LEDs are NOT a cost effective way to light a reef... (at least not for me)

[VJV]

Hi, this is my opinion and based on my observations, so it may not be an universal truth for every tank out there. But in mY opinion, the whole argument that LED's are a more cost efective option in the long run is not accurate. The problem with LEDs is that manufacturer's grossly overestimate the actual coverage of their fixtures, leading you to believe that too can get away with just two for the typical 4 foot tank. This is simply not true and if you follow this recommendation you end up with massive shadowing and uneven coverage that leads to coral colonies dying from the bottom up. I am not saying LED's can't grow corals. Of course they can. You just need to have more fixtures. MUCH more, than the manufacturer recommendation.

So in my case I have a 80 inch long tank. To provide adequate coverage using a popular fixture such as the RADION I would need at least 5 fixtures (I would actually say 6 but do not want to stretch it) to provide adequate coverage of even light and minimize hotspots. This would compare with a single 72inch Giesemann Spectra (3x250w MH and 4x80w T5)or two ATI sun powers 8x39w (because of the length of my tank 2 ATIs would provide better coverage than one single 8x80)

The tables below highlight the average running cost per year of each unit and how long (in years) would it take to recover the initial investment in the LED fixtures. Under these assumptions as you can see it would take 8,2 years of savings to recover the initial investment of the Radions vs the Spectra, and almost 14 years of bulb replacements in the case of the ATI ;Woot.

I can already hear some saying "oh, but you should also factor in the cost of a Chiller"... should I? I mean, won't 950w of LED produce as much heat as 1070w of MH+T5?... I am honestly not an expert so although my guess is that it should, I have nonetheless run the calculations adding the chiller cost to the Giesemann Spectra.

In this scenario it would still take me 6 years to recover my initial LED investment vs the Spectra, and 10y vs the dual ATI T5 fixtures. I am not adding the electricity cost of the chiller because on the other hand I would also need to add the cost of the heater to the LED options.

So, at least to my tank, LED is not a cost saving option, quite the opposite.

In all fairness, puck style fixtures like the Radions will never be good value for money if your objective is to provide a even blanket of light vs a high PAR hotspot. There are other alternatives out there wide large arrays of LEDs such as the Philips CoralCare or the Orpheck Atlantik fixtures. Either of these would be much more cost effective given I would need much less units to provide coverage for my tank. At €2.850 4 Philips Coral Care units would be significantly less expensive upfront, and recovery much faster.

Happy to be called out if the calculations are not right (quite possible as my 4th kid was born just 3 weeks ago ;Singing) but the whole "efficiency" argument for led's simply does not add up.

 

[mcarroll]

Congrats on the newb to the family!!!

Did you compare your calc's with any that others have done in the past? Some have even counted actual power use. I'm still getting to your PDF's, but your results are unusually close compared to others where I've been able to "check work".

 

[mcarroll]

Oh! What are the tank's actual dimensions out of curiosity?

 

[SashimiTurtle]

The heat source with LEDs are not the diode, but the drivers for the diodes. 99.9999% the heat is being emitted behind the light and if the light is designed properly, the fan and heat sink will direct this up and out, away from the tank. I'm not saying there is no heat generated by the diodes, it's just negligible compared to MH and even T5.

 

[mcarroll]

How often are you changing bulbs? Once per year across the board?

 

[saltyfilmfolks]

It cost me 450 to light my 48" tank. I have too much coverage and too much power. I'm at 1% and 20% on 3 165w fixtures.

Corals grow great.

My cube was 250 I clould have gotten away with spending 125 for code to same thing.

But I opted for a brand name and an unnecessary feature. Ramping.

Fwiw
The heat (IR) generated by a mh t5 is radiated from the souce. Leds are not. So 600 or 1000 par generated by an led will not have the hear due to the narrower light spectrum.
Led give out very little in the spectrum of 700+Nm. IR.

 

[ndrwater]

In my very unscientific way, I can say LED's ARE cheaper in the long run.
I used to run MH+T5 on my standard 240. 96x24x24. I had 3 250w Halides and 8 39w T5. In that scenario, a chiller was a must for me. At my peak, not counting the annual $400+ bulb replacement costs, I was spending $800 a month on electricity. Keep in mind I live in Anaheim California which is some of the cheapest electric costs in the country.
After doing the math and spending the $3500 for my 5 G3 Pro's and deleting the no longer needed chiller my electric costs dropped to about $200 a month. Everything else stayed the same.
Doing some simple math, $3500 initial cost divided by $600 a month savings, the LED's pay for themselves in 5.83 months. This still doesn't account th $400 annual savings by not having to replace bulbs.
With the 5 Radions I have running front to back, spaced 10" apart, I get more than adequate coverage and my corals grow like weeds.
I am happy with my decision and have looked back.

 

[Randy Holmes-Farley]

Just to add to the complexity of determining relative costs, one might also need to understand what is happening to the waste heat in each setting.

In my home, we never air conditioned near the tank, but we did heat both the tank water and the house for much of the year. So the "waste" heat from the mh lights was useful more than half of the year, neutral for a couple of months, and probably required some extra cooling water use (I cooled my tank in mid summer with a tap water cooling system).

 

[Lasse]

The heat source with LEDs are not the diode, but the drivers for the diodes. 99.9999% the heat is being emitted behind the light and if the light is designed properly, the fan and heat sink will direct this up and out, away from the tank. I'm not saying there is no heat generated by the diodes, it's just negligible compared to MH and even T5.

Its because MH and in some extend T5 produce huge amounts of IR radiation (700 nm and over) that directly heating the water. If you do not use phosphorous coated LEDs - the IR radiation is zero. If you use low Kelvin LEDs you have a small amount of IR.

The two spreadsheets are the same - there is no calculation of chillers in the second.

Sincerely Lasse

 

[mcarroll]

Pick Of Fixture
First, as you implied, this LED is among the least favorable to compare with. A simple pair (or two; I don't know tank size) of Orbit Pro's might be a better comparison...about $88/year in power per set, IIRC. (The differential cost of a DIY would be so large as to make the discussion not worth having.) But the Radion still usually holds up in the numbers.

Keep It Cool
If your house is cool all the time (i.e. no chiller to eliminate or minimize), then your cost savings from power might be close to zero as the tank has a heat budget that needs to be met to keep it tropical. Cutting heat input from things like pumps and lights can end up doing nothing more than triggering more ON-time from your heaters. (I had to add more heaters for wintertime, in fact.)

An Example
Check out Post #67 from some lighting thread that I posted in. Sorta similar example.

Minor Criticism
It looks like you're using watts pulled from the wall for your halides but nominal watts for the T5 and (I think) G4. This throws off the comparison...ideally all would be sampled from the wall, but without that data I think I'd revert the halides to using nominal watts as well.

 

[VJV]

Congrats on the newb to the family!!!

Did you compare your calc's with any that others have done in the past? Some have even counted actual power use. I'm still getting to your PDF's, but your results are unusually close compared to others where I've been able to "check work".

Thanks! Not wanting to sound like a character from the Mercenaries movie, I'm getting too old for this ;Dead.

I did not compare to others. I only used the power ratings and I know this is not 100% accurate, but I believe it is a good proxy given the monumental upfront cost differences.. which leads me to another question: why do you say the calculations are close? In what regards, wattage? Because in terms of cost the difference is huge. It would take 8 years to make the Radions cost effective vs the Spectra unit, or 6 if I add a chiller.

 

[VJV]

Oh! What are the tank's actual dimensions out of curiosity?

80x24x24 inches.

 

[VJV]

The heat source with LEDs are not the diode, but the drivers for the diodes. 99.9999% the heat is being emitted behind the light and if the light is designed properly, the fan and heat sink will direct this up and out, away from the tank. I'm not saying there is no heat generated by the diodes, it's just negligible compared to MH and even T5.

Thanks for posting! This piece of technical info is very useful. Which means I really need to add the chiller to the equation I guess. Even when I do, the recovery time to make the RADION cost effective would still be some 6 years (and no one keeps their LEDs for that long ). So you theoretically put about the same level of heat into the room, just not directed towards the tank.

 

[VJV]

1 year yes.

How often are you changing bulbs? Once per year across the board?

 

[VJV]

It cost me 450 to light my 48" tank. I have too much coverage and too much power. I'm at 1% and 20% on 3 165w fixtures.

Corals grow great.

My cube was 250 I clould have gotten away with spending 125 for code to same thing.

But I opted for a brand name and an unnecessary feature. Ramping.

Fwiw
The heat (IR) generated by a mh t5 is radiated from the souce. Leds are not. So 600 or 1000 par generated by an led will not have the hear due to the narrower light spectrum.
Led give out very little in the spectrum of 700+Nm. IR.

Thanks salty. I guess at 150$ per 165w fixture (I'm assuming LED here) these are some black box units? There are also 8bulb T5 units on Amazon for 100$ or so but to me at least those would not be an option. My comparison (and this is just the result of my decision process, not meant to be science) is based on comparing upmarket alternatives of both technologies.

The heat issue is important to me though. Using the Philips CoralCare fixture and comparing with MH+T5 the numbers change very significantly. I still like the simplicity of having one light fixture hanging over my rimless peninsula tank but on the other hand do not like the added complexity of adding a chiller to the system...

 

[VJV]

In my very unscientific way, I can say LED's ARE cheaper in the long run.
I used to run MH+T5 on my standard 240. 96x24x24. I had 3 250w Halides and 8 39w T5. In that scenario, a chiller was a must for me. At my peak, not counting the annual $400+ bulb replacement costs, I was spending $800 a month on electricity. Keep in mind I live in Anaheim California which is some of the cheapest electric costs in the country.
After doing the math and spending the $3500 for my 5 G3 Pro's and deleting the no longer needed chiller my electric costs dropped to about $200 a month. Everything else stayed the same.
Doing some simple math, $3500 initial cost divided by $600 a month savings, the LED's pay for themselves in 5.83 months. This still doesn't account th $400 annual savings by not having to replace bulbs.
With the 5 Radions I have running front to back, spaced 10" apart, I get more than adequate coverage and my corals grow like weeds.
I am happy with my decision and have looked back.

Thanks for posting. The heat issue also depends on where you live. In places where you need to heat the tank using LEDs means you will end up having to also have 600w heater to heat the water which you can do without with MH/T5. In my case that happens like 50/50 during the year. But to your specific case, in my area and assuming running the MH/T5 unit full tilt for 8h plus chiller (TECO TK2000) would spend some 1500 watts per hour or 12.000 watts per day (assuming the 8h above). At 0,1659 cents per KW/h (again, this is in my area) the monthly electricity running cost would be some 60$ so one tenth of the power savings you have achieved.

 

[VJV]

Just to add to the complexity of determining relative costs, one might also need to understand what is happening to the waste heat in each setting.

In my home, we never air conditioned near the tank, but we did heat both the tank water and the house for much of the year. So the "waste" heat from the mh lights was useful more than half of the year, neutral for a couple of months, and probably required some extra cooling water use (I cooled my tank in mid summer with a tap water cooling system).

Thanks for posting Randy. Precisely. That is why although I factor in the cost of buying the chiller I am not considering added electricity costs as I believe those are offset by the need to heat the tank if going with LED (in my area anyway, which I believe behaves in a similar fashion to where you live).

 

[VJV]

Its because MH and in some extend T5 produce huge amounts of IR radiation (700 nm and over) that directly heating the water. If you do not use phosphorous coated LEDs - the IR radiation is zero. If you use low Kelvin LEDs you have a small amount of IR.

The two spreadsheets are the same - there is no calculation of chillers in the second.

Sincerely Lasse

Thanks for posting! Well, I did include the chiller but only the upfront cost, and added to the cost of the Spectra. Without the Chiller the Spectra would cost me approx 1500€. I am adding to that 850€ from the chiller cost. Not adding the electricity cost of running the chiller because I would also need to factor in the cost of running a heater to heat the water in the LED option. So basically it only increases the upfront cost of the Spectra option.

 

[mcarroll]

@VJV I wanna try running your numbers, but until that happens....do you have a power meter like a Kill-A-Watt? You can eliminate quite a few assumptions from your calculations with a day or two of real-time monitoring.

 

[saltyfilmfolks]

Thanks salty. I guess at 150$ per 165w fixture (I'm assuming LED here) these are some black box units? There are also 8bulb T5 units on Amazon for 100$ or so but to me at least those would not be an option. My comparison (and this is just the result of my decision process, not meant to be science) is based on comparing upmarket alternatives of both technologies.

The heat issue is important to me though. Using the Philips CoralCare fixture and comparing with MH+T5 the numbers change very significantly. I still like the simplicity of having one light fixture hanging over my rimless peninsula tank but on the other hand do not like the added complexity of adding a chiller to the system...

Why would they not be an option if you say what you're looking at is not affordable?

It would probably cost me $100 bucks to custome a single fixture. 200 if I had it done for me.

Still under 1000 if I had a builder for the whole thing for me.