LED lights for shop

[a1953mdl]

My brother in law has a large shop building he lights with a mix of 1,000 watt and 400 watt metal halides. He does metal fabrication and mechanical repair on heavy equipment. The shop is 50'x100' with a ridge height of 28'.

Anyway, I have been interested in LEDs for a while and have built a few smaller projects. We were talking about how efficient LEDS are and he is interested in possibly converting to LEDs. Enough so that I have a $350 (maybe a bit more, but not over $500) budget to build a prototype LED light, to see how it compares to his existing 1000w M/H.

I am thinking of using 2 ea high brightness 100w LEDs from ebay.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=400115299620&ssPageName=STRK:MEWAX:IT
I am hoping 2 will be enough to replace the 1000w m/h. (thoughts)
I have a box made of thin (maybe 22-24 ga.) aluminum which is open on one side. The box is 20"x20"x16". I am thinking of mounting the LEDS on a square sheet of 1/8" aluminum and then bonding that to the aluminum box on 4 sides as a large heatsink and fixture combo. I would have a few holes in the sheet and the top of the box for heat to escape from, but this would give around 3300 sq inches of surface area. (if my math is close)

Would this be enough heat sink for these LEDs? It probably gets 120F up near the roof in the summer here (Texas).

Is it reasonable to expect 200w of directional LEDs to compare to 1000 w of non directional metal halide?

I will be having more questions, if this seems like a viable project. I would be using an appropriate driver (also from ebay) for this first light. I have been reading on the need for constant current and/or constant voltage drivers for these higher brightness LEDs. Still a lot to learn. I am a carpenter by trade.

What are your thoughts on this project? Worth trying?

Lee

 

[jason 77]

I don't think those ebay LEDs are going to be very efficient. If you really want to try to build a light like this then I would go with a couple of bridgelux leds like the BXRA-C4500-00000

http://www.newark.com/bridgelux/bxra-c4500-00000/high-brightness-led-arrays/dp/60R6542

but I still don't know if they would be more energy efficient than what he has already... what are the specs of the lights he uses now?

 

[Lynx_Arc]

First off from the little research I have done metal halides according to some googled sites can approach up to 115 lumens/watt that would make a 1000 watt metal halide lamp at 100 lumens/watt a 100k lumen source. If you compare that to that bridgelux LED at 5000 lumens you would need 20 of them at ~70 each that is 1400.00. as for your ebay LEDs 140.00 for 14k lumens (for 2) you would need 7 of these sets at 140 a set or about 1000.00.
IMO 100 lumens/watt won't be beat using low to middle range LEDs only the higher bin LEDs can beat that unless you underdrive medium bins to get more efficiency. Oh, and on top of all the LED prices... you would have to buy power supplies and fashion heat sinks. I would be looking at LEDs that get over 115 lumens/watt to even bother with this project otherwise the savings is going to be negligable

 

[blasterman]

Like Lynx just said, a 1000watt metal halide throws out at least 100 lumens per watt, and even a mediocre MH reflector in 75% efficient. The best LED commercial fixtures right now are just creeping towards 100 lumens per watt. Most of them less.

Even using Bridgelux, which has a better lumen per dollar ratio than the Semi based array you linked, you'd be spending a fortune.

The disadvantage of the big metal halide is it's likely so far up in the air it wastes a lot of light. Plus, as a single point light source it's likely annoying to work under because of shadows. By using lower light sources more located over the specific work areas you can save some light.

However, then T5/T8 fluorescent come into play, and they are just as efficient as the LEDs were' discussing.

 

[AnAppleSnail]

I'll second that the problem with the existing lights isn't their efficiency but their placement. You can do things with LEDs that are impossible with other lights. Task lighting would be much more efficient for general work, though that depends on the work flow and other things at the shop. If there are no legal requirements for brightness at the metal shop, he could run fewer of his metal halide lights and use them as walking-around lighting, with task lighting at work stations. That could significantly reduce power usage, but you'll have to know about the lighting needs to really answer it.

tl;dr, industrial lighting fixtures are pretty efficient but customizing lighting for the workplace gives best savings.

 

[a1953mdl]

Thanks guys, lots to think about.
As for the specs on his existing lights,,,,,,he's not sure. But probably 100,000 lumens and fairly inefficient ballasts as the light fixtures are nearly 20 yrs old.

I had come up with the "hope" of 200w of LED replacing the 1000w MH because of some posts on the aquarium forums in their quest to replace the metal halides above their tanks. They had said that 250w of LED was replacing their 1000w mh.
I know its not the same but I also read on a report of some LED streetlights that were installed in iirc Raleigh NC and they were saying the LEDs were less lumens than the streetlights that were replaced but "seemed" brighter.

BUT it does make sense that lumens are lumens.

The bridgelux LEDs and the mean well drivers do seem like an efficient package.

Lee

 

[Lynx_Arc]

Thanks guys, lots to think about.
As for the specs on his existing lights,,,,,,he's not sure. But probably 100,000 lumens and fairly inefficient ballasts as the light fixtures are nearly 20 yrs old.

you would have to look up the ballasts, it may be just updating the ballasts could save you some money on bills.

I had come up with the "hope" of 200w of LED replacing the 1000w MH because of some posts on the aquarium forums in their quest to replace the metal halides above their tanks. They had said that 250w of LED was replacing their 1000w mh.
I know its not the same but I also read on a report of some LED streetlights that were installed in iirc Raleigh NC and they were saying the LEDs were less lumens than the streetlights that were replaced but "seemed" brighter.

BUT it does make sense that lumens are lumens.

The bridgelux LEDs and the mean well drivers do seem like an efficient package.

Lee

if you use optics and reflectors you can direct more of the light but I am guessing the streetlights design may have had more to do with perceived brightness than the actual output as many streetlights just have a bulb and globe and waste light throwing even up into the sky above it.
Like what was said earlier for spot lighting LED could be a good deal but for area lighting high efficiency fluoros can match LED on output easily while being a lot cheaper. LEDs need to get more efficient and the ballasts/power supplies need to get as cheap as fluoro ballasts then we will have a competition. As it is now LED power supplies can cost as much as a fluoro fixture, lamps, and ballast at times.

 

[blasterman]

They had said that 250w of LED was replacing their 1000w mh.

That's because on reef tanks the only thing that matters is light used for photosynthesis (PAR). LEDs are very good at generating lots of blue light that plants want. They aren't as efficient at generating white light at 5000k.

Also, a lot of reef guys are not comparing apples to apples when it comes to metal halide with LEDs and often cheat by using narrower optics. When you spread the LED light to the same angle of coverage as the metal halide, at most it's a 2:1 advantage for PAR. Also, in a shop or garage it's annoying as hell working under very directional light.

You're likely better off breaking the light down into smaller groups over the work areas, and lowering it. You can still save a lot of power this way and make things look better.

For the record, my DIY Bridgelux rails are getting close to the price of higher end fluorescent fixtures, including ballast. I'll add it up if anybody is interested. Cheap shoplights though with 2/4 48" T8s are dirt cheap, have good ballasts and have lots of firepower.

 

[AnAppleSnail]

LED streetlights that were installed in iirc Raleigh NC and they were saying the LEDs were less lumens than the streetlights that were replaced but "seemed" brighter.

I live there, I've been watching these go up on Hillsborough street. I'll get some pictures of 'em next week for your research.

 

[purduephotog]

I have a 60W led light source that's on an almost 7lb chunk of aluminum- 1/4" base, 3" fins (32 of them)... and without a fan it gets HOT.

So let's say you got a couple of these-

http://www.newark.com/bridgelux/bxra-c4500-00000/high-brightness-led-arrays/dp/

70$ a piece. 5000 lumens. And a fan. And a driver.
http://www.meanwell.com/search/eln-60/ELN-60-spec.pdf

ELN-60-24 - 2.5A, 24V max- but can be adjusted up to 26.4V (at the expense of amps, which is fine).

Then you'd need some reflector optics- and if you can find a ledlil that works let me know.

Good luck.

 

[a1953mdl]

Photog,
The first link doesn't work for me, but I'm pretty sure I know which LED it is.

With your 60w, do you have an idea if it is close to as bright as a 400w Metal Halide? What height do you have it mounted and what area does it give good illumination to? If you have it mounted, that is.

Do you have a light meter that reads in lux and is the light to where you could measure lux from some distance? Hopefully at least 10-15 feet from the light.

I had thought of a cooling fan, but was hoping the 3,300 sq inches of surface of the mounting plate and box would be enough.

Lee

 

[blasterman]

With your 60w, do you have an idea if it is close to as bright as a 400w Metal Halide?

We can't answer this if you aren't listening.:thinking:

The big Bridgelux BXRA-C4500 purduephotog was refering puts out around 4500 lumens when it warms up and consumes 55watts (minus driver).

A 400watt metal halide around the same color temp throws out around 35,000-40,000 lumens. There's no need for lux meters or anything else (unless we're dealing with OSHA complaints). These are are published numbers.

The metal halide will be at least 8x brighter and a bit more efficient. For the record, the big Bridgelux is among the least efficient emitters we can talk about.

The Bridgelux would have be no more than 1/3 the distance from the work area that the metal halide is to have the same brightness.

 

[a1953mdl]

Blasterman,
I am listening, I have pretty much decided this is not a practical project. But I am still trying to learn. My questions were more in an effort to put to rest some of the nagging questions I have in the back of my mind.
I have read many of your posts and had hoped you would be one of the people to respond to my questions. I like your common sense approach and the fact that you don't always go "mainstream".
I know that it gets hard to always answer us "newbs" questions, but it is appreciated and some of us do still want to learn. But what y'all see as simple and self evident is sometimes a hard thing for some of us to grasp. Therefore we ask some pretty "seemingly" ridiculous question.
Lee

 

[blasterman]

Nothing ridiculous about asking questions - ask away. Sorry if I came off as terse.

My main concern is that like the reefing question you aren't being 'swayed' by a particular lobby. Reef tank lighting in one of those instances where a very small group of people have caused a lot of those misconceptions.

For me at least it's easier to explain (and understand) things in the simpliest, logical chunks that it can be broken down.