Cambridge, MA – LED Streetlight Retrofit Project

Here’s a quick review of the Cambridge, Massachusetts, LED street lighting retrofit project. In early April 2015, I traveled there to review the installation and talk to Steve Lenkauskas, the city electrician, and Glenn Heinmiller of Lam Partners, the lighting design firm contracted by the city to facilitate the project.

SOLA interviewed Glenn Heinmiller and published this short video:


In 2005, the City of Cambridge purchased the streetlights on Cambridge streets from NStar, an electrical utility. The city already owned the lights in city parks and parking lots. In total, the city currently owns and maintains approximately 7,800 free-standing lights, over 3,500 of which are on wooden utility poles.

In 2008, the city commissioned a Lighting Study to review the inventory of current public lighting in the city, to develop standards for appropriate light levels for Cambridge streets, to evaluate energy-efficient light sources and luminaires for potential use, and to recommend a set of luminaire types to be used for street and park lighting. The city hired lighting consultants from PB Americas, Inc. and formed a Lighting Design Committee.

In 2010, Cambridge released the results of the study in the Cambridge Lighting Study Report. The report included results from a pilot test of LED fixtures, but concluded that it would be premature to install them at that time.

In 2013, Cambridge hired Lam Partners to facilitate a conversion of the city’s streetlights to LED. In consultation with Lam Partners, the decision was made by the city to include wireless adaptive controls in the design.

Phase One of the installation began in June 2014 and consisted of replacing 4,900 cobra-type fixtures with Cree LEDway type 2 short fixtures in 66-191 watt configurations. Each fixture included a Lumewave TOP900-TL adaptive control node, and were connected to the city’s infrastructure using Lumewave gateways and controlled with LumeStar software. Phase One was completed in March 2015.

Phase Two will retrofit the city’s remaining 2,000+ decorative fixtures using a variety of LED fixtures and the Lumewave adaptive control system. Estimated start date is summer 2015, to be completed in 2016.


Cambridge will become the first city in the U.S. to complete a city-wide LED streetlight retrofit using adaptive controls and dimming to match lighting levels to the IES RP-8-14 road classification for the specified time of the night. In addition, Cambridge has reduced initial illumination levels to 70% of maximum to allow for lumen depreciation compensation over the life of the fixtures. By employing reduced illumination levels for lumen depreciation compensation, more efficient LED fixtures and lighting to minimum illumination levels for individual IES roadway classifications, the city is currently realizing a 77% reduction in energy and CO2 over the previous High Pressure Sodium (HPS) system. This is resulting in a $500,000 annual reduction in energy costs.

It should be noted that this reduction is based on a renegotiated energy tariff with Eversource, the utility company, that is based on each fixture being operated at full rated output from dusk to dawn. The city is currently negotiating a further reduction to capture the energy reduction from dimming. When implemented, it should result in significant additional cost savings and an increase of the percentage of energy and CO2  being saved.

Eversource also gave the city approximately $820,000 in energy efficiency rebates for Phase One of the retrofit and will grant additional undetermined rebates for Phase Two in the near future.

The total cost of Phase One was approximately $3 million dollars, of which $700,000 or 23% of the project cost was for the adaptive control system. Using the dimming as described and the rebates, Cambridge will pay for the installation within 4.36 years. For the remaining 20+ years of the fixtures’ life, the annual energy savings to the city should save approximately $500,000 per year or about $10 million. In times of shrinking revenues, it is a powerful reason to consider using adaptive controls to their maximum ability.

Lessons Learned

During my meeting with Steve Lenkauskas, we discussed what lessons he could pass on to other cities considering retrofits. He credited his predecessor, George Fernandes, with making the choice to use adaptive controls. The primary reason for the decision was to improve maintenance and asset management. However, in 2010, the city also launched a Sustainability Initiative that led it to a Green City designation by the Commonwealth of Massachusetts. As part of the initiative, the city pledged to reduce CO2 20% by 2013 from 2008 levels. The LED streetlight retrofit project was approved partially in response to this effort.

Public reaction to the project has been overwhelmingly positive, but there have been concerns expressed by some residents about several factors. First, a number of complaints have been made regarding the perceived brightness of the new LED fixtures. While the illumination levels of the new fixtures are generally lower than the previous HPS fixtures, the increased visibility from broad-spectrum white light is perceived as brighter. This may also be responsible for the increased number of complaints related to light trespass. Steve indicated that the majority of the complaints were received during the first few weeks of the project, prior to commissioning, while the illumination levels for all fixtures were at 100% of output. After the fixtures were integrated with the adaptive controls, the fixtures were dimmed to 70% prior to 10:00 p.m. or midnight, and to 35% thereafter. He indicated that complaints were significantly reduced after the dimming schedules were implemented.

Steve said the major lesson he learned was that in the future he would integrate each fixture into the adaptive controls network as installed, so that they would be operated at the proper lighting levels from the very beginning.

Another byproduct of the LED fixture is its ability to better control the distribution of light so that it doesn’t spill where it isn’t needed. However, even with this increased control, the city has faced difficulties with the extremely short setbacks many of the residences have to the streetlight poles. In many cases, they can be three feet or less. Even using the lowest backlight fixture rating available, significant light trespass can still occur. The city electrical department staff has responded to every complaint and has made adjustments to the fixtures as needed. In some cases, the problem has resulted from fixtures not being installed parallel to the road due to tilted poles. In other cases, staff have had to install house side shielding which has provided very accurate beam control.

Steve indicated that by being responsive to the residents’ complaints, he has been able to resolve most of them quickly and amicably. Another lesson learned is that communication with the public is tremendously important in making the conversion a success.

The last issue of public concern is in regard to the color temperature of the new LED lights. While Cambridge did some pilot testing of the 4000k CCT that was selected, some residents have complained that the lighting is more glaring and that the cooler color changes the character of the community. A few residents have also expressed concerns about the possible impact of the new lighting on sleep and circadian disruption of humans and animals. The decision to use 4000k CCT was made by the lighting consultant and was based on a compromise between efficacy and color temperature. At the time of the decision, 3000k CCT was available by special order, but its efficacy was 20% less than the 4000K CCT option. As energy and CO2 reduction was a large part of the city’s requirement for the project, 4000k CCT was selected. Cambridge will specify 3000K- 3500K CCT for pedestrian scale fixtures in Phase Two.

The delicate balance between visual comfort and neighborhood character versus maximum energy efficacy remains an important element of a successful LED conversion. It is often the source of public dissatisfaction without being articulated. The public typically complains of increased brightness although actual illumination levels are reduced. Although more should be done, the research that has been completed links higher CCT with increased perception of glare and visual discomfort.

The efficacy delta between warm and cooler CCT continues to shrink. However this delta is dwarfed by the energy savings possible from dynamically adapting  fixture output to match IES minimum illumination recommendations for traffic conditions, and stopping the common practice of over-lighting.  Additionally, cities should not underestimate how dramatic the shift in color from HPS, at about 2000K CCT, to LED will appear to the public. Residents who had become comfortable with the warm yellow glow of HPS are not all eager to see their neighborhoods bathed in the same color light that they might park under at the local shopping center.

(See news stories on neighborhood brightness concerns)

Should the decision to use a higher CCT be made solely on energy efficiency grounds, when much greater energy  savings are ignored? For the sake of a smooth transition and public satisfaction, the warmer 3000K CCT option with adaptive controls may be the best solution.


Cambridge has successfully done what few cities have attempted: a LED conversion using dimming that maximizes energy efficiency and carbon reduction. Some cities have experimented with pilot tests, but Cambridge fully committed. As a city of about 100,000 residents, it may not draw the same level of attention that New York City (NYC) or Los Angeles (LA) does, but it should.

Both NYC and LA were not willing to “take the risk” of lighting their streets to the recommended minimum illumination levels of the lighting standards that they profess to follow. LA installed 140,000 and NYC has announced that they will install 250,000 streetlights next year, all without dimming or adaptive controls. In April of 2014 during a meeting with the NYCDOT, I was told that they would not consider dimming due to potential public safety concerns. They also said that they don’t get citizens complaining that the lights are too bright, if anything, they want more light. When queried regarding the additional energy and CO2 reduction possible using dimming, they said that the 30% projected savings were more than enough.

Interestingly, LA recently announced that they will upgrade all of their streetlights with Philips City Touch controls. However, they still have not announced plans to dim the lights for constant lumen output or to meet RP-8 recommended minimum illumination levels.

The big takeaway from Cambridge is that you can achieve a 77% or more reduction in CO2, energy, and costs with no compromise in public safety by using current industry minimum illumination standards already in place. There is no legal risk from doing so and not pursuing maximum savings due to the mere threat of frivolous lawsuits is bad public policy.

If the Federal government is really serious about climate change and energy conservation, it needs to pass legislation to indemnify cities that are willing to implement LED retrofits with adaptive controls and dimming. Can we really afford to watch city after city follow the example of NYC and LA and forego half the potential energy savings due to fear of litigation over the next decade? The fixtures being installed now may easily last 20 years or more. The labor costs to backtrack and install adaptive controls later will be more difficult to justify, and will easily negate several years of the potential energy savings.

At an absolute minimum, every city should plan to install adaptive controls to allow for the initial reduction to compensate for lumen depreciation. This feature alone will pay for itself in energy savings, and the 20-50% increased fixture life that running them at reduced current will produce. By installing them at the same time as the fixtures, it will be possible to make the transition to dimming at a later date. It will quickly become obvious that the increase in perceived brightness from broad spectrum white light will allow the use of RP-8-14 minimum illumination levels throughout the night. We must get past the antiquated idea that more light makes the public safer. Overlighting is a luxury that we can ill afford, as it comes with a much higher price tag than just that of wasted energy. LED white light will appear brighter and visibility will be improved; now is the perfect time to readjust our illumination levels and faulty assumptions regarding safety.

Thanks to Cambridge for being a pioneer of smart urban lighting best practices and to Steve Lenkauskas and Glenn Heinmiller for sharing the details of the conversion. If you have any questions or comments please post them below.





12 thoughts on “Cambridge LED Streetlight Retrofit Project”

  1. FYI-I live in Cambridge, and there have been a lot of problems with the LED street lights. Your article seems more like an advertisement, than an accurate report of the retrofit project in our City. Our poor City Council has been swamped with complaints about the lights, and the public reaction is not as positive as you express in the article. The City did not do enough testing before they purchased the lights, and we now have a terrible night time glare problem on the streets in Cambridge.

    1. The glare and light trespass complaints by the public are primarily the result of two factors, the high color temperature of the LED fixtures and Cambridge using IES RP-8 recommended lighting levels. The situation is compounded by the fact that many of the narrow streets have houses that are set back only 3 or 4 feet from the street light fixtures. Even if they had used warmer CCT LED fixtures, the poles are positioned so close to the houses that most fixtures could not be mounted in a way that could direct the light away from the windows of the houses. Unfortunately the minimum recommended illumination levels specified by IES are too high for residential neighborhoods when white LED is used. Cities are often required to adhere to these standards by their legal departments. So until IES revises these lighting recommendations, cities are using too much light. I do commend Cambridge for at least dimming by 50% at 10 pm to meet the lower recommended levels for less traffic, most cities are leaving the lights at the same lighting levels required at rush hour for the entire night. So while I agree that Cambridge’s retrofit is not perfect, they tried to balance reducing energy and CO2 with standards that have not been revised quickly enough. The good news is if and when IES reduces their recommended lighting levels, the city has the controls installed that will let them dim the lighting more. Unfortunately, the 4000K CCT color temperature can’t be reduced without replacing the fixtures. SOLA recommends that residential communities use <3000K CCT to maintain a warmer neighborhood ambiance. Thanks for your feedback.

  2. You say that savings from lower colour temperature lights are less than those of higher colour temperature. Dimming is not the answer. If you are really concerned about savings, then instead of dimming, why not switch them off altogether after 11p.m. until dawn when most people are not around? Curfews on lighting used to be standard practice until 1969, and are now being re-introduced in some areas, ostensibly to save money. After re-introduction criminality has declined by as much as 50%, ergo criminals need light. Residents would also sleep better as light intrusion onto properties will be eliminated. The onus should not be on residents to buy thicker curtains or take other measures to reduce the effects of light trespass. The polluter – namely the local authority – should therefore institute measures to reduce the problem.

    Ideally, the colour temperature should be around 2000K and should not exceed 2700K, in order to minimise environmental effects.

    1. Colin,

      If we could wave a wand and make things perfect we would have to start again in 1890. The outdoor lighting that we have today is the result of a century of public policy that has trained the public to believe that more light will make them more safe. So like it or not, changes today have to accommodate the wishes of a voting public that is afraid of the dark. If you try to radically change lighting policy you will incite these citizens to petition their government to increase, not decrease lighting.

      After 20 years of working with communities and their elected officials, I have found that gradual change is the only path to success. If we use new control technology to adapt lighting levels to meet current recommended practices for the volume of traffic thoughout the night, we can reduce total lumens by 70% or more. If we ask the public what CCT they want, we’ll use predominantly 2700K. If we can show that improved visibility and the quality of light has more impact on safety than the quantity, the public will embrace better lighting.

      SOLA best practices recommend quality fixtures that deliver no uplight, no glare and no light trespass. Our mission is to educate stakeholders of the value of Community Friendly Outdoor Lighting. We’re in it for the long haul. Over the last two decades I’ve watched cities and states change their lighting policy to use predominantly fully shielded lighting. Nothing happens overnight. Like erosion, it takes a sustained effort over time to change something that evolved over a hundred years. Education and raising public awareness is the only tool that absolutely works. We need to support these organized efforts and know we’ll affect change. The more outreach, the faster it will happen.

    1. Not the whole country, but Paris was one of the first citywide lighting curfews in the world. They also have resisted the conversion to white LED based on the issue of ambiance.

  3. Bob,

    The essence of good governance is to know what the people want, and to know what the people need, and to have the wisdom to understand the difference. There has got to be a universal culture change in our attitudes to outdoor lighting, so I stand by my original comments. The needs of the environment demand it.

    1. Well I think we can all agree that good governance is in short supply in a lot of areas. My point is that you first need to educate the stakeholders. That doesn’t happen overnight. Then the transition can start. Using adaptive controls helps cities start the process of reducing lighting levels without scaring the public. The public’s fear of the dark is a strong driver and it can’t be ignored. We need to unlearn 100 years of training that more light makes you more safe.

  4. Education is important, but some people either think they know it all, have a different agenda, or won’t accept anything.

    If LP increases at 10% and you go for a 1% improvement, you still loose and exponentially so just like credit card debt. Einstein was the one who said “compound interest is the most powerful force in the Universe.”

    Cambridge, MA went for a 30% reduction in the number of streetlights out of necessity more than 15 years ago and got it. They also went for 28-watt CFL which is far better than what is being done with LED. Other New England towns just turned the light OFF. Cars have headlights, and that should be enough. There is no need to drive over the speed limit. If you want to CYA at home, get a dog.

  5. In my previous message I said Cambridge reduced its lighting by 30% when it should have Lexington.

  6. Instead of dimming, in residential and suburban areas, low colour temperature LED streetlights can be motion operated. The lights will only come on when someone is in the vicinity of the light. The lights remain on as long as it takes to reach the next light, so the person is never out of light. In this way the public’s fear of crime and darkness can be overcome. It’s not rocket science and I know for a fact that at least two lighting manufacturers are considering that option. When combined with the lighting curfews I mentioned previously, savings will be maximized without loss of amenity. Public outreach and education are required in order to convince people of the rectitude of this policy. So go for it.

Comments are closed.

Reducing global lumens