In addition to the energy savings that can be achieved with LED traffic signals, their reduced power has another potential benefit related to maintenance and safety. If electric power is cut to an intersection because of inclement weather or some other emergency, short term backup battery power is feasible with LED traffic signals, unlike with conventional incandescent signals. In these cases, the signal could revert to a flashing red mode for several hours, indicating to drivers that they should stop before proceeding into an intersection. Such a feature has important safety implications (Lundberg, 1997b). In some cases, the relatively low power consumption of LED traffic signals is a disadvantage. In the city of Stockholm, Sweden, where a procurement of traffic signals using LEDs of all three signal colors has recently been completed, electrical "noise" in the power lines that feed the signals occasionally resulted in a signal being inadvertently switched on at an inappropriate time (Lundberg, 1997a). Load switching compatibility is another important power issue. The lower power and current of LED units can be incompatible with switching gear designed for incandescent lamps with higher power and current (Behura and Evans, 1998).
Because traffic signals are a 24-hour-per-day, year-round operation, and because there are so many of them in use throughout the U.S., power factor and total harmonic distortion are important concerns. Early LED traffic signal units had power factors less than 0.6 (Wyand, 1996). Through incentive programs mainly offered by utilities, for whom low power factors increase their costs in delivering electricity, products with power factors over 0.9 are presently common (Lundberg, 1997a). Utilities usually encourage devices with low total harmonic distortion (THD), to minimize noise on system lines. Low THD can be difficult to achieve with low power devices, so tradeoffs between power and THD may be necessary for efficient operation of LED traffic signals (Behura and Evans, 1998).
LEDs are sensitive to temperature changes in two ways: luminous output and peak wavelength. Currently-manufactured LEDs are rated for operating temperatures of 25 degrees C; at lower temperatures they produce more light and at higher temperatures, less. At -40 degrees C AlInGaP LEDs have an output that is 192% of the output at 25 degrees C; at 55 degrees C, the luminous output is 75% of that at 25 degrees C (Hodapp, 1997). This point is crucial when considering that traffic signals must operate year-round in warm climates where solar radiation produces high temperatures inside the traffic signal enclosure. The peak wavelength also shifts toward longer (red) wavelengths slightly with increasing temperature, but this shift is usually too small to be of significance for most applications.
Potential consequences
Due to these characteristics, and the relatively small amount of data about the range of conditions that can be experienced in traffic signal installations, unforseen reliability problems such as inadvertent switching on or off, lower life, or reduced luminous output may occur. Of course, these problems have significant life and safety consequences and any steps to minimize them should be taken.
Options and recommendations
One potential option for LED manufacturers is to develop LEDs that would be optimized for operation at higher temperatures than those currently designed. Other potential avenues that may be worth exploring include, for example, the possibility of venting the signal housing or including a heat sink, to decrease the interior temperature. This is unnecessary for incandescent lamps because their luminous output is not dependent upon ambient temperature, but it could be an effective approach for LED signals if other factors such as humidity or dirt depreciation could be controlled. Of course, information gained by studying the optimization of LED traffic signals could be applied to other potential applications of LEDs, including displays and architectural lighting.
Conway et al. (1997) point out that the unique characteristics of LEDs as light sources for visually critical applications perhaps could allow totally redesigned traffic control devices, which serve the same purpose and function as existing traffic signals but in a radically different form, such as a narrow-profile sign using multi-color LEDs rather than three discrete signal heads. While the likelihood of such ideas gaining acceptance even in the long term is low, a deeper understanding of LED technologies and of human visual needs in this and other applications could perhaps encourage solutions uniquely suited to the characteristics of these light sources.
The development of a research agenda and priorities for problem solving would be a valuable step in increasing the effectiveness and reliability of traffic signals and other visibility and architectural applications for LEDs.