Text-Alternative Version: Pedestrian-Friendly Nighttime Lighting Webinar

Below is the text-alternative version of the "Pedestrian-Friendly Nighttime Lighting" webcast, held November 19, 2013.

Linda Sandahl: Welcome, ladies and gentlemen. I'm Linda Sandahl with Pacific Northwest National Laboratory and I'd like to welcome you to today's webinar, "Pedestrian-Friendly Nighttime Lighting," brought to you by the U.S. Department of Energy Solid-State Lighting Program. We're very happy to have as our speaker today Naomi Miller of the Pacific Northwest National Laboratory. Naomi Johnson Miller is an engineer in the Solid-State Lighting program here at Pacific Northwest National Laboratory, focusing on lighting quality, energy efficiency, sustainability and the acceleration of LED technology deployment.

Before joining the lab, Naomi was principal of Naomi Miller Lighting Design in Troy, New York. She is a fellow of the Illuminating Engineering Society of North America and the International Association of Lighting Designers. She holds a bachelor of arts degree from Massachusetts Institute of Technology and a master's degree in lighting from Renssellaer Polytechnic Institute.

Naomi?

Naomi Miller: Thank you, ladies and germs. We're going to talk about nighttime lighting now, concentrating on LEDs but also talking about other technologies.

First off I have to start with some caveats on the talk we're about to go through. First of all, we're not talking gospel here. This is mostly a set of observations and collected wisdom on several pedestrian projects. So we're not trying to espouse principles that everyone will use on every project. It's full of evolving ideas. It's based on mockups, it's based on observations and feedback from multiple pedestrian sites, and it relies heavily on work that is not my own but work by Rita Koltai of Koltai Lighting Design, Terry McGowan of Lighting Ideas and Dr. Bill Neches, who is the head of the Chautauqua Property Owners' Association in Chautauqua, New York.

This couldn't have happened without input and products from a large number of manufacturers, so we are extremely grateful for their participation. And finally, it couldn't have happened without the Department of Energy funding the GATEWAY Demonstration program, which gave us the ability to look at all these different projects and try to reach some conclusions.

So let's talk about conventional outdoor lighting and how it has been traditionally approached. For very good reasons, the focus has been on disability and environmental impacts of outdoor lighting. So issues you're very familiar with are illuminance on the pavement, or in the case of roadway lighting luminance on the pavement, uniformity of light on the pavement, usually a maximum to minimum ratio for luminance, or average to minimum values, depending on what kind of spaces you're looking at.

Vertical illuminance on faces and targets, and by that we're talking about how much light is falling on the face of somebody who is walking towards you, so you can recognize them. Or how much light is falling on a ball that is bouncing out in the middle of the roadway so that you can see the ball, see to avoid it, and hopefully avoid the kid that's charging out to get the ball.

We also have put focus on pole spacing because if you don't have to use as many poles, if you can space them farther apart you save money on a project. And pole spacing is important because it works hand in hand with the uniformity of light that's on your ground or your pavement. We also care about cutoff or BUG system ratings for dark skies considerations and for light trespass and for glare. And we have cared about efficacy, which is basically lumens per watt of the lighting system that's installed. And those are all very good goals.

What we'd like to introduce here is that when you have pedestrians involved the goals may shift a little bit and maybe the order in which these issues are addressed will shift a little bit. So pedestrians: very concerned about making sure they're not tripping and falling on uneven pavement, or walking off a curb, for example.

They're worried about being hit by bicycles and cars and skateboarders and all kinds of vehicles. They're trying to make sure that they are not in harm's way, that if they see people that make them nervous, or if they see critters like alligators, that they can avoid them.

Residential windows. You don't really want to be sleeping in a bedroom with a lot of light pouring in from the streetlight outside because it keeps you awake at night. Appearance of the neighborhood—well, gosh, you know what? Light fixtures and the poles that they're mounted on carry baggage, stylistic baggage. So depending on the neighborhood, who lives there and what the appearance of the houses is, for example, you may want to have different looks of fixtures involved.

And glare. There are at least two flavors of glare that we'll be talking about today, maybe a third one too. Discomfort glare is glare that makes you uncomfortable but doesn't necessarily reduce visibility. But disabling glare is actually an uncomfortable brightness that reduces your ability to see, or to see contrast in the visual field. And both of these affect adaptation. You know that your lighting system adapts according to the range of luminances that are in your visual field, and it tends to adapt to the brightest thing in the field of view.

Now let's talk about safety from tripping and slipping and falling. This also affects visibility when you're outdoors. The angle of light that is on the path ahead of you enhances the contrast of the tripping hazard. So you want to make sure that the tripping hazard, let's say it's a tall rock that's in the middle of your sidewalk, you want to make sure that either the rock is lit brightly with respect to the background behind it so that it should be brighter than the dark pavement behind it, or conversely, it should be darker than the pavement so that you're seeing the contrast between that object and its surrounding, that shadow, that highlight and shadow combination makes objects visible, or uneven pavement, for example.

We care about illuminance uniformity along that path because you want to make sure you don't have really dark patches of pavement where you wouldn't be able to see the unevenness of the pavement. You want to make sure there's enough light spilling over the edge of the path. It doesn't have to go very far, but if you look at the photograph that's on your screen right now there is a bright reflection of those gumball fixtures in the distance on the wet pavement. And you can't see details very easily in that wet pavement because the wet pavement turns into a mirror. So you're seeing actually a reflection of the light fixture; you're not actually seeing the pavement. But because the grass on either side of the pavement is less specular, less shiny, you're able to see the edge of the pavement. And just seeing edges helps the pedestrian navigate safely along the path.

Disability glare. We want to control disability glare because that bright spot of light in your field of view shifts your adaptation range, which means that your eyes are adapted to that bright spot of light and your visual field, depending on the individual and exactly what their adaptation is, they usually can't see more than a range of about 1,000 to 1 in luminances within that visual image. So anything that's brighter than the high end of that luminance range gets blown out; it turns white. And everything that's lower than about that 1,000 to 1 luminance range turns dark. So just like a camera needs to—you need to change the F stop so that your luminance range matches the film, just like that camera adaptation you need to do the same thing with your visual system to make sure that the important details that you're looking for are within that 1,000 to 1 luminance range.

Disability glare also superimposes a veil of light across the visual field. And if you just think of it as a smear, think of it as a light-colored smear that gets thrown across your visual image, that is washing out your ability to see contrast in your visual field. So it's reducing the contrast.

Other issues that pedestrians worry about: personal security from harm and intimidation. And I have a couple of slides here, couple of photos of someone standing not very far from me, and depending on the lighting pattern in the space you can either see their face or you can't see their face. On the upper right you'll see that those fixtures are spreading a lot of light fairly evenly and at fairly high angles so you can see the face of the person that's walking towards you.

In the photo at the bottom of the screen the walkway is lit with bollard fixtures; they're full cutoff bollards, so they're directing the light downward only. So there's very little light bouncing against the surface of the pathway back onto the face. And consequently you're seeing that face essentially in silhouette. And if the background is black also you're not seeing enough contrast to identify the facial features. So it's the pattern of light that helps determine how much light you're going to see on faces and bodies of people around you.

This is important because we all have concerns about individuals out there, whether they are potential foe or potential friend, whether they're alligators that are going to nip at our heels or scary dogs, for example. We have concerns when we're walking outside. So being able to identify the face and the body or the clothing of the alligator that's walking towards you is very important.

When I was a baby lighting designer I heard from Dr. Peter Boyce that there were two principles for the perception of safety. And those two principles have stuck with me for a very long time. One is you want in your visual field, you want in your area of view to be able to see at sufficient distance to identify danger that may be coming towards you in enough time to react to it. So that distance is related to how much time it would take for you to identify the danger, turn, and react. Now whether that means walking in a different direction, for example, there are many different kinds of actions that you can take. Or maybe it's just being more aware as you approach that individual.

The second principle is once you identify potential danger, seeing where to go for safety or refuge if you need to find safety or refuge. So it's being able to see at enough of a distance around you so you know where to go for safety if you need to make that reaction.

Other pedestrian-focused issues. We talked about unwanted light in residential windows. Backlight from a streetlight can be very annoying, and usually that backlight is the light that's emitted from 60 degrees up to about 90 degrees from the luminaire. So if you think of the luminaire nadir as straight down from the luminaire's center and you go backwards towards the apartment building usually what's pouring into bedroom windows and is problematic is emitted from 60 to 90 degrees.

Next issue: appearance of the neighborhood. Now we mentioned that luminaires carry baggage, stylistic baggage. So the style of the luminaire makes a difference, but also the pattern of light on the grounds makes a difference, and the pattern of light on the building. Sometimes if there's a really funky pattern of light falling on the building it's kind of distracting and ugly. Another issue is the color of light. We're going to talk about that pretty extensively today when we go through our examples.

Now let's get into glare for the pedestrian. Discomfort glare. Usually when we talk about discomfort glare for the driver we're talking about discomfort glare coming from the luminaire emitted from about 75 degrees up to 90 degrees. So why is that so high? It's because you're in a vehicle that's moving. And you're in a vehicle that has a hood on the car, unless you're driving a convertible at night, in which case you've got some other problems to worry about. So if you have a hood on your car that is automatically providing a visor on your eyes, just like wearing a baseball cap. So you're seeing the light from a distance; usually the glare angles are 75 to about 90 degrees from the luminaire nadir.

It's different if you're a pedestrian. For one thing, unless you're on a skateboard, you're moving a lot more slowly than you are in a car. So you're exposed to that light fixture for a longer period of time and you are underneath the—for most of the time you are close to the luminaire and the luminaire that's above you us causing discomfort. Even though you may not be able to see the luminaire if your eyes are aimed straight ahead, parallel to the ground, as a pedestrian you will normally move your head around to look at streetlights, to look at trees, to look at the moon, to look at clouds, to look at—well, I hope you're not peering into bedroom windows but you may be glancing at walls and upper story windows and roofs and birds and you name it. You are never keeping your direction of view on the ground 100 percent of the time. You are naturally glancing around.

And as a consequence you are also looking up at areas of brightness, which do tend to draw your eye. And those areas of brightness can very easily be the light fixture itself.

The pedestrian glare angles are primarily zero to 75 degrees. When you're at quite a distance away from the luminaire, yes, you may also be distracted and find glaring, the light fixture, when you're in the 75 to 90 degree zone, just like the driver is but for the vast majority of the time that you're near the luminaire you're in the zero to 75 degree zone.

Now let's talk about some examples of disabling glare. Disabling glare, as we just said, is the glare that scatters light in your visual field. And the two images that you have on the bottom of the screen are illustrating that. They've been simulated; it's very difficult to simulate exactly how the eye sees, but this gets very close. So if you look at the image on the right you'll see a young couple that's walking towards you from the supermarket and there's a light fixture above their heads. If that light fixture is not sending a lot of high angle candlepower into your eyes then you can see the object—or excuse me, the two people walking towards you.

But if there's a lot of glare from that light fixture, if there's a lot of candlepower coming straight to your windshield and straight through to your eyes from that light fixture it superimposes a smear of light over the visual image. And you can see the image of that young couple is washed out by that disabling glare. This also affects your adaptation level, so your eyes will adapt to the brightness of that light fixture, which also means that the young couple are going to be more difficult to see.

Here's a great illustration that also came from the International Dark Sky Association. Here is an individual looking towards the back fence, and there is a wall pack on the building to the left. So there's that very bright light source that is producing glare in the visual field of the person who's looking at the back fence. Now let's put up a hand and let's block that glare. Now essentially your visual system is adapting quite rapidly to a lower range of luminances because that very high luminance is no longer in the field of view.

Now as you see there is a person standing at that back fence and your visual system is able to accommodate the range of luminances that bring out the detail of the individual standing at that back fence. So let's go back to the earlier image just to prove to you that that individual was standing there the whole time, but you can't see it because your eye has adapted to the brightness of that wall pack fixture. That's why we are concerned about glare.

So what kind of places might we think about putting pedestrian-friendly lighting? We don't want to put it everywhere; it's not appropriate everywhere. Don't get me in trouble by writing to me and saying that you have lots of spaces where this won't work. I know it won't work everywhere.

But in some places, like summer camps, clubs, retreats, cultural institutions, those are good places to think about pedestrian-friendly lighting. College campuses, places where there's a lot less vehicle traffic, for example, a lot more pedestrian traffic. Private schools, boarding schools, places again, campuses where there are very few vehicles. Parks, cafes, outdoor festivals, again, places where if there's traffic at all it may be moving very slowly. Quiet neighborhoods where neighbors know each other, where neighbors spend time outside, where they walk their dogs in the street, and possibly where crime is less of a concern.

There are places where crime is a big concern. And in those places pedestrian-friendly lighting may not be the cup of tea that's appropriate. But where crime is not the number one concern think about pedestrian-friendly lighting.

Here are a couple of places where we looked at pedestrian lighting. The first one is Stanford University, and the second one is Chautauqua Institution. Stanford is located in Palo Alto, California, or actually in Stanford, California, which is next door to Palo Alto. When Stanford hired Rita Koltai of Koltai Lighting Design to advise on options for pedestrian lighting they expressed their interest in the pedestrian lighting on campus. They really wanted soft, warm, lighting that was better in color rendering than their existing 100-watt high-pressure sodium lighting. You can see the photo of the main campus buildings there; it's a warm, colored stone on most of the buildings in the campus, red, Spanish tile roofs, very warm-colored kind of environment.

Stanford wanted to reduce energy use, improve the campus appearance and reduce glare for pedestrians. Pedestrians are on this campus almost 24 hours a day. Even in the middle of the night there are laboratories that are functioning, computer centers that are open, all kinds of activities going on, even in the middle of the night.

Stanford was interested in unifying the fixture appearance on the campus, and also their residential neighborhoods. And they wanted to maintain a somewhat traditional style of luminaire. They wanted, wherever possible, to use existing poles and spacing of the poles because that saves money.

So here's their existing lighting. It's a 100-watt, high-pressure sodium lamp in a glass refractor, utility pedestrian fixture. These fixtures have been installed and I'm going to make a guess here: 40 years, 50 years? A long time. The high-pressure sodium, the lumens emitted from the luminaire, produce 51 lumens per watt. Now that's not the light source efficacy; that is the luminaire efficacy and that's the number that we're going to try to use throughout this presentation to make it consistent. And the high-pressure sodium lamps last about 24,000 hours, or a little bit longer, in general.

Now let's talk quickly about some options that didn't work, and then we'll moved into the options that made it to the final round. First of all they looked at some full cutoff LED lanterns that had a traditional appearance but open sides. So instead of having luminous glass refractor on the sides it was open on the sides, and all of the optics are in the hood of the luminaire.

The lamp used was a 3000K 100-watt ceramic metal halide lamp, clear lamp, and the—as you can see, the aperture was sealed with a clear glass, flat glass. 110 watts total, 50 lumens per watt is the fixture efficacy. They tried mocking up a whole range of fixtures along one particular roadway, pedestrian roadway, and this is one of them that produced extremely harsh shadows on the ground. You can see the pattern cast by the top of the pole from that luminaire. This is what happened when you had a clear arc tube lamp and clear glass below the lamp producing those very sharp shadows. And the facility people and the individuals who filled out surveys did not like the strong shadow cast around the base of the pole.

Let's talk about some other options that didn't work. They tried looking at the existing high-pressure sodium luminaire, and reusing the luminaire, pulling out the existing high-pressure sodium lamp and ballast and reworking it to accept LED options. Some of them looked like corn cob lights; some of them had an interesting configuration of LEDs. But all of them reused the existing glass refractor.

These lamps ranged between 50 to 70 watts and that's including driver. They tried three different models. Unfortunately they produced a light distribution on the ground that was not as good in this particular optical system, and observers found the pattern of light on the glass to be very glaring because the LED products that were put in as retrofits had vertical rows of very bright LEDs. And those vertical rows combined with the glass refractor produced some very bright, and frankly ugly stripes on the glass refractor. So that dot pattern, or the stripes, were rejected.

Now that's not to say that this couldn't have been corrected with some frosting on the interior of the glass, but this iteration didn't end up being successful, except in one respect: they tried 4000K units, 3000K units and 2700K units, and they learned that 4100K was rejected by the observers because it was considered too cool and too bright; 3000K units were good but still too white. The 2700K units set exactly the right tone for this campus. So it was a bit like papa bear, mama bear, and baby bear: 2700K was just right.

Now let's get into the options that did work. Option A was reusing the existing fixture, changing out the lamp and ballast to a 60 watt—it's called a Cosmowhite ceramic metal halide lamp. It's a 2800K lamp, 70 CRI, producing 67, or drawing 67 watts and producing 67 lumens per watt in fixture efficacy.

The facility people very much liked the color of light, and this gave direction on the subsequent effort for looking at options. However, the lamp in the vertical orientation only produced about 18,000 hours of rated lamp life, so it paled in comparison to potential LED light.

Option B was looking at a totally new luminaire head. So pulling off the existing Holophane RSL-350 luminaire head and replacing it with another Holophane product called the PUL, it was a utility series full cutoff luminaire. So all of the optics are in the head of the luminaire.

Using 70 watts, 3000K, 80 CRI LEDs this fixture was never born with conventional lights in it; it was—or conventional lamps, I should say—it was born as an LED luminaire. It had flat glass in the luminaire, and that flat glass, unfortunately, originally produced similar sharp shadows to the metal halide product I showed you a couple of slides back. So that flat glass was changed to a diffuse glass instead, and that eliminated the sharp pole shadows on the ground. 63 lumens per watt fixture efficacy, estimated 50,000 hours life or more.

Option C, again, looking at the existing glass refractor, but instead changing out the optics of the fixture to a brand new set of optics that were dedicated to LED. This is a Holophane retrofit product with 50 watts, 3000K LED module built into it. The glass refractor, the original glass refractor, was retained. That's an old, glass refractor that's been around for decades. It's extremely rugged, and that product stays very clean, it's very easy to clean so it was retained for this mockup.

The light distribution on the ground was no worse than the original high-pressure sodium. The glare was considered very acceptable. It looked the same as the original product, so that was a definite strike in favor of it because people tended to like the original product. 62 lumens per watt fixture efficacy, long life.

Now, did I mention that the interior of the glass was frosted? Maybe not. If I did forgive me, I'm getting old and my memory is evaporating. But the luminaire, when it was mocked up originally with this kit, there was a lot of glare on the glass refractor from the LED kit, that the LEDs are now on the hood of the luminaire.

So the manufacturer suggested taking out the glass refractor and sandblasting the interior of it, which would frost the inside and essentially spreads the brightness of the LEDs over a larger area. That was done, and that was a very successful approach.

Everyone likes this option, except that 3000K was just a tad too cool. So could the color be warmer? Yes, it can. Now this is what the LED product looks like. This is the full retrofit head. The only thing that remains the same is that glass refractor.

Let's go to option D, which is exactly the same, with the frosted interior of glass but with a 2700K LED module. The glare, again, was acceptable, the color tone was perfect for the campus, the 2700K LED is not as efficacious as the 3000K LED, so the efficacy dropped to 57 lumens per watt, but LEDs – very long life. So pending a life cycle cost analysis this is probably the way Stanford is going to be going with their pedestrian luminaires on campus.

Now let's talk about the neighborhoods. There are some neighborhoods immediately around the Stanford University campus that are—with houses that are owned by Stanford faculty and staff but they don't actually own the land; they lease the land. So while you're on the faculty of Stanford you're able to live in these houses. The neighborhoods are full of houses that were built in the early 1900s, and several different fixtures were mocked up in this neighborhood.

Here are a couple of them that were mocked up and they were rejected. And the main reason is although people liked the lighting quality very much they did not like the contemporary look because it didn't match the age or the style of the houses that were in much of the neighborhood.

But they did like the warmer colored lamps, 3000K and 2,800 colored lamps; they just wanted a traditional appearance on their luminaire instead.

So here is the solution that is the best so far for these neighborhoods: replacing the existing luminaire, which is in poor shape right now, with full cutoff LED lanterns with open sides, and it has a diffusing glass on the LEDs. 70 watt luminaires, 3000K LEDs, 59 lumens per watt, fixture efficacy.

The diffused LED matrix produced tolerable glare and the color worked well for the neighborhood. Notice the shadow around the base of the luminaire, no sharp shadow because of the diffusion on the LED glass.

Now let's move to a different pedestrian-focused neighborhood. This one is in upstate New York, along Lake Chautauqua, which is one of the Finger Lakes that never quite made it as a full finger.

Chautauqua was founded in—I believe it's 1875, as a place to train Methodist Sunday school teachers, and it grew up as a tent facility and now it has grown into essentially a cultural summer camp with symphony programs, music programs, arts, theater. It's a wonderful place to come for lectures every morning. There are themes every week that are political themes or environmental themes or historical themes. Fascinating place to go.

Generally people go for a week at a time. It's a vacation spot; you got for a week, except that you don't have to put your brain on hold for a week. There's lots to keep you entertained culturally and intellectually for that week, as well as recreation.

The grounds of this Chautauqua Institution have very dense housing and incredibly cute streets and plazas. Vehicles are very much discouraged; it's a pedestrian community. Pretty much anywhere you need to go on the grounds are within a mile's walk. So it's very quick getting around. There are bicycles and pedestrians absolutely everywhere. It's a community that is very environmentally conscious, they're bat conscious—and by the way, that keeps the insect population down—they're critter conscious, they're dark sky conscious, and everybody is very concerned about making sure it's easy to sleep at night so they can bounce out of bed refreshed the next morning to go to their art classes.

This is what you see driving into the grounds of Chautauqua Institution. It's a sign that says, "Listen, folks, if you're a car you are not welcome, except as dropping off pedestrians or groceries. Get the flock out of here as soon as you can. You're restricted to 12 miles an hour." So it's a sign that warns everybody that this is a place where pedestrians rule.

The existing utility-supplied poles and fixtures on the grounds of Chautauqua are deteriorating at this point in time. They consist of older mercury post-top fixtures, and believe it or not, incandescent fixtures, even for roadway. And the goal of the street and pedestrian lighting here at Chautauqua is to provide soft, warm lighting and to minimize glare so that it supports pedestrian and their activities at night. They're very concerned about minimizing light trespass in the windows and they want to make sure that the luminaire style supports the early 1900s appearance of most of Chautauqua.

They also want to reduce energy and maintenance on the grounds. And just to show you that they're very serious about controlling glare here are six examples of fixtures that have been creatively modified on the campus in order to make sure that glare is mitigated for people who live on the grounds or are renting space on the grounds. All kinds of shields, funky interior baffles to block glare, etc.

Here's a view of a street. On the left you can see one of the streets with the old 175-watt mercury carriage lanterns and the second one down, the carriage lantern essentially deteriorated and the local utility replaced it with a high-pressure sodium fixture which really doesn't match the character of the grounds and is more glaring than the original fixtures.

So the residents of Chautauqua and the Chautauqua Institution itself decided that they needed to make some changes. So here is a series of demonstrations they went through to find a new pedestrian luminaire for the grounds. It has a 20th century ambience, early 20th century ambience and it used this 60-watt Cosmowhite ceramic metal halide lamp, dimmed at night. The lamp started out at 6,900 lamp lumens; it's an indirect design so that there's light that bounces up onto the upper reflector and then outward onto the grounds below.

Because of the indirect design it's not very efficient, or efficacious, so the lumens per watt for the luminaire is only about 18 lumens per watt, very low efficacy. But the color was very well-received. Warm, familiar color, very much like the original incandescent that the grounds had. Very little glare. The light levels were deemed acceptable but very little glare.

The second demonstration was a post-top fixture with an LED matrix located in the hood. And it was supplied at the request of the Chautauquans with a diffusing glass. 7x7 matrix of LEDs, and it produces 3,000 lumens, warm 3000K color, 58 system watts and 72 lumens per watt for the luminaire, quite high. The expected life was very long so the grounds really like the idea of reducing the cost of maintenance.

The light is all directed downward. So this was a good solution, except that the people who were surveyed responded that they thought that there was unacceptable glare. Even with the diffusing glass there was unacceptable glare.

So option number three: similar appearance, old-fashioned appearance. Round configuration of LEDs, 76 watts, warm color, expected life of—very long time. Light aimed downward. But even this luminaire was received as having unacceptable glare. And those LEDs that are embedded into the hood are directed so that some LEDs are directed in one direction, others are aimed down, others are directed out to the opposite side of the luminaire.

But as you are a pedestrian and glance up at the luminaire, there is always an LED that is aimed towards you and that LED will always be received as being glaring, at least that's what the surveyed residents told us.

So number four: this was the post-top lantern, also mounted on a 12-foot pole. This one was a little bit different in that it had an LED product with a remote phosphor panel. It was also installed with a diffusing glass lens, 81 lumens per watt, so quite high in fixture efficacy, 3,000 lumens came from this module, 3000K color.

The color was very well-received; light is aimed downward so it's dark sky friendly and residents who filled out surveys responded that this glare was much more tolerable, still a little bit too high but much, much more tolerable than the others. This was only drawing 37 watts.

Here's a view of that installation from above so that you can see the light pattern that's produced. And by the way, the two images in the lower left, one is of the light module inside the optical system before the flat glass was changed out to a diffusing glass, and the left one is that module seen with a diffusing glass.

So because this luminaire had a dimming driver the residents were able to go out and dim down the luminaire to see if dimming improved the acceptability of the glare. And bingo, it did. They dimmed it to 60 percent of the original level, which brought the lumens emitted from the luminaire down to about 1,800 lumens. It still had a light distribution that was deemed very good, about 75 feet total, or 37-1/2 feet on either side of the pole, and the average foot candle level when it was new was about 0.4 foot candles.

Now it's accepted that the light levels will drop over time because LEDs will drop in light output and there will be some reduction in light output due to dirt. And that's why the residents were happy with 60 percent of the light output. They still thought that was a little too much light but they recognized that there would be some automatic dimming over time just due to dirt.

There is a module that is a 24-watt module that draws about 27 watts including the driver. It comes to 66 lumens per watt and that is the product that's going to be installed for a mockup of nine poles that's planned for next year.

So what factors affect pedestrian glare? The viewing angle for the pedestrian. Remember, it's different than the viewing angle for the driver. The luminance of the luminaire relative to the viewer adaptation luminance: it can't be so high that it raises the viewer's adaptation luminance, or elements that are in the darker areas simply aren't going to be visible.

The luminaire's luminous distribution—now what I mean by that is the LED products are not uniform brightness, except for the one with remote phosphor. It's a bright dot next to a darker space next to a very bright dot next to a darker space. It's true also for a lot of the metal halide products where you have a very bright arc tube maybe next to a darker portion of the reflector.

The maximum luminance doesn't seem to be related to the perception of glare for the pedestrian. This is studied at Chautauqua and actually the product that was measured as having the highest glare, the highest luminance, was the one deemed least glary. So we think that there's something else going on with the perception of glare. It may have to do with the distribution of luminance, or brightness, across the surface of the luminaire.

This needs more study. It may be that small, intense patches of brightness may appear more glaring than if there's either diffusing glass or a glass refractor that spreads the light over a larger area. And higher correlated color temperature was usually perceived as brighter than lower correlated color temperature.

Here's an example that was generously shared with me by Acuity Brands. They looked at two LED products, one with a clear glass and one with a small, prismatic glass. You can just see the difference in appearance, especially from an angle, as in the two bottom images.

What happens when you put some diffusion over it? Well, you kind of spread the brightness over a larger area. So the question is doesn't this diffusion turn the optics to mush? And the answer is yes, it does.

Here is an example: this is a type three distribution from that previous luminaire, the LED luminaire with the clear, flat glass on the bottom. And the blue outline is the light pattern on the ground from that luminaire at—oh, if I remember correctly it's—the illuminance contour has a foot candle, at a 15-foot mounting height.

Well, if you put in the prismatic glass suddenly the light pattern changes; you can see the red pattern. So it draws in, that half a foot candle occurs at a distance closer to the pole. So you're right, you're not getting as wide a spread of light from it. Diffusion definitely does reduce the spread of light.

Does it matter? Perhaps that's something that we need more study to determine. Here is the IES classification system for outdoor luminaires. Unfortunately the classification system doesn't take pedestrian glare specifically into account. The glare zones are on the forward side and the backward side between 80 and 90 degrees, the BVH and FVH that you see there; those are between 80 and 90 degrees. F means forward, very high; BVH is backlight very high.

That doesn't correspond to the pedestrian's perception. It may very well correspond to the driver's perception. So we know that probably that BUG system will not be predictive of glare for the pedestrian.

So how do we mitigate glare? We can't use the BUG system to help us figure out what's going to be more glaring for the pedestrian; what can we do? Think about using lower lumen output luminaires. We may not need light levels as high as we originally thought. Luminaires that spread brightness over a larger area may be less glaring, luminaires with less optical punch and less sharp cutoff at the edge. And some of the luminaire people call this a rapid runback, which means you have that killer candlepower at 70 degrees or 75 degrees and then suddenly at 76 or 77 degrees suddenly your candlepower drops back to a very, very low number. That's called a rapid run back.

That works very well for driver glare, mitigating driver glare, and for producing very even light levels on the streets, but it does not work for pedestrians because that flash of light at the high candlepower, and the fact that it drops back quickly or shuts off quickly when you step out of the zone may actually be received by the observer as more glaring, not when you go from high candlepower to low candlepower, but if you're walking towards the luminaire and suddenly you go from low candlepower to very high candlepower in the eyeballs that could be perceived as very glary.

Luminaires delivering warmer color light seem to be perceived as less glaring. Should we think about those in pedestrian-intensive areas? Perhaps. Here are some tradeoffs that are inevitable. You've got warm color, soft, low-glare pedestrian luminaires, but if you're using lower lumen output luminaires to produce slightly lower light levels you are probably also reducing visibility slightly. We need to recognize that.

Warm colored lighting delivers lower scotopic-photopic ratios, so it may be a tad more difficult for the observer, the pedestrian, to see clearly in off-access areas. So not in your foveal vision, which is straight on with your central vision, but picking up activity in your peripheral vision at very, very low light levels; it may be more difficult because you have a warmer color light source that delivers lower SP ratios.

3000K LED packages are less efficacious than the 6500K packages. 4000K packages, which are becoming much more common, is about 8 to 10 percent in some cases less efficacious than 6500K. When you go to 3000K or even to 2700K you may be looking at 20 percent less than 6500K. That is a fact of life; we may have to deal with slightly lower efficacy in order to get pedestrian-friendly lighting.

We're hoping that this is going to improve with time; in fact we've already seen 2700K and 3000K packages improving in terms of lumens per watt and we hope that will improve more and more with time.

Mushy light distributions produce less uniform ground plane lighting, but keep in mind that when you have a slightly mushier light pattern you may also be improving the light that's hitting faces. So you may improve vertical light on faces of people that are walking towards the pedestrians and improving slightly the vertical light on objects ahead of the pedestrian.

As soon as you put in lenses and diffusers you're going to be losing some light output, between 10 to 20 percent, sometimes more. And that is also a fact of life. So yes, there's a reduction there. Are there more efficient ways to do it? Yes, there are better optical systems and we're hoping that perhaps this talk will encourage manufacturers to develop even more efficient optical systems for giving us diffusing without forcing us to sacrifice efficacy.

Conclusions: every project is different. The needs are going to vary. There are so many different clients out there. There's so many different users. Everybody has different expectations and needs. We recognize this. These principles will not apply to everybody. It's going to vary from project to project. There's no glare metrics that we are aware of that work for pedestrian lighting at this point in time because the glare metrics do not seem to take into account the distribution of luminance over the surface of the luminaire.

They also don't take into account the fact that pedestrians may be seeing the luminaire from angles that are lower than the angles that we used to think were affecting glare perception.

The problems of pedestrian lighting occur with all technologies, not just LEDs. However, LEDs are giving us optical options and opportunities we have never had before. And that's why I have so much faith that we're going to come up with better solutions and LEDs are going to be a part of that solution.

Should the Illuminating Engineering Society investigate pedestrian-friendly lighting and a modified recommended practice? I hope so. I think it's high time we thought about that and perhaps sponsored some research into that very issue. This talk is obviously meant as a thought experiment. We're trying to stimulate discussion. We're trying to stimulate research. And we're trying to get people to talk about new ways to think about pedestrian lighting and luminaires.

And I hope this has been of some value to you. You can see that this is the last slide. I thank you very much and I am prepared for the onslaught of rotten tomatoes. So Linda, do you have some questions for me?

Linda Sandahl: I sure do. We had a number of questions; I think you spurred a lot of interest and there's a number of things people would like to hear more about. So let's go ahead and start.

Regarding the list of pedestrian concerns you shared, was that in priority order?

Naomi Miller: No, not necessarily. That's a good question. No, those were in no special order, and they should not be in any special order because frankly, it's going to vary according to the needs of every community.

Linda Sandahl: OK. Which glare metric should be used?

Naomi Miller: I don't know of a glare metric that we can use. So unfortunately I don't have a response. I know that the BUG system doesn't seem to work. We know that measuring luminance of the luminaire doesn't seem to work, and we have looked at some glare metrics, we've even looked at UGR, for example, and similar glare metrics, veiling luminance ratio, for example. Unfortunately those glare metrics rely on looking at the candela value from the luminaire and the candela value doesn't represent the glare—the perceived glare of the luminaire because it does not take the luminance distribution across the face of the luminaire into account.

Linda Sandahl: OK. Can you talk about any inertia or special considerations that are being given to promote more solar-powered lighting solutions?

Naomi Miller: Hmm. Solar-powered solutions are a really good idea if you can pair it with a battery system that is going to be able to reliably store power that is generated by the sun during the day and then release it at night. There are—I think with the improvement in battery systems in recent years, solar-powered lighting is going to be a much more viable option, especially in places where the utility electric rates are high even at night, or especially in places where there is little electric service available at night.

Linda Sandahl: OK. Is pedestrian-friendly lighting criminal-friendly lighting?

Naomi Miller: I don't think that pedestrian-friendly lighting needs to be criminal-friendly lighting because one of the things I said is that although perhaps the spread of light on the ground isn't necessarily as uniform as it is with what we've traditionally considered street lighting, there may be more light emitted at high angles because as soon as you put diffusion on a luminaire it spreads a little more light at high angles. So that little bit of light at high angles may help light up faces. It may help light up cars. It may help light up building surfaces so that perhaps it is just as easy to spot the criminal. I would hope so. That is something that we need some research for. And I would hope that we can find a happy medium between lighting that is pedestrian friendly and criminal lighting that does a really good job for criminals that's just horrendously disturbing for pedestrians.

Linda Sandahl: This is a similar question: why not use pedestrian-friendly lights in high crime areas? Don't you want to see people's faces?

Naomi Miller: Yes. I think that we can use pedestrian-friendly lighting in high-crime areas. You may have to use higher light levels in high crime areas but there's no reason why we can't use luminaires that are deemed to be less glaring in high crime areas; in fact, it may be an advantage.

Linda Sandahl: The next question goes back to the earlier solar question: have you seen effective use of solar-powered lighting in pedestrian-lighting applications?

Naomi Miller: I have not personally seen solar installations that are effective but the principles would be the same. Solar lighting, up to this point, has had to be fairly low wattage because it's very difficult to get solar panels big enough to collect enough solar radiation to power high-power luminaires at night. So for the most part the installations I've seen have been limited to, say, 30-40, maybe 50 watts of LED power at night.

I think that with the improvement in battery power we're going to be able to see more effective solar lighting installations, and as long as we apply the principles for pedestrian-friendly lighting I think they could work hand-in-hand.

Linda Sandahl: Do linear sources, including linear LED luminaire styles, provider better or best results?

Naomi Miller: As long as the LEDs in that linear luminaire don't end up being a series of glaring dots. Yes, I think it would be very possible to apply diffusion, or to use optical systems that are able to control the distribution of light carefully, but at the same time diffuse the high intensity of the LED for glare control purposes.

Linda Sandahl: All right, next question: for Stanford option A did CMH lamps produce 2700K? Looks like a higher CCT.

Naomi Miller: Yes, the lamp that was used in that case was a 3000K lamp; I think there is at least one manufacturer that makes a 2700K metal halide lamp as well. That was not tried. The optics weren't pursued, or I should say because there were other products that seemed to do a better job there wasn't additional work done to try the 2700K lamp. That may be a good option in the neighborhood areas, for example, since they—well, actually I'm not sure. In the neighborhood areas the best solution is an LED solution; the ceramic metal halide lamps still are limited in terms of lamp life compared to LED.

Linda Sandahl: OK. On slide 20 why was option C, which was 3000K, acceptable; an earlier slide says that 3000K is too white.

Naomi Miller: Now that may have been the neighborhood area, is that right? And that's not to say that the 2700K option wouldn't have been even better received. You have to understand that this was a process of trying one fixture after another and there wasn't always the opportunity to go back and try additional options of every single luminaire.

Linda Sandahl: OK. Related question: please explain with more detail on the frosting of interior of the glass on option C.

Naomi Miller: Basically if you take the Holophane RSL-350 prismatic glass out of the luminaire—it's a very heavy prismatic glass that, I'm trying to remember—I think the prisms are on the outside of the lens and it's smooth on the inside of this cylindrical glass. That cylindrical glass was taken to a shop where a work person took sandblasting material and sandblasted the interior of that glass refractor. So what that did was to add diffusion; it adds a little bit of tooth to the smooth surface that's on the interior of that prismatic glass.

Linda Sandahl: OK and another related question: did frosting the glass have any impact on efficacy?

Naomi Miller: Yes. Frosting the glass, if I remember correctly, the reduction in efficacy from frosting the glass was something like 7 to 10 percent perhaps. So yes, there was a tradeoff.

And I should mention one more thing: as soon as you frost the interior of the glass that changes the optical distribution from the luminaire. It will send probably a little more light upward and a little less light downward. So what it does is to actually make that luminaire a little bit less dark sky friendly in that light that was directed downward is now being diffused, so it's going upward.

However, the tradeoff in lumens emitted from the luminaire going from the original high-pressure sodium lamp to the much lower wattage LED solution meant that even with the small increase in light going upward from the LED solution, you're still sending far fewer lumens into the sky.

Oh, and I just got a correction here: the estimate on the reduction of light output from the frosting on that prismatic glass was about 8 percent.

Linda Sandahl: OK. For campus lighting does the luminaire efficacy meet the DesignLights Consortium requirement for rebates?

Naomi Miller: I'm lost. I don't know the number. Does anyone else know the number for the DLC rebates? I have to find that out and respond to that individual later.

Linda Sandahl: OK. Was induction considered? Or is induction still being considered in these applications?

Naomi Miller: Good question. Induction was considered for the luminaire on the Stanford campus. It was one of the first products examined, and the advantage of induction of course is that the lamp life—well, so to speak—the lamp life is about 60,000 hours, maybe longer. The driver life is—excuse me, the lamp life is about 100,000 hours; the driver life is about 60,000 hours, if I remember correctly. So it was long life lamp.

However, that option was available in 3000K; it was not available in a warmer color version, and it was only available from one manufacturer. So the decision was to look at other options as well.

Linda Sandahl: Were optical models rendered for the Stanford project? Or did you just try different fixtures?

Naomi Miller: No, optical models were not developed for the Stanford project? Rita Koltai and the facilities group relied on looking at fixture samples in their offices, looking at photometric reports, looking at product cut sheets and then ultimately in mounting them on campus poles and taking a look.

There's unfortunately nothing like seeing the light fixture with your own eyes and looking at the pattern of light on the ground with your own eyes to make those kinds of decisions.

Linda Sandahl: OK. Instead of conventional diffusers have you tried binary optical diffusers?

Naomi Miller: Well, I just learned about something new. I don't know what a binary optical diffuser is. So I'd love to learn more about it.

Linda Sandahl: Something we can look into after this. The International Dark Sky Association recommends filtered LEDs, so CCT lower than 3000K to minimize blue emissions. What are your thoughts on that?

Naomi Miller: The IDA is trying to reduce the emission of blue light because there's some concern that critters, and possibly even humans, may have biological systems that respond to blue light at night and it may disrupt circadian rhythms and other biological systems. I understand that interest and I think it makes sense in areas of extreme environmental concern, such as national parks, any kinds of animal conservation areas, critter conservation areas, waterways, for example.

The moon is 4125K. Now it may make sense to use higher CCT light levels in places where visibility is a very critical issue; I'd probably recommend that we not use color temperatures that are dramatically higher than moonlight at high levels. But remember that moonlight levels at night are extremely low. It's about 100th of a footcandle, or 0.1 lux.

So street lighting levels and parking lot levels using high CCT products may have some negative consequences on critters; I'm not entirely sure that I'm the expert who can give you that kind of a positive response. But let me say this: in places where the environment is really critical it may indeed make sense to use warmer CCT color temperatures and lower light levels.

Linda Sandahl: You mentioned many different segments of pedestrian lighting, so municipal, commercial, and universities. Are you seeing any more interest in a particular segment or an increase across all segments?

Naomi Miller: I think an increase across all segments. In fact I went to the Street and Area Lighting Conference back in September and I was surprised and pleased to find out how many luminaire manufacturers were already producing products that were controlling the brightness of the individual LEDs, either through somewhat indirect optical systems or using diffusion because they recognized that a number of segments in the lighting industry, a number of applications were looking for fixtures that were less glaring and more conducive for pedestrians.

Linda Sandahl: I'd be interested in a few examples of what would not be good for pedestrian lighting and why not. Is it largely a function of driving needs?

Naomi Miller: Examples of what would not be good for pedestrians? Yes, any kind of lighting where you really need to be wearing a baseball cap to be comfortable walking around. That can be a problem. And that can be glare caused by metal halide, glare caused by LEDs. It can be all kinds of light sources. But especially ones with perhaps clear arc tubes or with very bright LED points of light.

Linda Sandahl: Wouldn't a diffusion provide a more consistent lighting level under the lamp?

Naomi Miller: Well, it does provide a more consistent lighting level, although that's not to say you couldn't do that also with an optical system that was designed to spread light evenly. What diffusion does not do is to help you send high candlepower values out at extremely high angles. So it doesn't help you send out candlepower at 70 degrees, 75 degrees, 80 degrees from the luminaire. Consequently those are the angles that some manufacturers have really worked on for high candlepower in order to get as uniform a light distribution as possible, which gives you wider spacing of poles, and a less expensive lighting system, consequently, because you're using fewer poles.

So if you're talking about pedestrian-friendly lighting, that is probably not going to work with the kinds of lighting systems that are designed to give you killer uniformity with very wide space in the poles.

Linda Sandahl: All right, next question. One of the stated advantages of LED is better recognition of color. So a person with a white versus a red jacket. Did any of your surveyed users care about that?

Naomi Miller: The questions in the surveys were about colors of objects, but there was no specific comment about a specific color. But I'd have to revisit all of those survey forms to give you a definitive answer. I'm sure that the response to LED color was very good, partly as a consequence of comparing it to high-pressure sodium, which of course does not give you clear color differentiation.

Linda Sandahl: OK, what would be your suggestion for lighting gas station fuel islands and the space between the store, as well as the entry area when drivers and pedestrians are mixed and store operators are concerned about security?

Naomi Miller: Well, one issue is you need to keep the lighting within the store—I should say you want to make sure that the operator, the cashier inside the store can see through the windows out to the islands. And that often means that you need to keep the lighting on the interior of the store, especially surfaces that may reflect in the windows, lower in luminance, so that that piece of glass doesn't turn into a mirror. So you want to promote visibilities through those windows and you also want to have enough light out there lighting mostly vertical surfaces. The horizontal surfaces will sort of get lighted by default; lots of bounced light will hit those surfaces. But the vertical surfaces, which mean lighting on the vertical surfaces of cars, lighting on the pumps themselves, and lighting on the bodies of people, that's very important. So that's where the lumens need to be applied so that they provide good vertical illuminance on the bodies and the surfaces that are vertical in that canopy area.

Linda Sandahl: When you're going from a 100-watt HPS lamp to a 27-watt LED the SP ratio color temperature and CRI will look brighter to pedestrians but the actual measured walkway illuminance will be reduced. How do you address the liability risk, as a light owner, from lawsuits filed by individuals who may claim that the new light didn't provide as much on the ground as the light that was replaced?

Naomi Miller: Well, that's of course a very serious consideration and that's one reason why I think the Illuminating Engineering Society needs to address conditions under which pedestrian-friendly lighting and lower light levels may be appropriate. It does help the lighting designer in that kind of lawsuit or an argument if they can point to the IES standard and say, "Listen, a lot of people agree with me that these light levels that I've provided are appropriate for this kind of a space, or this kind of an outdoor application."

The light levels that were provided were not necessarily dramatically lower than what you would provide—well, they weren't dramatically lower than what the IES already recommends for pedestrian areas where there are low levels of traffic, for example. The light levels in general on both of these projects, oh, I think the light levels at the Chautauqua installation, horizontal light levels range between about 0.1 footcandle up to one or 0.9 footcandles and in the Stanford area they may have ranged slightly higher than that up to two footcandles in some cases.

So we're not dramatically out of range. It is certainly light levels that are lower than what you find in some shopping center parking lots, and that's appropriate because you have less vehicular traffic, you may have different populations that have different kinds of activities, and you have perhaps more people around, more people that are helping make the space feel safe.

You have to remember that one of the things that make outdoor spaces feel safe is other people. So if you are on the streets in Paris at night at 11:30 at night, you walk out of the subway and you feel perfectly safe because there are people sitting in outdoor cafes, even in March, for crying out loud. And there are people everywhere. So you're less concerned—you are visible and they are visible to you.

If we can create spaces that encourage people to be outside, that provides a level of comfort and a level of safety that we can accomplish in many cases by using slightly lower light levels, and light fixtures that are more comfortable to be around.

Linda Sandahl: CRI seems conspicuously absent from the talk; why is that?

Naomi Miller: Well, frankly, CRI wasn't addressed much because in general the CRI was pretty doggone good. The lowest CRIs we had was from the Cosmowhite CMH lamps which was 70. So the color rendering was pretty good; there was nobody who really complained that the color rendering was poor, except in the cases of the existing lighting, the high-pressure sodium that was down at 22 CRI.

Linda Sandahl: OK, and we're getting short on time; I'll just ask another question. Are you aware of any other case studies that show how nighttime lighting strategies have enhanced public safety or crime prevention?

Naomi Miller: I know that there was a researcher in England named Kate Painter who did a number of studies about lighting systems and how the lighting systems affected crime in these neighborhoods. I believe that you can go online and find her work.

I know that the International Dark Sky Association also has accumulated some papers on light and crime, and I believe that—it many have been Ron Gibbons who gave a talk at the Illuminating Engineering Society Street and Area Lighting Conference who talked specifically about some studies that addressed light and crime. So those are probably your best bets.

Linda Sandahl: OK. Well, we are now out of time, but Naomi, I want to thank you so much for your presentation today and thanks for everybody who submitted questions. These were great questions and I think we learned a lot from that too. So again, thank you for participating in today's webcast brought to you by the U.S. Department of Energy. Thank you very much and have a great rest of your day.