Updating the WPG rule (theory)

[Tong]

I have been doing a lot of lighting research on my own and czcz mentioned I should post my findings here. First of all, I completely agree with everything Wizard~Of~Oz and czcz have previously stated.

Their research on lumens/watt seems to have all been done on 4' tubes. However, the same bulb at different lengths produce different lumens/watt. This is true for all bulb types. Take for example the 48" T12 GE Chroma 50 (Sunshine) bulb. It produces 2250 lumens with 40 watts giving it 56.25 lumens/watt. Now take a look at the 24" T12 GE Chroma 50 bulb. It produces 875 lumens at 20 watts giving it 43.75 lumens/watt. The pattern is that the smaller the bulb size, the lower the lumens/watt number eventually becomes.

This means that unless you have 4' tubes, you simply can not add up your wattage and multiply it by your bulb types conversion factors they have found. Instead, you must use the rated lumens in your calculations.

Formula: Total Lumens / Surface Area / 5.63

(5.63 is Wizard~Of~Oz lumens/inch^2 to achieve 1 WPG)

Example:

Total lumens: 620 (1 GE Chroma 50 18" T8)
Surface Area: 200 (Stadard 10 gallon)

620 / 200 = 3.1 lumens/inches^2
3.1 / 5.63 = 0.55 WPG

I bet all you 10 gallon people thought you were getting more light off that single strip. And I'm only going to make it worse for you.

Has anyone ever considered ballast factor? In Wizard~Of~Oz's findings he assumes 1.0 ballast factor. However, most magnetic ballast have a ballast factor of 0.94 while most electronic ballast have a ballast factor of 0.88. Wizard~Of~Oz assumed the WPG rule was developed over a 55 gallon take using 4' T12 bulbs. I'm going to add to this by assuming they were using average magnetic ballast. In his example he uses a 4' T12 bulb rated at 2350 lumens. With a magnetic ballast, this results in 2350 * 0.94 = 2209 lumens or 2209 / 40 = 55.23 lumens/watt. This means that 1 WPG (55 watts) of this lighting gives us 55 * 55.23 = 3037.38 lumens or 3037.38 / 576 = 5.27 lumens/inch^2 (576 is 55 gallon surface area in inches^2).

So our final formula will be Total Lumens * Ballast Factor / Surface Area / 5.27

Example using previous 10 gallon tank:

620 * 0.88 (Electronic ballast) = 545.6 lumens
545.6 / 200 = 2.73 lumens/inches^2
2.73 / 5.27 = 0.52 WPG

Is 0.55 WPG and 0.52 WPG really that big of a difference? No. But it does play a bigger part on larger tanks with more lighting. That and some ballast have really extreme ballast factors. I have seen some as low as 0.6 and as high as 1.2. In those cases you may want to consider this approach. In most other cases the first aproach should be fine and is simpler.

For a smaller tank you know you already have less light than what the WPG rule gives. I just want to make the point that you have even less light that what you previously thought.

I hope this wasn't too confusing and that it helps.

Edit: More on ballast factors. Most electronic ballast have a ballast factor that changes depending on how many of the leads you use. For example an electronic ballast with 4 leads is rated at 0.88. However, when only 3 leads are used, the ballast factor is actually 1.0. With only 2 leads used it becomes 1.12 and with only 1 lead use it is 1.24. This is normal and is similar to overdriving lights. Just something to think about.

 

[czcz]

Attached is the new standardized wpg described by Tong for all the tanks in Ozz's survey. What do you think? Do you find this calculation holds true for your tank?

I really like the calc he described above and think it is a big leap forward for the concepts in this thread, fwiw.

Here is a temporary link to my updated Excel compatiable calc: It is like the one linked earlier in the thread and here, except it includes Tong's standardizing wpg using Ozz's standardizing lumens/sq in. It does not account for the lower lumens/watt rating of small bulbs.

--

Tong,

What should we call the resulting wpg from your formula? Some name to reference it while noting it is not the same as bulb watts/gallons. I can't come up with anything better than "Tong's wpg"

 

[Tong]

Tong's WPG...haha that would be so vain. I've kind of been calling it Effective WPG or something along that line.

 

[Purrbox]

Looks good czcz, however I think that we'll need take account of the lower lumens per watt with smaller bulbs before I'd be able to make any judgement on the accuracy of the calculation as far as my tanks. The numbers for the 5 Gallon Hex feel right on, but the numbers for the 2.5 Gallon and 5.5 Gallon definately look too high.

 

[Tong]

Hey! I have 4.8 WPG in my 10 gallon! Woot! Okay I know better than that. Yes, we are going to have to take into account smaller bulbs if we want any sort of accuracy. I know 18" T8s actually get about 41 lumens/watt. So my 10 gallon is actually more like 2.2 WPG. And that to me feels about right.

czcz mentioned it another post somewhere, but I would bet when we figure out all the light loss for smaller bulbs, we'll conclude that CF lose the least and are therefor the way to go for smaller tanks.

 

[MarkP]

So, I'm looking over this new equation and applying it to my current setup. While double checking the numbers, I'm not certain that this works for all lights.

Specifically, I'm looking at my F30T8 bulb. From http://www.gelighting.com/na/busine...ibrary/catalogs/downloads/cat_fluorescent.pdf (as well as various other sources) I see F30T8 listed as approx. 2000 Lumens (The source I'm quoting actually as the F30T8/D as 1850 Lumens).

The spreadsheet referenced shows 30 watts of T8 Lighting as putting out 2733 Lumens.

I can think of two explanations. The first is that since lumens is the amount of light in a given area, GE and others might be mearsuring lumens without using a reflector. Unfortunately, I'm not sure this is true.

I don't see any other datapoints in this thread referencing F30's, but I do see F32T8 referenced as putting out between 2800 and 3100 Lumens, giving about 90 Lumens/Watt. The GE Specsheet referenced above gives similiar numbers for their F32 bulbs, leading me to suspect they have accurate for the F30 as well.

It appears that the length of your lamp has a huge affect on the Lumens Per Watt (LPW).

From GE Specs- (Initial Lumens, using numbers from 6500K Temperature Bulbs when availible, noted when not, above referenced PDF is source)


F32T8/SP65/ECO (4 Foot)

2700/32 = 85 LPW

F30T8/D (3 Foot)

1850/30 = 61 LPW

F15T8/D (1.5 feet)

700/15 = 46 LPW

F13T8/CW (1 foot, New Color Temperature, 4100K, no 6500K info)

565/13 = 43 LPW

This seems to indicate that efficency drops significantly with the size of the bulb- Perhaps explaining why 4 WPG of T8 light over a 4 foot tank is extremely high light, and 4 WPG of T8 light over a 12 inch 10 gallon isn't?

I don't beleive this point was previously raised, but am currently unable to access some of the links to other articles- But it would seem to argue against a constant for T8 Lumens per Watt, which I had seen previously quoted.

 

[Tong]

Their research on lumens/watt seems to have all been done on 4' tubes. However, the same bulb at different lengths produce different lumens/watt. This is true for all bulb types. Take for example the 48" T12 GE Chroma 50 (Sunshine) bulb. It produces 2250 lumens with 40 watts giving it 56.25 lumens/watt. Now take a look at the 24" T12 GE Chroma 50 bulb. It produces 875 lumens at 20 watts giving it 43.75 lumens/watt. The pattern is that the smaller the bulb size, the lower the lumens/watt number eventually becomes.

This means that unless you have 4' tubes, you simply can not add up your wattage and multiply it by your bulb types conversion factors they have found. Instead, you must use the rated lumens in your calculations.

It does not account for the lower lumens/watt rating of small bulbs.

The issue was raised. We just don't have enough data yet to come up with a conversion number for every bulb length. In the mean time, if you know the lumens of your bulbs, then it is preferable to use those versus the actual watts (there are some exceptions to this). We use watts for simplicity because not everyone knows the lumens of thier bulbs but almost everyone knows the wattage of thier bulbs.

 

[MarkP]

The whole time I was writing that, I was thinking it HAD to have already been addressed, but couldn't find a mention of it. Why? Because I didn't check page 4. :-(

Ok, then, lets take the next step. Using the data points provided by GE in the above referenced PDF, we have

4 Foot T8 Multiplier- 1

3 Foot T8 Multiplier- .71

1.5 T8 Foot Multplier- .54

1 T8 Foor Multiplier- .50

This only gives us a single data point for each number, so these numbers would be considered, at best, as *extremely* rough.

Taking a 45H tank with a single F30T8 bulb (36 inches by 12 inches). Spreadsheet gives you Tongs WPG of 1.12. Multiply that by the (in)efficentacy multiplier of .71 for 3 foot T8 gives you .79 Adjusted Tongs WPG. (Numbers chosen from my tanks setup when I first purchased it from a yard sale).

Additional Data Points for F30T8- Sylvania, Grainger, and Phillips all give me 1850 Lumens for a F30T8/D Bulb (6500K). 1000Bulbs.com gives me 1500. http://www.donsbulbs.com/cgi-bin/r/b.pl/f|30|t8|765.html lists it as 1900.

Perhaps the best course would be for me to stop by Home Depot today and check to see if the bulbs there are labeled with Lumens.

Or, perhaps the best course would be to buy 4 foot tanks.

 

[czcz]

Ok, then, lets take the next step. Using the data points provided by GE in the above referenced PDF, we have

4 Foot T8 Multiplier- 1

3 Foot T8 Multiplier- .71

1.5 T8 Foot Multplier- .54

1 T8 Foor Multiplier- .50

I like this, and looking ahead, in my mind it will be easier and more intuitive to write calcs that allow the user to select bulb length from a drop down menu or something. Thoughts?

Additional Data Points for F30T8- Sylvania, Grainger, and Phillips all give me 1850 Lumens for a F30T8/D Bulb (6500K).

IMO this meets the sample size requirement. Anyone have smaller wattages and bulb types, including source, etc? Maybe looking at and sorting the data will keep our minds churning.

BTW, Ozz has CSV available and I have plain text for most data, 'case anyone wants it.

--

Only some points for others who question by nature. Here is a thread off-site that includes valid arguments against lumens, and therefore any standard derived from lumens, as well as a response from Ozz. It is important to remember the realistic goal of this thread is a better guess, and WPG is simply old and broken. Guidelines help but as we know nothing is written in stone.

For the resulting "Effective WPG" to hold any water, it has to pass the smell test. To accomplish this we not only need large samples for the bulbs, but also your tanks. Please submit data to Ozz's survey.

 

[Tong]

Here is a thread off-site that includes valid arguments against lumens, and therefore any standard derived from lumens, as well as a response from Ozz.

Agreed. One case I know of that is a bad use of lumens is with plant and aquarium bulbs. These usually only have a fraction of the lumens as daylight bulbs but IME, grow plants just as well.

Keeping a color temp of 6500K constant is a good thing but remember to keep the CRI ratings fairly close as well. A bulb with a CRI of 78 will produce more lumens that a bulb that has a CRI of 98. Yeah...just thought I'd through another monkey wrench in there for you.

 

[czcz]

F32T8/SP65/ECO (4 Foot)
...
F30T8/D (3 Foot)

Can someone please explain watts/length of T8? If 4ft T8 is 32w, why is 3ft T8 30w instead of 24w? Why do different T8 manufacturers list different watts for the same length bulb? A T8 should be a T8 in terms of watts, right?

Is it fair to say many sites use the 10watts/ft of T12 for T8 in error, simply because T8 fits in a T12 fixture? Is it really 8.x watts/ft?

If so lumens/watt still changes dependent on bulb length, but it doesn't break as badly as before. If a 36" T8 is 24w and 1850lumens, it is 77.1lumens/watt, only ~10% less than the 4ft T8, for example.

 

[Tong]

Can someone please explain watts/length of T8? If 4ft T8 is 32w, why is 3ft T8 30w instead of 24w? Why do T8 manufacturers list a watts for the same length?

Oh yeah, I forgot about that. Yes, T8 wattages are kind of erratic. Most common are 4' = 32 watts; 3' = 25 watts; 2' = 17 watts; 18" = 15 watts; 16" = 14 watts.

I've never looked at the 3' and 4' bulbs but I know the 2' bulbs will range anywhere from 15 watts to 20 watts.

If so lumens/watt still changes dependent on bulb length, but it doesn't break as badly as before. If a 36" T8 is 24w and 1850lumens, it is 77.1lumens/watt, only ~10% less than the 4ft T8, for example.

Are you sure? Because if a 4' bulb uses 32 watts then an 18" bulb should use 12 watts. But it doesn't...it uses 15 watts.

620 lumens / 12 = 51.67 lumens/watts

vs.

620 lumens / 15 = 41.33 lumens/watts.

 

[Purrbox]

For some of the brands the number of watts is equal to 10 x the length in feet of the bulb. AFAIK this is the highest wattage that you'll find on any of these bulbs.

 

[czcz]

Tong, I used 8.x watts/ft because 24" T8 is sometimes listed as 17w or 18w. I am not sure at all. I should have clarified: if lumens/watt stays fairly constant at, say, 3ft-5ft, I don't think we should care about those samples so much yet. I agree samples in the lower lengths (typical for, say, a 10-20gal) are far more useful and interesting to play with for your points.

 

[Tong]

Agreed. I've also found that the loss for 3' tubes is a moot point.

 

[CHBGator]

Thought I would throw this into the topic as well if we are actually getting more light than what we thought because of the newer technologies then this might also have a place here to put even more dirt on the grave of how much light we are getting and whats really needed. Now if someone can work out an excel spreadsheet to combine the 2 formulas (what light we are actually getting and whats actually needed). That is of course if both thoughts are correct.

http://www.fitchfamily.com/lighting.html

 

[Tong]

Yes, I've seen this before. Funny thing is though is that when you work the the formulas by hand taking into account light loss for smaller bulbs and such...our numbers come fairly close to the numbers Amano recommended. Assuming the numbers they give is the wattage needed for high light.

 

[czcz]

Also, the fitchfamily analysis is awesome and arguably the best support for wpg breaking in small and large tanks, but it still assumes a watt is a watt. It does not address bulb efficiency (CF vs T12 vs T8...).

Earlier in this thread CGGorman noted that the Amano books do not list bulb type.

I checked my Amano books. They don't specify lighting type. Just wattage and number of lamps.

 

[CHBGator]

So.... can an excell guru write up a quick spreadsheet to combine the 2 ?

 

[MarkP]

My understanding is that the F30T8 really does use 30 watts. There is also a F25T8 bulb, that follows the 8 Watts Per Foot rule.

The F30T8 differs electrically from the F25T8. The F25T8 operates at a higher voltage, but draws much less current, giving you a lower total wattage.

The F30T8 is the older style, pre-heat bulb. The F25, and I beleive the F32, should be a rapid start.

Some of my information is from The Krib, and confirmed from a few lighting forums. I wondered the same thing when I saw a F30T8 bulb.

I do not beleive F30T8 vrs F25T8 is "shorthand" or leftover from T12 days. In the above referenced PDF, GE lists the F25T8/XL/SPX65/ECO with Initial Lumens of 1950 and color temperature of 6500k. The F30T8/D has initial lumens of 1850, with a color temperature of 6500k. In addition, Grainger, and Phillips offer both F25T8 and F30T8. In addition, I beleive that there is a tax break for changing out older F30T8 fixtures for newer, F25T8 ones.

Czcz- I was actually thinking of your spreadsheet, but instead of one entry for each light type, a line for 4 foot lights, 3 foot, etc. Then I thought it would get too crowded.
I could probably cobble together a VBS application that could do it, but I've never done a graphical front end, it would just be text in a DOS box, at least initially.

Edit- I've just realized, I'm even further off base then I thought.

F30T8 isn't the same fixture as the F25 and F32 T8's, and yet I've been including it under the same calculations. I'm seeing a huge difference from F32T8 to F30T8, others are saying that "3 foot T8's are similiar efficency".

Apples and Oranges.

On an unrelated note- The guy on the linked forum who challanged Lumens as a mearsurement- He's got a point, but his suggestion of Watts of Radiant Flux isn't any better. That includes any energy- Infrared, X-ray, etc. Black Lights would score very high on that scale, yet not grow many plants.

What we REALLY want is a Mearsurement of Photosynthetic Flux- That is, the mearsure of Radiant Energy that is usefull to plants- Lumens for Leaves instead of eyes, if you will. I haven't a clue where to look for it- Maybe Hydroponic literature?