Posts Tagged ‘lamp life’

LED Lighting, About Life

Friday, August 28th, 2009

Another potential benefit to LED lighting is that they promise to last a lot longer than incandescent or even fluorescent lighting. It is often claimed that an LED will last 100,000 hours. In practical terms, how long is 100,000 hours?

  • Left on 24 hours a day, 100,000 hours means about 11 1/2 years.
  • Left on 12 hours a day, for example all night all year long, 100,000 hours means about 22 years.
  • Left on 9 hours a day, for example all day in your office, 100,000 hours means about 30 1/2 years.

Those calculations are pretty impressive, which is why they are used. Especially when someone is trying to convince you to buy a $36 LED light bulb. If you compare a single $36 LED lamp against 100 to 133 $0.75 lighting bulbs the economics seem to make sense.

However, it is not a fair comparison. Similar to the posting about LED efficiency, you can’t take the life of the LED module itself and apply it an LED used in a lighting application. To be fair, the major lamp manufacturers who are entering the LED game have toned down the rhetoric, and usually claim somewhere between 30,000 and 50,000 hours. However, the big claims are still out there.

Like efficiency, everything you do to an LED will tend to shorten its life. Much come back to the heat that has such a negative effect on efficiency. That same heat shortens the lifespan. If a module and the fixture design are good at dissipating the heat the lifespan isn’t shortened as much as it would be with a poor design, but there is always going to be some effect.

The problem at predicting the effect is that the LED lighting fixtures being designed and sold are still so new there just hasn’t been enough time to adequately test the claims.

Technical note: you don’t necessarily need to know this bit.

Lifespan for lighting is odd. It measures the number of hours a lamp type in aggregate is going to average until the lamp is no longer useful. For incandescent lamps this is fairly simple. If an incandescent lamp has a 750 hour life, than if you take a large sample of lamps after 750 hours you would expect about half of them to still be on and half to have burned out.

Lifespan for fluorescent lamps is more complex, since they slowly decrease the light they put out over time. Therefore, it is possible to have a lamp still function that isn’t putting out enough light to be useful. So if you take a very large sample of lamps with a rated life of 10,000 hours, after 10,000 hours more than 50% might still be illuminated, but only about 50% will actually be useful.

Lifespan for LED modules is like fluorescent lamps. They slowly decrease over time. There is an industry testing protocol (LM-80), but not everyone is using it. For example, some people use the “B50″ claim, which is the point when 50% of the LED stop turning on. Others might be a bit more reasonable and use the “L50″ claim, which is the point when the lumen output is 50% of the original. Others use the “L70″ claim, which is the point at which the lumen output has dropped to only 70% of the original. LM-80, the testing protocol from the IES, uses L50 or L70 based on the application.

To return to the main point: The problem at predicting the effect is that the LED lighting fixtures being designed and sold are still so new there just hasn’t been enough time to adequately test the claims.

If you take a 100,000 LED module and stick it in a fixture, you have to test the life of the LED in the fixture. Say we think it will last 50,000 hours instead of 100,000 hours. That means to adequately test the claim (not just run the computer simulations) you have to build a bunch of fixtures and test them for over 5 years. These test are in progress, and there are plenty of tests that have been completed. However, in terms of the overall industry those amount to spot checks, and what we really need is the sheer massive quantity of testing completed that will allow us to make confident claims about the industry as a whole. We’ve been using incandescent lamps for over a century, and fluorescent lamps almost as long. (Earlier posting: fluorescent precursors were invented before incandescent lamps but not really commercially viable or available until the 1920s.)

In general, the longer the life claim of an LED the more skeptical your approach should be. Chances are, if a manufacturer says their LED fixture has a more limited life of around 20,000 hours it is because they have gone through the testing procedures and there is the paperwork demonstrating that rated life and they know they can’t get away with claiming anything longer. If they claim 100,000 hours, chances are they have taken the number straight from the original LED laboratory results and not actually tested their own application.

Finding a More Efficient Lamp

Monday, July 13th, 2009

As I was finishing up my previous post about efficiency pitfalls, I realized that I had started an example which could be useful in and of itself. So I’m going to take it a step further.

Situation Summary: I have an existing fixture in my home that I want to make more efficient. It is a surface mount that looks like a glowing bowl stuck on the ceiling. It uses two 75 watt bulbs. The light in the room is more or less acceptable, although it is a little dim and I would like it to be slightly brighter if possible. Looking up the information about the existing bulbs, I see that the total lumen output of both bulbs is 2,340, so I know I want to at least match that and if possible increase it slightly.

I could have just substituted my 75 watt incandescent bulbs for 20 watt self-ballasted CFL bulbs, based on the label on the box that says to replace my 75 watt bulbs with the 20 watt CFLs. The energy savings is good, since now my fixture only uses about 40 watts instead of 150, meaning I’m saving about 73%. However, the mean lumens of those CFLs are only 965, so the new total would only be 1,930. Since I thought the room was already a bit too dark, going from 2,340 lumens to only 1,950 is going to make me unhappy over the long term, and I run the risk of someday switching back to the incandescents. (Note, the initial lumens of this bulb are listed as 1,150, which brings my total up to 2,300 and almost a perfect match to my existing. However, that is initial lumens and won’t last for the lifetime of the bulbs. They will continue to depreciate and my room will get darker and darker, so it is an unfair comparison to use the initial lumens.)

Fearing the decreased light output, I buy two 26 watt self-ballasted CFLs because I see that the mean lumens for those are 1,365, meaning my new arrangement will provide me 2,730 lumens! I’d be saving about 65% energy by using 52 watts, but I discover that the 26 watts lamps are bigger than my existing incandescents and they won’t fit in my fixture. The bowl hits the ends of the lamps preventing me from reattaching it.

Now I recall reading my own post from yesterday. I go to my local electrical supplier and buy a replacement fixture for the existing. Based on my post, I’m now using a long “blob” instead of a bowl on the ceiling. I get 2,772 lumens, so I am very happy with the new lighting level in the room, and I’m only using 32 watts. That’s a savings of about 75%, better than the savings from those CFLs! Granted, I’ve now bought a new fixture as well as lamp, and I had to be very careful about my lamp and ballast choice, but I’m much happier with the final result.

Of course, this just shows that in order to make the best choice you need to have some knowledge, but that’s what this blog is for. Feel free to send me questions or reply to my postings to get more information.

Efficiency Comparison Pitfalls: Age

Thursday, July 9th, 2009

Following up on yesterday’s post, efficiency pitfall number one: age.

An incandescent or halogen light bulb is a fairly closed system. It can be stuck in a socket and what you see is what you get. A 100 watt light bulb is going to be pretty consistent over time. There will be some decrease in the light output as the tungsten filament gets burned up and bits of it get stuck to the inside of the glass envelope, darkening the glass. However, for the most part the lamp will burn out and the filament break before the decrease in light output becomes much of a factor. Tungsten-halogen lamps do slightly better, since the halogen gas inside the bulb makes the tungsten that gets burned off reattach to the filament and basically “recycle” itself, but even though the lamp will last longer and suffer from less darkening it will still break before those changes tip the efficiency equation too much.

Unfortunately, the same cannot be said for the various discharge sources like fluorescent, compact fluorescent, or the various HID types. While a filament does exist in these lamps, the real work of making light is done by the gas discharge. (Check some of my previous posts about fluorescent lighting for a summary of how it works.) The light bulb does not just “burn out,” but instead slowly degrades over time putting out less and less light.

This means that the efficiency of a fluorescent or compact fluorescent decreases over time if you are just using the lumens per watt calculation. For example, the day you install a brand new CFL 26 watt twist self-ballasted lamp you’ll probably get about 1,700 lumens. However, since the output of fluorescent lamps decreases over time, manufacturers also publish the “mean lumens,” which is the lumen output at about 40% of the lifespan of the lamp, which in our example case is 1,365 lumens. So the efficiency has dropped from 65 LPW when brand new to 52 LPW. The lumen output will continue to drop until the lamp reaches the end of its life.

Granted, the lower efficiency of 52 LPW is still much better than the 17.5 LPW of the 100W lamp we’ve replaced (see yesterday’s example). However, remember that you also have 20% less light. So after a while, the new CFL lamp isn’t providing the amount of light you used to have with the old incandescent lamp. This is the point when some people give up and go back to the old lamp, or they add a reading lamp using another incandescent lamp.

Professionals try to counterbalance the lumen decrease by working depreciation factors into the system when they are planning out what lamps to use in a project. We will intentionally over-light a space from day one knowing that eventually the lamps will decrease in output to the level we want for the longer term. The first 100 hours of a CFL or fluorescent is when the decrease is the most rapid, which is called “seasoning.”

As another age related problem, fluorescents don’t necessarily “burn-out” like incandescents and simply stop working. Sometimes, often times, they will just continue to get darker and darker and darker and don’t simply stop turning on. This is why professionals recommend having a maintenance schedule for replacing your fluorescent lamps based on time, not waiting for the lamps to stop working. (What? No one told you this? Yes, if you are going to use CFLs you should keep track of how old the lamps are replace them on  a schedule. And yes, it’s much more of a hassle than just waiting for the bulbs to stop working.)

Dimmers in the Home

Monday, June 15th, 2009

Over the next few days I’ll give some tips for using dimmers in your home. I always recommend people replace their switches with dimmers. It is one of the quickest, easiest, and cheapest ways to improve your lighting environment. Low cost dimmers are available in hardware stores and big-box supply centers (Home Depot, Lowe’s, etc.) all over the country.

Some quick info:

  • Dimming your incandescent loads 10% can save you 10% electricity and can make the bulbs last twice as long.
  • Dimming 50% saves about 40% electricity and can make the bulbs last 20 times longer!
  • Dimming incandescents shifts the color toward the red end, which most people perceive as warmer and more comfortable.
  • Dimming can help with your sleep patterns.

Intro to Fluorescents, Day 5

Friday, June 12th, 2009

Maintenance for fluorescent lamps is very different than incandescents. With incandescent bulbs you just put them in and replace them when they stop working. Fluorescents don’t work that way.

Fluorescent lamps typically don’t “burn-out” and stop working. They slowly get dimmer and dimmer over time. They need to be replaced based on how long they have been on. It is different lengths of time for different lamps, but the box will probably list the expected life of the lamp, and it will be a number like 8,000 or 10,000 hours. If you have your lights on for about 6 hours a day, say 1 hour before work and 5 hours after (6 to 11), you’ll have them on a little more than 2,000 hours a year. The expected life is not the length of time until it stops working, it will work long after reaching its allotted time, it will just have gotten so dim as to need replacement. When you change the bulb it will seem much brighter.

There actually are filaments (similar to an incandescent light bulb) in a fluorescent lamp located at each end. When you see the ends of the lamp getting darker over time, that is the lamp aging and material burning off those filaments and depositing on the glass. If the ends get really dark you should replace the lamps.

Fluorescents need to be on for about 100 hours to stabilize the phosphors when brand new. During this time they will be brighter than their rating and the color may shift, but not enough for you to see it. It really only matters for dimming fluorescents, since if the stabilization isn’t done before the lamps are dimmed it can dramatically shorten their life.

Fluorescents are temperature sensitive. They work better in warmer temperatures, up to a point. When you first turn them on it may take a short time to come to full intensity as they warm up.