It all started in 2008 when the City of Anchorage called wanting to create new streetlighting guidelines. Our process has always begun by asking A LOT of questions. Questions like “What streetlights do you prefer?” and “Are there areas that have high accident rates at night?” We discovered that, overall, the existing 400W HPS streetlighting produced very high lighting levels, and yet, accident rates were also high. While pondering this question, the idea of demonstrating different technologies, including the new up and coming “LED” street lighting technology was formed.
We started with a demonstration project in a residential area with different manufacturers’ products. On one very cold night in the Alaskan autumn, we invited the public to partake in a series of organized surveys. The public response was fabulous! Through our surveys, we discovered that the citizens of Anchorage preferred the LED “white” light, but at much lower light levels than before! Overall, we were very surprised at the results. As researchers, we questioned if we did something wrong, and whether or not we could duplicate the results in a commercial area of the City.
With the City extremely excited about exploring the new “LED” technology and the prospect of reducing energy use (from 400W HPS to 100W LED), they agreed to fund another study. This time, Clanton & Associates proposed to team up with Virginia Tech Transportation Institute (VTTI) to conduct the survey. VTTI came equipped with instruments to measure reaction times on object detection in addition to our subjective surveys of the public. The experiment was repeated in a commercial district and we came up with similar results! In addition, the object detection distances showed that using the LED “white” light and lower lighting levels significantly increased detection distances. This all seemed so counter intuitive, since the philosophy of streetlighting has typically been “More light … better sight”. We were proving the opposite that “Less light … better sight”. The City of Anchorage was very pleased with the results and decided to change out all of their 400W HPS street lights to the much lower wattage LED lights. In addition, they installed automatic dimming systems that dimmed lighting when snow was on the ground and during low activity level times.
Answers Always Lead to More Questions
Our research team had many ongoing questions such as: Did the snow on the roads affect our results? Did on-coming headlights make a difference? What is the influence of white light versus HPS orange light? How did glare impacts enter into the results? So, we decided to head south to the forever summer conditions of San Diego, CA.
In San Diego, the Clanton team and VTTI repeated the experiment with some “improvements”. Yes, there was no snow, only Palm trees lining the streets. Also, we closed the streets to traffic to avoid the possible “head lamp” influence of on-coming traffic. What we didn’t account for were the police warning lights flashing as they blocked traffic from entering our test roadway. We continued and repeated the Anchorage experiment with object detection and subjective evaluations. Once again, the public opinion and detection distances were significantly better with “white” LED streetlighting versus the HPS lighting, and with significantly lower light levels.
More Data… Always More Data
Did we have enough data? Well… as researchers we always ask “What if?”, so we wanted to repeat our experiments one more time without the police car warning lights as a distraction. On to San Jose!
As luck would have it, we soon got our chance. The City of San Jose hired Clanton & Associates to write new streetlighting guidelines and all of the same questions were asked. The surprising answer to our question about high accidents showed that, at intersections where an amber low-pressure sodium (LPS) streetlight was viewed by drivers next to the traffic signals, many drivers confused the red stop light with an orange warning light. As a result, there were many intersection collisions due to this visual misperception.
The team chose a remote area of town in a business park where streets were easily closed during the evening. No snow, closed road, and no police warning lights… all lessons learned from our first two cities. Since the streets were isolated with minimal ambient lighting, we had as close to a “perfect” experiment as we could ask for. We varied the color of the lighting (3500K to 5000K), had different dimming levels (100% to 50%), and performed the same object detection and subjective evaluations.
Results came in… the same as the first two cities! White light and lower light levels not only increased detection distances but also were preferred by the public.
Are we finished? Do we have enough results? Well … once again we kept asking “what if”. Our opportunity came when the Northwest Energy Efficiency Alliance (NEEA) approached us in 2013 about repeating the experiment in Seattle (City number 4) with more variables. For this experiment, we varied the light distribution of the test areas and also wanted to see how “low” we could go with light levels (down to 25%!). Our setting was in an urban area with a higher street presence ambient lighting from the store fronts and bars. No snow, closed road, no police lights, and some ambient lighting to simulate typical urban atmosphere. One more variable… dry and wet roads were tested since it rains a lot in Seattle.
This time, we wanted a lot of people taking the surveys and riding in the test vehicle to gather detection distances. Our goal was to have balanced gender and variable ages in each test situation so no one would go through the test more than once. The variables for each unique test group were light levels (100%, 50%, 25%) and different pavement conditions (dry and wet). We solicited volunteers through social media and had them sign up for a time to balance out the gender and age diversity for each test. We were able to get more than 300 volunteers! You may ask how we did it? We paid them! The neighbors used this experience as fund raisers for schools, churches and sports activities.
Since we were pushing the limits of the variables especially with lighting levels and lighting uniformities, we expected that the results would taper off. To our surprise… they didn’t! In fact, the 25% light level, white LED lighting, and poor uniformities produced significantly higher detection distances. The walking public did not like the poorer uniformities on the sidewalks which gave us all a moment to reflect on how we should handle streetlighting differently from pedestrian sidewalk lighting.
Changing the Way We Light Our City Streets
The big loser in all of the studies were the existing 250W HPS streetlights with the lowest detection distances and public preference. Not surprisingly, 250W HPS is the most common type of streetlight in the majority of cities.
A Tale of Four Cities produced some of the most valuable results for roadway standards in the future. Ongoing studies with other cities such as Salt Lake City are answering even more questions concerning environmental impact of lighting, and further citizen preferences.
Our team did make a difference. We are now involved in the IES Roadway Lighting Committee and are questioning the standards criteria such as “Are roadway lighting levels too high?” and “Uniformity ratios need to increase”. Ideally, spectral contrast-based standards will greatly increase visibility of objects and pedestrians. More experiments!!! These changes have the potential to save communities a lot of money and energy use by reducing lighting levels and decreasing the quantity of streetlights. Soon, streetlighting will become obsolete with autonomous vehicles and all lighting will be for pedestrians and cyclists.
The big question we have now... What’s next???