how efficient are LEDs really?

[sphelps]

LEDs may be similar to other lighting types in the sense they have a bulb and ballast but my point was in relation to numbers. I've seen a tank use 400 LEDs while it would easily be lit by 3 or 4 halides. You can't argue that the chances of failure aren't greater for the LED.

Also with current fixtures, as far as I'm aware, the LEDs are soldered directly to a PCB strip and it's not very easy to replace a single LED and therefore the entire strip will need replacing. Perhaps if you're an experienced electronic assembler this might be something you can handle or if you made your own fixture it might be difference but certainly not for the average Joe.

As for the warranty I'm an engineer in a manufacturing field and I can assure you warranty is tied directly to product lifespan, and specifying something will last 10 years but only providing a single year warranty is a huge conflict. You don't need evidence to support that theory, it's common sense.

Also let's not get too carried away and start demanding evidence for this and that as this is all speculation on both sides. LEDs aren't exactly proven technology in this application and only time will provide the true answers.

The cost analysis is out of whack as well. Let's not compare DIY methods but rather apples to apples and use something you would find on the average tank. A standard 4 foot MH fixture runs for around $1000 and a LED fixture runs for $2200. For argument sake lets assume the MH will use 500W while the LED will use 250W both will use a little more but I think it's fair to assume the LED could potentially be twice as efficient. So a difference of 250W. At 10 hours a day that difference only adds up to $9.13 per month or $109.56 per year. Also it's awfully convenient for LEDs to be compared over 10 years, let's rather compare 5 years and 15 years.

For LEDs
5 years: ~ $2750
15 years: ~$6050 (new fixture after 10 years)

For MH
5 years: ~ $2100 plus bulb replacement = $2850 (2 bulbs per year @ $75 each)
15 years: ~ $6550

And that's provided nothing goes wrong with either fixture but like I said earlier the LED fixture is much more likely to have problems and when it does it'll cost more to fix.

Really the savings aren't there, with a larger up front cost the difference could easily be made up with the interest most will pay since it's too large of a sum for most to pay in cash. The only up side is the LED fixture has features the MH can't touch which is nice but not my point and not relative to efficiency or operating costs.

 

[evilc66]

You aren't factoring in the other support equipment that would potentially be used by a MH setup (chiller, fan, etc...) that would be eliminated by an LED setup. Removing the equipment cost and operating costs for that equipment adds up pretty quickly also over time.

 

[evilc66]

Another thing to add, this arguement is assuming that you use the more conventional "flood your tank with light like T5's" approach to an LED array. If you take a page from the Japanese methods of spotlighting, then energy savings are considerable. For example, my 40 breeder is currently lit by 4 15W PAR38 spotlights and I can grow anything I like. The look is different, but it works for me. You can't possibly argue that I'm not saving a considerable amount of money by using only 60W of LEDs.

 

[sphelps]

You aren't factoring in the other support equipment that would potentially be used by a MH setup (chiller, fan, etc...) that would be eliminated by an LED setup. Removing the equipment cost and operating costs for that equipment adds up pretty quickly also over time.

Maybe so but in all fairness I could argue the reverse, less heat would result in more work required by the heater and drive up operating costs. In either case the use of MH doesn't guarantee the need for a chiller. I use 500W of halide on my open top tank and my temp only climbs on the few hot days in the summer which is easily fixed with a small fan which only uses a couple watts and only runs when it needs to. An LED fixture needs fans running all the time to cool the heat sink so the additional operating costs for cooling could be higher for the LED.

 

[sphelps]

Another thing to add, this arguement is assuming that you use the more conventional "flood your tank with light like T5's" approach to an LED array. If you take a page from the Japanese methods of spotlighting, then energy savings are considerable. For example, my 40 breeder is currently lit by 4 15W PAR38 spotlights and I can grow anything I like. The look is different, but it works for me. You can't possibly argue that I'm not saving a considerable amount of money by using only 60W of LEDs.

Agreed, spot lighting is an alternative to reduce electrical costs but can you argue this for LEDs? The Japanese spotlights are typically 150W halides (as far as I know) so they are still quite substantial. How many LEDs would you need to match that and can you concentrate them over the same area? Might work well for smaller tanks but considering an average size tank you may not be able to pull it off. Typically LEDs are like T5s, you need to run as many as you can fit over the area.

I think LEDs work great for smaller tanks but currently I don't see a huge amount of potential for the average tank and larger. BTW I'm considering the average tank to be around 90 gallons.

 

[evilc66]

Like I said before, I'm lighting a 40 breeder (36x18x16) with 60W of LEDs (4 lamps, 5 LEDs per), and I'm able to keep anything I like. Granted, it's not a huge tank, but the appropriate MH setup for this would be at least a pair of 150W MH, not including any supplimental lights. I mentioned the Japanese method as it pertains to the use of spotlights in general, not the use of 150W MH spotlights specifically.

 

[beeker]

Honestly at 8 bucks an led and the fact that i wouldn't bet more than 1-5 leds would be failing per year...i've had an led flashlight for over 3 years and have never ever had to replace the leds in it...and Cree has far higher standards for their Leds than the flashlight i have does...

Believe me buying 120 bucks worth of MH bulbs every year plus the energy they consumer and the heat they put off are 3 reasons to go with Leds

Yeah leds upfront cost is expensive and i can agree with that but when i get my 200g DD tank up and running using only PAR38 bulbs about 12-16 of them (this will be about 2000k for the whole setup on track lighting) so decent upfront cost but better longterm savings

i'll be a lot happier and have a lot more options for controlling where the light would be placed on the rockwork using the PAR38's, i love MH's don't get me wrong but replacing the bulbs every year is a pain and the electricity i'd be using to light this tank would be a ton, i'd have to have atleast 500w's of MH and we all know most 250w MH's consume more like 320w's of power

12 PAR38's at 15w's a piece is a ton less electricity at only 180w's and even if i go to 16 bulbs it's only 240w's of power this would put 8 bulbs over each side of the 200g DD which IMO would be a lot of light and a lot of ability to focus the light where i want and need it to be for the Sps

I've looked into making a DIY led fixture but cost would be about the same factoring i'd need about 60 LED's on each side's of the tank, drivers and heatsinks and power supplie's as well as the mounting hardware, the only reason i'm doing the PAR38 spotlights over the DIY led is because to me replacing 1 led bulb or replacing a burnt out led bulb in a PAR38 fixture is much easier than replacing one thats mounted to a heatsink that already has 50+ leds on it and is already wired up

i'm not a big DIY person but i could make an LED fixture and make it work, PAR38's are simple though and the output i've seen on them is good enough to convince me that being able to control the direction the bulb is facing and where it hits the rockwork is better than lighting the whole tank when i wouldn't be using all that available light

 

[Skeptic_07]

I use 500W of halide on my open top tank and my temp only climbs on the few hot days in the summer which is easily fixed with a small fan which only uses a couple watts and only runs when it needs to.

things are much different for people who don't live in the far north. We don't have a few hot days, we have hot months.

An LED fixture needs fans running all the time to cool the heat sink so the additional operating costs for cooling could be higher for the LED.

Running a couple of computer fans is nowhere near the cost of running a chiller. Also, many MH fixtures such as current sunpods, employ the use of similar fans and not all LED fixtures require the use of a fan for cooling. Many times a large enough piece of aluminium is enough to dissipate the heat. Furthermore the heat generated by LEDs is conducted through what it is mounted to and flows upwards into the room while the heat from MH is projected downwards heating the aquarium water. Try putting your hand under an LED fixture while it is on and under a MH fixture when it is on.

With all due respect, I think you should do a little more research on LED's before having this argument because it is clear that you don't know all the specifics.

 

[g8gxp]

Maybe so but in all fairness I could argue the reverse, less heat would result in more work required by the heater and drive up operating costs. In either case the use of MH doesn't guarantee the need for a chiller. I use 500W of halide on my open top tank and my temp only climbs on the few hot days in the summer which is easily fixed with a small fan which only uses a couple watts and only runs when it needs to. An LED fixture needs fans running all the time to cool the heat sink so the additional operating costs for cooling could be higher for the LED.

An INCORRECTLY designed cooling system will require constant fan operation. And yes, some of the commercially designed ones that i have run across seem to be capitalizing on the LED mania currently making the rounds, and perhaps they haven't devoted enough design time. So speaking in regards to commercially available MH vs commercially available LED setups i think you are correct that LED's setups arent as cost efficient. But it is only a matter of time before costs start coming down.

As for arguing that less heat results in more work for the heater, that's just silly. This is basically saying that a lighting source that produces heat is desireable because you want to see a decrease tank heating costs.

 

[sphelps]

things are much different for people who don't live in the far north. We don't have a few hot days, we have hot months.



Running a couple of computer fans is nowhere near the cost of running a chiller. Also, many MH fixtures such as current sunpods, employ the use of similar fans and not all LED fixtures require the use of a fan for cooling. Many times a large enough piece of aluminium is enough to dissipate the heat. Furthermore the heat generated by LEDs is conducted through what it is mounted to and flows upwards into the room while the heat from MH is projected downwards heating the aquarium water. Try putting your hand under an LED fixture while it is on and under a MH fixture when it is on.

With all due respect, I think you should do a little more research on LED's before having this argument because it is clear that you don't know all the specifics.

I'm not sure why you're so uptight. Never did I say LEDs run as hot as MHs nor did I say a chiller is the same as a computer fan so with all due respect perhaps you should read my post again and don't quote individual lines out of context for the sake of argument. Using a Metal Halide doesn't guarantee the need for a chiller, more than likely those systems also use high heat pumps and have closed canopies and may even require a chiller if an LED light was used. Good metal halide fixtures are made from aluminum and don't even have fans built into them, like mine. We may not get too many hot days but it still gets close to 40 degrees C here and all I need to keep my tank at 80 is two small computer fans, the same as you might find on an LED fixture. My house also doesn't have AC which is typical in your hotter climate areas experiencing "hot months".

 

[sphelps]

An INCORRECTLY designed cooling system will require constant fan operation. And yes, some of the commercially designed ones that i have run across seem to be capitalizing on the LED mania currently making the rounds, and perhaps they haven't devoted enough design time. So speaking in regards to commercially available MH vs commercially available LED setups i think you are correct that LED's setups arent as cost efficient. But it is only a matter of time before costs start coming down.

As for arguing that less heat results in more work for the heater, that's just silly. This is basically saying that a lighting source that produces heat is desireable because you want to see a decrease tank heating costs.

Yes well when costs are more reasonable LED lighting will make much more sense but that hasn't happened yet. They also need to be verified in application to meet the given specifications. H2 and electric cars are great to but also still in this development stage. As much as we'd all love to live in the future we don't.

And yes it sounds a bit silly and it was only to prove a point but in all fairness all heat recover systems work on this principle, reusing energy lost in the form of heat to reduce the amount of energy required by something else. In either case 6 months out of the year here that extra energy isn't exactly wasted.

 

[der_wille_zur_macht]

The cost analysis is out of whack as well.

Perhaps I did not make myself clear - the cost analysis I provided was real and based on actual plans for my specific 360g tank. It's based on DIY because that's what I plan to do. And, to be fair, a majority of LED users are DIY'ing their rigs right now - LEDs certainly have more appeal from the perspective of someone likely to DIY, which IMHO makes my DIY comparison quite valid.

Of course, you or I can make up a cost analysis to prove anything we'd like, but that DOES NOT change the FACTS:

1) LEDs, themselves, last a LONG time. You also seem concerned about other components failing. FWIW, the popular Meanwell ELN LED drivers have an MTBF of 603,000 hours per MIL-HDBK-217 at 25C ambient. Bought today, one of them would literally have a good chance of outlasting it's owner.
2) LEDs use less electricity than comparable reef lighting fixtures, and require no regularly scheduled lamp replacements. Generally this means they're cheaper in the long term.

I'm also concerned that you missed something I noted in my first post. Your cost comparison assumes that the LED fixture will be replaced at 10 years, and elsewhere in your posts you've made comments about an implied 10 year lifetime. That's simply not the case. The fictitious "10 year lifespan of LEDs" that people often quote is based on a misinterpretation of the 70% lumen maintenance lifespan most manufacturers publish, which is 50k hours - at that point, they're not DEAD, they're just 30% less efficient than they were on day one. At 10 hours per day, that equates to more like 13 - 14 years - and it's under pretty rigorous conditions. FWIW, an MH lamp reaches that point of lumen maintenance after at most, a few thousand hours of use. The difference is, you have to replace the MH lamp at that point. With an LED fixture, you can just turn up the current and it'll still likely be more efficient than other forms of lighting. Or, worst case, replace the LEDs themselves and keep using your (603k hour MTBF) drivers for another 10 years.

Also, you're using a worst case scenario for the LEDs (assuming the entire fixture must be replaced) vs. assuming that no extra maintenance will be required on the MH rig. IME over almost two decades of this hobby, that's not going to be the case - any system is going to have SOME components fail.

An LED fixture needs fans running all the time to cool the heat sink so the additional operating costs for cooling could be higher for the LED.

A few fans will consume 4 or 5w each. I really don't see how that is a factor in overall system efficiency or power budgeting, by at least one order of magnitude.

You can't argue that the chances of failure aren't greater for the LED.

Chance of failure isn't something to be argued, it's something to be tested, documented, and then looked up by the consumer/designer of the finished product. Same with cost of failure.

And at any rate, even if individual LEDs failed sooner than individual components in an MH rig, the cost of replacement is much less. I could eat three or four meanwell ELN drivers before I hit the replacement cost of a single electronic MH ballast. And since that's about the ratio you'd find them in a finished system (three or four LED drivers supporting the same light output as one MH ballast), the LED drivers could have a far worse MTBF (which they don't!) and still be more cost effective.

specifying something will last 10 years but only providing a single year warranty is a huge conflict.

Nothing has been claimed as 10 years of life. See above. Also, it's important to differentiate between statements made by a component manufacturer (LED lumen maintenance of 50k hours, or driver MTBF of 603k hours) vs. a warranty provided by the vendor of a finished product. Warranties made by product vendors are based on the vendor's approach to customer support and the implied durability of the object (is it considered consumable by the customer?), not just on projected lifetime - it's as much a marketing statement as a real projection. Meanwhile, MTBF and lumen maintenance projections made by manufacturers are based on experimentation and published standards.

On the same note, you cost comparison assumed that NOTHING in the MH rig would fail over a 15 year lifetime - can you reference an MH fixture that has a 15 year warranty? I'm not claiming an LED fixture will go 10 years without anything failing, FWIW. What I will claim is that component failure is not a differentiating factor among these technologies, so it's a moot point.

As I said above, I'm happy to discuss this topic at length, and I hope it doesn't appear that I'm claiming LEDs will save the universe or mow your lawn. They're certainly not mature as a technology, to be fair. Please feel free to tell me I'm being a nit-picky jerk, because at this point it's probably justified. But, the other side of the coin is that LEDs are being misrepresented in this thread, and I'm just doing my best to set the record straight based on published data and my own experience.

 

[sphelps]

Honestly at 8 bucks an led and the fact that i wouldn't bet more than 1-5 leds would be failing per year...i've had an led flashlight for over 3 years and have never ever had to replace the leds in it...and Cree has far higher standards for their Leds than the flashlight i have does...

Believe me buying 120 bucks worth of MH bulbs every year plus the energy they consumer and the heat they put off are 3 reasons to go with Leds

Yes if you build your own fixture this it might be easy and cheap to change out bulbs but I'm generally not talking about DIY stuff here. As a sustainable product in this application you have forget DIY and you can't assume everyone is willing to piece together there own fixture. If you purchase a fixture and a bulb fails it won't be that cheap to replace, and it won't be as simple and replacing a T5 or halide.

But let's even say they are equally easy to replace. If we assume the rate of failure is the same for LEDs and MHs and say that 10% will fail within 5 years. Comparing the previous scenario will result in 1 halide bulb failing (2 bulbs per year, 10 bulbs total) and 9.6 LEDs failing. @ $75 for a MH bulb that's $75 and @ $8 per LED that's $77.

 

[der_wille_zur_macht]

Yes well when costs are more reasonable LED lighting will make much more sense but that hasn't happened yet. They also need to be verified in application to meet the given specifications.

What sort of verification is missing? There are dozens of well-documented, active threads on this forum of successful tanks with LED lighting, and evidence to suggest that the total number of users is at least in the hundreds. It works fine.

 

[der_wille_zur_macht]

As a sustainable product in this application you have forget DIY and you can't assume everyone is willing to piece together there own fixture.

No one's trying to argue that LEDs are currently a "sustainable" product, and we're not trying to force DIY repairs on anyone. A question was asked about how "good" LEDs are. We're trying to answer based on our own experiences. Since most current LED use is DIY, that's the perspective we're bringing to this thread. Apologies if any more was implied. Back in my very first post, I commented that I AGREE that LEDs are not yet commercially viable to the average consumer.

But let's even say they are equally easy to replace. If we assume the rate of failure is the same for LEDs and MHs and say that 10% will fail within 5 years. Comparing the previous scenario will result in 1 halide bulb failing (2 bulbs per year, 10 bulbs total) and 9.6 LEDs failing. @ $75 for a MH bulb that's $75 and @ $8 per LED that's $77.

FWIW, HP-LEDs haven't cost $8 in years. Even the best there is (XP-G R5) can be had for around $6. Run of the mill Q5 XR-E are more like $4.50.

 

[Skeptic_07]

I'm not sure why you're so uptight. Never did I say LEDs run as hot as MHs nor did I say a chiller is the same as a computer fan so with all due respect perhaps you should read my post again and don't quote individual lines out of context for the sake of argument. Using a Metal Halide doesn't guarantee the need for a chiller, more than likely those systems also use high heat pumps and have closed canopies and may even require a chiller if an LED light was used. Good metal halide fixtures are made from aluminum and don't even have fans built into them, like mine. We may not get too many hot days but it still gets close to 40 degrees C here and all I need to keep my tank at 80 is two small computer fans, the same as you might find on an LED fixture. My house also doesn't have AC which is typical in your hotter climate areas experiencing "hot months".

I apologize for being uptight but when I see people going around posting incorrect information, it just irks me. Very little of what you have posted about LEDs has been accurate.

 

[wdt2000]

Most people probably have not heard of this yet but Cree is actually able to get ~160 lumens per watt now. This is in a laboratory, and I'm pretty sure its not yet available to the public, but this just further illustrates what DWZM was saying about how fast this technology is improving. See article below.

http://www.cree.com/press/press_detail.asp?i=1227101620851

208 lumens per watt for a white power LED
http://www.cree.com/press/press_detail.asp?i=1265232091259

 

[sphelps]

What sort of verification is missing? There are dozens of well-documented, active threads on this forum of successful tanks with LED lighting, and evidence to suggest that the total number of users is at least in the hundreds. It works fine.

Nothing to suggest lifespan and reliability of actual fixtures (not DIY) over the so called 10 year life span. Yes the manufacture of the LEDs will specify so many hours which equates to 10 years or 13 or whatever. However metal halide manufactures also specify a lifespan in the range of 10,000 to 15,000 hours before a 30% reduction in output which is 2.5 to 4 years with 10 hour days so why in our application this is reduced to 1 year. There are always variables and one must always use common sense when looking at studies over the internet, this article for example could be used to say MH bulbs can last much longer than one year and in fact can even get more efficient over time:
http://www.advancedaquarist.com/2008/7/aafeature1/view?searchterm=None

And yes there was some serious assumptions in the previous post but those assumptions where carried to both MH and LEDs. And yes your cost analysis works for your application but what I meant is it's out of whack for the average application and is a little bias towards your side of the argument.

I'm not trying to discredit LEDs either, but people need to settle down and stop worshiping them, they are not the holy grail.

 

[Skeptic_07]

208 lumens per watt for a white power LED
http://www.cree.com/press/press_detail.asp?i=1265232091259

woah! :eek1:

 

[sphelps]

I apologize for being uptight but when I see people going around posting incorrect information, it just irks me. Very little of what you have posted about LEDs has been accurate.

Like what?