonality were recorded and further analyzed. A summary of the conclusions include:
Clearly, a common requirement for snow and ice-covered LED traffic signal lenses is a wind of sufficient speeds to blow the snow at an angle towards the signal lens and below the visor, usually aided by strong wind gusts. All storms had at least 10 mph sustained winds with gusts generally exceeding 20 mph. Another extremely consistent variable is air temperatures very near freezing, falling from higher temperatures above freezing shortly before the snow storm. Humidity is high and moisture is high, increasing the water content of the snow and the snow's ability to adhere to the LED lenses. Statistical AnalysisThe primary focus of the statistical analysis was to determine if it was statistically possible to compare storm events that produced snow-covered signal lenses to storm events that did not, and to identify weather attributes that were significant to the differences. Furthermore, a statistical analysis was also attempted to determine if the change in weather variables within a specific storm had statistically identifiable changes in variable values. To consider the differences in weather attributes between storms that led to snow-covered LED traffic signal lenses and storms that did not, three of the sites with the most robust and direct weather data were selected. These locations included:
Specifically, each of these three locations had abundant weather data; storm durations that were well documented; weather data from more than one reporting source; and perhaps most importantly, a reporting weather station located directly within the jurisdiction reporting. Note that in some of the jurisdictions, the reporting weather station was some distance from the jurisdiction. The statistical procedures attempted to model the relationship between the explanatory and response weather variables included in the data sets. Since the data set contained both quantitative and qualitative measures and multiple response variables, a binary response statistical model was considered and used in the generalized linear models methodology. This methodology can be easily repeated using one of several off-the-shelf or Internet-based statistical modeling software packages. The discrete nature of the responses required a special function, usually referred to as a "link function," to serve as a bridge between the linear predictors and the mean of the distribution we use in the analysis. For the binary responses in the analysis, the logistic or "logit" link was used. Logistic regression improves the analysis when categorical dependent variables are also part of the analysis. The results of the climatological modeling of Green Bay, WI, Madison, WI, and Sioux Falls, SD, data are presented in Table 2.
The results of the statistical analysis suggest that the variables in Table 2, at a 95 percent level of confidence, are statistically significant predictors of snow-covered LED lenses in the winter storm events considered. Detailed descriptions of each of the Table 2 variables include:
Using the four statistically significant variables reported above and the associated coefficients in Table 2, it is possible to create a statistical model such that future data can be used to predict the probability of a storm event that leads to snow and ice-covered LED traffic signal displays. Nevertheless, as described in the next section, the model predictability will likely have little practical significance. In other words, it is possible to model past storm events, but unlikely that the attributes of a future storm event will be holistically consistent with past events, with or without the buildup of snow and/or ice on LED traffic signals lenses. Meteorological EvaluationTwo meteorologists were consulted to assist in the analysis, given their direct expertise on the primary consideration in this research. A meteorological researcher with expertise in winter storm events provided insight and expert recommendations during the analysis process. Also, a professional meteorologist who forecasts weather for local television and radio media outlets, and analyzes and provides daily weather data throughout the United States, was retained to complete a thorough review of the data. Additionally, he looked more formally at professional weather databases and historical models to help identify weather attributes that may not be captured in the public weather data sites used. The results of each of these analysis steps are provided below. First, both weather experts agreed that the databases considered from NCWC and Weather Underground contained the necessary variables important to this analysis. There was nothing additional in their local databases that would add to this research. Although meteorologists consider the same variables identified in this research, most weather forecasting is completed by computer weather models which may not directly report the variables pertaining to this research. Nevertheless, all weather variables are considered and reported as necessary. The meteorologists concluded that the major factors in these winter storm events that lead to snow and/or ice buildup on LED traffic signal lenses are simply wind speed and temperature (wet bulb). The fact that most of the storms identified occurred in late winter or early spring is expected as these warmer-type storms tend to occur later in the winter as the typical storm tracks brings the storms from the southwest more often (as opposed to clipper-type systems from the northwest that tend to be colder). Nevertheless, it is possible for storms moving in from the northwest or other directions to produce the same conditions, i.e., wet snow and wind, although it may be less probable. Humidity is always fairly high in any precipitation situation, even with colder storms which will not produce wet snow. This is because there is an abundance of moisture in the atmosphere due to the precipitation, and that moisture will evaporate into the atmosphere. Atmospheric pressure tendency may not be a significant predictor, as pressure will always fall as a weather front approaches and rise after it passes. Precipitation can occur on both sides of the front, though the wet snow is much more common ahead of the front (when pressure is falling). The actual minimum pressure could be an indicator variable. The question remains as to what types of storms produce the type of conditions described above such that snow and ice-covered LED traffic signal lenses can be predicted. Unfortunately, meteorologists agree that the simple answer is all types. Nevertheless, the key variable appears to be the temperature (wet bulb) as the storm approaches. If the temperatures are above freezing at the surface before the storm moves through, chances are there will be at least some wet snow. Storms that produce more wet snow generally move in from the southwest, although some parts of the United States and Canada can experience storms from most any direction. Southwest storms also tend to be stronger storms, which mean stronger winds (note that the strength of storms is typically determined by the atmospheric pressure). As a general rule, the lower the atmospheric pressure, the stronger the wind. In the opinion of the meteorologists, a specific set of weather variables that are a perfect predictor of snow and/or ice-covered LED traffic signal lenses is not practical. The meteorologists often consider statistical modeling and evaluations of previous storms, which is part of their forecasting methodology, but through experience, they know that models of previous storm are often imperfect predictors. The fact that models are not always good predictors is why weather professionals often have extreme variability in weather forecasts. Considering typical weather data and generalizing the results, it is likely that stronger storms with minimum pressures under 29.8 inches of mercury (wind producing), moving into an area with above freezing temperatures, is a good indicator of the need to be aware of the potential for snow and/or ice buildup on LED traffic signal lenses. The storm would most likely have to be several hours in duration to provide time for the temperatures to move through the above to below freezing range. The potential for snow and/or ice-covered LED traffic signal lenses also increases if the forecast snow totals are higher, as this often associates with the snow-to-liquid equivalent ratio. The snow-to-liquid equivalent ratio is the predictor of how many inches of snow is produced as compared to an inch of liquid (rain) water. The ratio for wet snow that is commonly consistent with snow-covered LEDS signal lenses is 10:1 or less. Both meteorologists suggested that from a forecast standpoint, it could be reasonably predicted that an upcoming winter storm had no potential for causing problems with LED traffic signal lenses as well as that an upcoming winter storm had a reasonable potential to cause snow and/or ice buildup on LED traffic signal displays. Unfortunately, more specific predictions are not possible since every storm is different and would need to be considered individually to determine the potential risk. Results of the Anecdotal, Statistical, and Meteorological AnalysesThe results of the combined analyses, including consideration of weather data in both the anecdotal and statistical methods, along with the evaluation, analyses, and opinions of the meteorologists, suggest that near freezing temperatures (typically wet bulb temperatures starting above freezing or dropping below freezing) and sufficient wind speed (associated with atmospheric pressure) are the two key attributes of the winter storm events that produce snow and/or ice-covered LED traffic signal lenses. These conditions are more common in late winter storms, which are consistent with the dates of the storms described in this research. Moisture conditions leading to predicted snow-to-water equivalency ratios of 10:1 (or less) is common. Results of the data analysis and meteorologist's opinions also agree that these weather attributes are potential predictors, and may or may not lead to snow and/or ice-covered LED traffic signal lenses. Nevertheless, winter storms with the following attributes have a highly probable potential for snow/ice covered LED traffic signals displays: predicted temperatures near or passing through 32 degrees Fahrenheit; humidity levels at approximately 90 percent; atmospheric pressures at or below 29.9 inches of Mercury (i.e., less than average sea level pressure); and wind speeds greater than 10 mph with northerly high speed gusts. CountermeasuresSeveral reactive and proactive countermeasures to solve snow and ice-covered LED traffic signal lens problems have been employed by traffic engineers around the country. The following example focuses on the City of Green Bay, Wisconsin, which provided details on the countermeasures it has considered. A summary of attempted countermeasures from other locations included in this research is also provided. Reactive countermeasures employed include manual brush cleaning, compressed air spraying, and antifreeze or deicing spraying. Each of these countermeasures requires manual labor and staff to be in the field during or immediately after the storm event to clear the effected traffic signal lenses. Figure 2 depicts a long handled brush commonly used. Figure 3 shows a lens cleaning operation in Utah.
Proactive countermeasures include the Fortran Snow Sentry,™ the McCain Snow Scoop visor, and the before-storm application of chemical deicers. The Fortran Snow SentryTM was a product developed after multiple experiments with lens heating elements (see www.ite.org/standards/led/files/13 %20Fortran.ppt). Heating elements were evaluated in both environmental chambers and field installations. Results showed that the heating elements were not successful in eliminating the snow and ice buildup on lenses, and may also lead to discoloration and deformation of the lens. These issues were in addition to the cost and complexity of adding power to operate the heating elements. The Fortran Snow SentryTM is a lens cover designed to guard the signal lens from snow and ice buildup. As shown in Figure 4, the slope of the lens cover generally matches the signal visor, sloping down and towards the signal head. The shape is believed to minimize the ability for snow and ice to stick to the lens cover, and the clear acrylic material used to make the Snow SentryTM was independently tested and shown to have no impact on the transmission of light from the signal indication. The McCain Snow Scoop visor was designed with a louvered vent at the top that funnels wind across the signal face. As shown in Figure 5, the wind funnel air speed across the signal lens decreases the probability of snow and ice adhering to the signal lens. Several traffic agencies in the U.S. have experimented with one or both of these devices. Green Bay, Wisconsin reported improvements with the Snow Scoop visor and the Snow Sentry,TM although neither was reported to completely prevent snow and ice buildup on the lenses. Figure 4 provides an example of the Snow Sentry™ and Figure 5 shows the Snow Scoop. Figure 6 depicts the results of an in-field comparison test of the two devices. Although Figure 6 does not show a comparative signal display without the Snow Scoop visor or Snow SentryTM, both devices reportedly provided improvement in preventing snow and ice buildup on the LED traffic signal lens from occluding the indication.
Pre-storm spraying with deicer was also shown to reduce the amount of snow and ice buildup, although the application life of the deicer is undetermined. Figure 7 shows a commonly used deicer product and the empirical results of an in-field test. The Wisconsin Department of Transportation (WisDOT) reported an experiment with Intelight's Automatic Lens Defrosters. This device includes a flexible heating element to thermodynamically heat the inside of the LED's signal casing, providing heat to melt snow and ice buildup. A heating unit sits inside the door of the LED signal lens and consumes 34 Watts as part of a 120-V AC connection. WisDOT reported difficulty with this device in non-Intelight products. The devices were never field installed. A review of the countermeasures identified found no device or application that proved to be completely effective at preventing snow and/or ice buildup on LED traffic signal lenses. Nevertheless, each method provided some reduction in the snow and ice buildup and improvement in traffic signal indication visibility.
Table 3 provides a summary of the primary countermeasures and their reported uses by the agencies included in this research. In discussing countermeasures with traffic engineers who are part of the National Committee on Uniform Traffic Control Devices (NCUTCD) and participate in the Transportation Research Board's Traffic Control Devices Committee (AHB50), most suggest that brush cleaning is the only countermeasure employed, and only when self-cleaning through melting is not timely. Table 3. Countermeasures Attempted by Responding Agencies
1With hot steam spray 2Has also experimented with Automatic Lens Defrosters Federal Aviation Administration CountermeasureAs a final note on countermeasures, other transportation-based agencies were explored for related information on their use of LED lighting and signals. Perhaps not surprisingly, the Federal Aviation Administration (FAA) has addressed the issue of snow and ice-covered LED runway and taxiway lighting through a specification referred to as an "arctic kit" or "heating kit." FAA Engineering Brief 67D, dated March 6, 2012, section 2.13 states:
The kit is a thermostatically controlled heater, designed to replicate an incandescent bulb, activated automatically when temperatures fall below 41 degrees Fahrenheit. The device uses a 20 to 45 Watt isolation transformer connected to heating elements, as shown in Figure 8. This device could be applicable to traffic signal LEDs with minor modifications and provide a heat source to prevent snow and ice buildup on the lenses. However, experimentation with existing LED traffic signal products is required to determine the trade-off in energy consumption and accelerated degradation and deformation of the LED signal lens with this potential countermeasure. Additionally, the impact and feasibility of operating a series of heater elements on battery backup power should be considered.
RecommendationsThe research results have led to the following recommendations:
Needed Research
PPT files can be viewed with the Microsoft PowerPoint Viewer. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
 |
United States Department of Transportation - Federal Highway Administration |
|
