Google 'Sanjay Joshi' for articles on power consumption of various halide systems.
Some more casual searching suggests that most decent LED's produce about 100 lumens per watt, with a current record of just over twice that. Since a decent T8 will produce 100 lumens per watt, and T5 just a bit more than that, I'm still not sure what all the buzz is about. MH might produce about 85 lumens per watt, so it is perhaps not so much that LEDs are so efficient, but rather that MH is just not very efficient.
In addition to the "all lumens are not equal" comment evilc66 made above, you have to consider one big difference you're omitting is
where that light goes. Even if we assume a lumen from LED is equivalent to a lumen from T5, the LED puts that light out in a
useful manner - i.e. in a relatively small viewing angle that we can orient towards the tank with no losses. We can tighten that viewing angle even more with optics that will be more efficient than any current reflector available for T5 or MH.
So, even if an LED rig put out X amount of useful light per watt and a T5 rig put out the same amount, the LED rig would "win" the practical test because that light is naturally going in a useful direction. With the T5, we
HAVE to use a reflector, because the light is going in ALL directions. And, even the BEST reflector is still going to be lossy compared to NO reflector.
This is all good-natured ranting, BTW
. I'm pretty conservative when it comes to new technology -- this tends to gain me some enemies in this hobby
.
I'm a natural skeptic, so we'd probably get along well.
However, I'm also firmly convinced that LED lighting is not simply a fad trend. I'm basing this on my own experience with other types of reef lighting systems over the last 18 years, and on my recent experience with LEDs - PLUS the theory and math, which I'd gladly argue all day long if people wanted. IMHO though, despite being very new, LEDs have a golden mix - they're
actually working over people's tanks, AND the theory/numbers are there to back up their success. Hard to argue with something that's backed in theory
and in practice.
And, the results are there in a vastly wider respect than this hobby. My last employer had ~600 acres of buildings and parking lots. They were switching street lighting, conference room lighting, hallways, etc. from other technologies to LED lighting as fast as they could - in general, it's a proven technology for lighting that's not going to die out overnight, and the acceptance well beyond this hobby means that R&D budgets are going to be vast - we can just ride the wave.
The biggest problem I have with LEDs is the complexity. Yes in theory they are great and if they last for the rated 10 years or whatever then the price isn't bad either. However with so many components something is likely to fail prior to the system reaching it's lifespan.
LEDs are absolutely no different than any other lighting technology we're likely to use in this respect. Ever looked inside a T5 or (electronic) MH ballast? It's a very similar level of complexity as an LED driver. I've had MH and fluorescent ballasts fail in the past. I'm sure I'll have some LED components fail before their rated lifetime, too. No big deal. Sure, there may be some instances where certain early commercial LED fixtures used components that failed often, but that's no difference than the junky MH or T5 fixtures bought at bargain prices on eBay - which fail often! It's not inherently due to the technology, it's a case of poor implementation.
One thing is for sure - with an LED rig, I won't miss having to replace expensive lamps every 6 - 12 months - and the cost saved there will more than pay for component replacement from time to time.
I've seen multiple issues from current fixtures such as burnt out drivers and LED arrays, which usually have to be replaced in sets rather than singles.
Not sure if I'm following you there. If I have an LED burn out (I haven't yet, in a "finished" fixture over a tank), I will replace that single LED ($5). These things don't have to be replaced in sets. You replace what's broken, same as any other type of lighting.
The lifespan is off as well IMO, in the conditions these fixtures will be used should shorten the rated lifespan as it does with all lighting technology. I huge hint is the warranty these fixtures come with, most offer only 1 year for a system that's apparently designed to last 10 times that. That right there is clear evidence they simply won't achieve that lifespan.
Are you speculating, or do you have evidence to back up these claims?
All said an done, even with electricity and bulb savings, I don't see the potential for savings.
I'm building a 360g tank. Before I chose LEDs, I was going to use four 250w MH lamps and ~200w of fluorescent supplementation (say, four 4' T5s). Now, I'm going to have ~400w of LEDs. I'm pretty confident that these two rigs will produce comparable results in terms of coral growth. I have short-term results on smaller tanks to prove this, and others have longer-term results if you're interested in digging them out.
Let's run some numbers.
The MH/T5 rig will consume ~1400w of power (assuming 85% efficiency, which is probably generous). The LED rig will consume 460w (same 85% efficiency, which in this case is probably worse than I'd actually get).
Upfront costs:
1) MH rig: $800 total. $200/ea for MH lamps (reflector, ballast, socket, lamp). T5 rig: $250 total. Entire fixture: $1050.
2) LED rig. $2000 total. I can break out line items if you'd like, but this is based on an average DIY cost of $10/LED, for 200 LEDs. I've kept numbers from all the larger DIY builds reported on these forums and that works out to be conservatively accurate. My own personal build will be cheaper, but I'm "cheating" by DIY'ing things like drivers.
Some ongoing cost assumptions:
1) Assuming the national average of 12 cents/kW for electricity.
2) Lamp replacement for the MH/T5 rig will be $75 per MH lamp every year, and $100 for T5 every year.
3) 10 hours/day of illumination.
Now, let's figure ongoing costs. I'm planning a 10 year lifespan for my current tank, so I'll use that timeframe.
The MH/T5 rig will draw 14kW/h per day. 5110 kw/H per year. That's $613/year in electricity. Lamp replacement annually will be $400. So, $1,013 per year.
The LED rig will draw 4.5kW/h per day. 1643kW/H per year. That's $197/year in electricity. No lamp costs.
In summary:
The MH/T5 rig will cost $1,050 upfront plus $1,013/year, or $11,180 to own for 10 years.
The LED rig will cost $2,000 upfront plus $197/year, or $3,970 to own for 10 years.
The LED rig will be
$7,210 cheaper to own over 10 years. Even if you had to replace
THE ENTIRE THING from the ground up three times, and the MH/T5 rig NEVER broke or required additional repairs, the LED rig would still be cheaper.
Also what happens after the lifespan of the LEDs? Will it be cheaper to replace the whole fixture?
The "lifetime" figures you see most people quoting usually fall in the 10 year range. These figures are based on manufacturer estimates. Most manufacturers have estimated these LEDs will last 50,000 hours for 70% lumen maintenance. That means that after 50,000 hours of operation, they will give off 70% of their original light for the same power draw. So, the fixture isn't "dead" at that point - you just need to add 30% more LEDs, or turn it up 30% to compensate. That said, there will be component failures along the way, but again - same thing will happen on T5 or MH.
On another note, it's important to be as accurate as possible with these numbers. At 10 hours per day, 50k hours is more like 13.7 years. And, that 50k figure is calculated under conditions (drive current, die temperature, etc.) that are likely the same as or WORSE than what the average DIY LED fixture will see (most people are going WAY overboard with heatsinks). So, the "average" DIY LED rig will likely see 70% lumen maintenance lifetimes LONGER than 50k hours.
All things considered I don't believe LEDs are realistic at this time but do make for a cool project for the electronic junky.
That's a very valid perspective. I agree that they aren't "ready" in the sense that the current off-the-shelf products are where they should be. It's a very new technology in this application, and in a certain sense, the vendors selling finished products to us hobbyists haven't figured it out yet. Certainly, some of the best LED results are being had by experimenters who are putting a lot of time and effort into this, and in that sense, it's probably not yet a "mainstream" technology in the hobby. But, the same was once true of MH, T5, protein skimmers, artificial saltwater mixes, or marine aquariums in general.
Another perspective is that LEDs are rapidly improving right now, while MH/T5 are making only very small improvements, if any at all. Just a few months ago, 100 lumens/watt was about as good as it got for LEDs. Now, the Cree XP-G is available for about the same price as those 100 lumens/watt LEDs used to cost, and it is in the 130 - 140 lumens/watt range. Clearly, we are near "perfection" for how good MH/T5 will get - both in general and in this hobby, but LEDs are still in a phase of very rapid improvement.
I could go on for many pages about functionality that LEDs provide, but I've probably already lost people.
Suffice to say, I would be using LEDs even if they were more expensive, more failure prone, and less efficient long-term than other forms of lighting. They allow functionality that's just not there in other types of lighting, or at least not as refined:
1) Color selection when initially designing the fixture. Most "colored" LEDs are a very pure, very narrow wavelength of light. This, combined with the high LED count in most fixtures, means you can VERY finely tune your output when selecting components. When selecting MH or T5 lamps, you typically have a smaller number of choices which means less resolution. Many people use 50% cool white LEDs and 50% royal blues. If I want a little more midrange blue, I can swtich to 50% cool white, 45% royal blue, and 5% plain blue. That's a smaller "jump" than switching from a 10k MH lamp to a 12k, for instance.
2) Dimming. Some other technologies allow dimming, but it's arguably easier to implement and more widely available for LEDs - nearly ALL of the drivers people are using are available with dimming functions, whereas you really have to dig to find dimmable versions of T5 or MH products. What this means is I can program a $30 Arduino controller to ramp my LEDs up and down for sunrise/sunset, imitate cloud cover, seasonal variations, etc. But, it also means I can VERY precisely tune the color and intensity of the light without component changes. If my white LEDs are on one circuit and my blues are on another, I can adjust the color balance on a whim with a knob or a bit of code. Or, if I want a 6" by 6" section of the tank to be a little bluer because I'm putting a really cool SPS there that would really "pop" with bluer light, I can turn the blue LEDs in that part of the tank up a bit.
3) Precision of light placement/angle/direction. Want to spotlight a single coral or rock outcropping with slightly different light than the rest of the tank? No problem with LEDs. Pack a few more LEDs into that part of the fixture, or adjust the proportions of the different colors, etc. Want a "shaft of light" look at a steep angle across the tank during dawn/dusk? No problem. Put a few LEDs at a steep angle with tight optics.
4) Ease of altering the viewing angle/distribution of light. Want to put your light fixture in the ceiling 5' above the tank, but have
nearly zero spill outside the tank? No problem. Slap on some tight optics.
And so on. . . .
At any rate, it probably appears that I've severely biased, but I'm hoping this post adds some useful information to the discussion.
4400 KILL A WATT
