The Planted Tank Forum banner

Lighting an Aquarium with PAR instead of Watts

436K views 453 replies 170 participants last post by  jeffkrol 
#1 · (Edited)
Before any discussion of aquarium lighting can proceed, first we have to debunk the myth about “watts per gallon” being a measurement of light intensity.

When the only practical source of light for a planted tank was T12 fluorescent tubes, someone decided that the way to pick out the best lighting was to figure out the “watts per gallon” that were needed to grow various types of plants. This would make sense if we could pour a teaspoon of watts of light into a tank, and get a light concentration of X watts per gallon of water, just as we pour a teaspoon of potassium nitrate in the tank to get a nitrate concentration of Y mg per liter of water. But light is nothing like a chemical - you can’t pour it anywhere, you can only shine it on something. That alone should debunk “watts per gallon” as a measure of light intensity.

But, there is more: Let us assume we have two 20 gallon tanks, with 40 watts of T12 Fluorescent light on each one - 2 watts per gallon. One tank is a 20L and one is a 20H. The 20L tank is 12 inches high, and the 20H tank is 16 inches high. If the fluorescent light sits right on top of each of the tanks, the light on the 20H tank is 4 inches farther from the substrate - 33% farther from the substrate. Because light intensity drops approximately proportional to one divided by the distance from the light squared, the intensity at the substrate in the 20H tank has to be about 56% of that at the substrate in the 20L tank. That alone should debunk “watts per gallon” as a measure of light intensity.

But, there is more: Let us assume we have two 20H tanks, one with a 40 watt T12 light sitting on top of the tank, and the other with the same light hanging 12 inches above the top of the tank. Again, because light intensity drops approximately proportional to one divided by the distance from the light squared, the intensity at the substrate for the tank with the light hanging 12 inches above the top of the tank must be about 32% of the light intensity of the tank with the light sitting on top of the tank. And, that alone should debunk “watts per gallon” as a measure of light intensity.

So, that is three strikes against “watts per gallon”.

But, there is more: Let us assume we have two 20H tanks, one with 40 watts of T5HO light, from a Tek light fixture, the other with 40 watts of T12 light. Anyone who has looked at both a T5HO bulb and a T12 bulb, when they are lit up, knows that the T12 bulb can be stared at without distress, but the T5HO bulb causes some temporary blindness if you look at it for more than a few seconds. The T5HO bulb is much, much brighter, and has to give much more light at a given distance than the T12 bulb.

“Watts per gallon” is dead!



PAR

Light intensity can be measured in lux, which is the intensity as perceived by human eyes. Or, it can be measured in PAR units, which is the intensity as perceived by plants. PAR is an acronym for “photosynthetically active radiation” - the radiation (light) that is used by plants for photosynthesis. The units of PAR are micromols of photons per square meter per second. So, a PAR of 1 is one millionth of a mole of photons striking a one square meter area every second.

Human eyes see the yellow green area of the spectrum of light very well - our eyes are very sensitive to yellows and greens, but we see reds and blues much less well. Plants are very sensitive to reds and blues, absorbing most of the light in those colors, but less sensitive to yellows and greens, reflecting a lot of the light in those colors. That is why most plants look green or yellow.



MEASURING PAR

The best way to find out how much light intensity we have in our planted tanks is to measure it. To do that we must use a PAR meter. A few years ago the only PAR meters available cost a few thousand dollars apiece. Now there are much cheaper PAR meters available.

You can buy a Quantum PAR meter, Model MQ, for $329 plus shipping. That is a near laboratory quality meter, with a guaranteed calibration, which can be re-calibrated at the factory when needed. It is the Cadillac of hobbyist PAR meters, usually bought only by clubs, where many members can use it.

A lower priced version of the Quantum PAR meter is just the sensor, Model SQ, for $139 plus shipping. To use this you need to either use a good millivoltmeter, which gives the best accuracy, or a cheap lux meter, like the Mastech LX1010BS, from Amazon.com, at about $20 plus shipping. Used with the lux meter you need to do your own calibration. You can use your digital multimeter, with a millivolt scale and the sensor, to determine the PAR from a light at a fixed distance, then connect the Mastech lux meter to the sensor to see what the meter reads at that PAR. This gives you a calibration constant for that combination of sensor and meter to convert lux to PAR.

Still cheaper is to buy one of the DIY PAR meters made by Mistergreen and/or O2surplus, for about $60. These are calibrated, and the meter reads in PAR units, but they may not be available when you want to obtain one.

Cheapest is to buy a Mastech LX1010BS, at $20 plus shipping, and modify it yourself per http://www.plantedtank.net/forums/showthread.php?t=179789 but you have to calibrate this yourself. However, the total cost should be $35 or less. If the Quantum PAR meter is the Cadillac of PAR meters, this is the refurbished Volkswagon bug of PAR meters.

SELECTING A LIGHT

Before we can even start to measure the light intensity, or PAR, that a given light will provide on our tank we first have to obtain the light. It may seem that we have to be working blind when we make this selection, given that knowing the “watts per gallon” won’t tell us anything about the intensity we will get. But, because there are now many PAR meters in hobbyist’s hands, we now have a lot of data on how much PAR we can get from several different lights, made by several different manufacturers. More data becomes available every month.

Today we can chose one of several different types of lights:
T5HO fluorescent lights with 1,2,3,4, etc. bulbs
T5NO fluorescent lights
T8 fluorescent lights with 1,2,3,4, etc. bulbs
PC power compact fluorescent lights with 1 or 2 bulbs
LED lights of many configurations - DIY or ready made
CFL screw-in fluorescent lights - DIY

For each of those types of lights a chart can be made showing the PAR produced by the light versus the distance from the light. These charts show the light intensity as measured without a tank of water being involved - just the intensity as measured in air. This is necessary to avoid the many variations in intensity caused by the tank dimensions and the cleanliness of the tank glass, both of which can have about a 10-20% effect on the intensity.





LOW LIGHT, MEDIUM Light, HIGH LIGHT

I don't believe there is any consensus about the definition of low, medium and high light. But, here is my definiition, subject to, and almost certain to change:
Low light - 15-30 micromols of PAR - CO2 is not needed, but is helpful to the plants
Medium light - 35-50 micromols of PAR - CO2 may be needed to avoid too many nuisance algae problems
High light - more than 50 micromols of PAR - pressurized CO2 is essential to avoid major algae problems

The following charts show the data that I now have for various lights. As I get more data I will keep updating the charts and adding new ones. If you want a light that isn’t included in the charts you can study the reflectors used in the light you want and compare them to the photos following the charts to see which charted light is closest to the one you want, to get a rough guess at what PAR that light will give you.

Fluorescent tube lights produce about the same light intensity for any length of tube, from about 24 inches to at least 60 inches. The longer bulbs are proportionally higher in wattage, so that the bulb wattage is mostly a measure of the bulb length, not the bulb brightness. For bulbs shorter than 24 inches, this may not be true.

CAUTION: Not all lights use a true, full power HO ballast. Some cheaper models use lower power ballasts, and will not produce as much PAR as those with good ballasts. Compare the chart for the FishNeedIt lights to the others for an example.






One layer of window screen over the bottom half of each bulb, right on the bulb, drops the PAR by about 30%






See http://www.plantedtank.net/forums/showthread.php?t=160396 for much more information on LED lights.

Photos of various reflectors used in T5 lights:




Hagen GLO 2 bulb light


ATI 4 bulb T5HO


Home Depot 2 bulb Diamond Plate Shop Light - note the reflections of each bulb.



One bulb T8 light with fairly good reflector


Aquaticlife 4 bulb Light
 
See less See more
15
#3 ·
I agree, BUT it's the NOBEL PRIZE, hehe. ( the prize was named after Alfred Nobel: The dude that invented TNT )
 
#4 ·
Hi Hoppy. I'm new here. Have you measured the PAR using 30" tanks like the 20Long, 29g and 37g? Most bulbs, PC, T5HO, T8 and T12, are approximately 22" long. This configuration has approximately 4" of space on each side that is not directly under the bulb. (I know that Coralife has some 30" T5NO and T5HO fixtures with 28" bulbs available and Zoo-Med has some 30" T5HO fixtures with 28" bulbs as well. We can use your presently available charts to find the PAR for these fixtures.) I am interested in the 22" bulbs over 30" tanks because so many people have this size of aquarium. You could probably extrapolate this reading of PAR.

One more question concerns your PAR measurements in the graph showing the distance from front to back. Why are there 10 micromoles difference at 4" from center in the aquarium with no water and the tank with water has 5 micromoles difference. Why is there a difference when measured from front to back? The back side of the aquarium has more PAR than the front side, if I am reading the chart correctly. Why is this? (I am not concerned with the 5 micromoles difference from an aquarium with water in it and one without water.) Why does the PAR increase on the back side verses the front side? It would seem to me that the front side and back side would have similar readings as the distance from center increases.

One last question. Why do the aquariums with water in them have a higher PAR than aquariums without water? I would guess that it would be the other way around. Aquariums without water would have a higher PAR reading that aquariums with 5" of water in them would be my guess.

Thanks,
clang
 
#5 ·
When light shines into an aquarium the air-water interface slightly focuses the light. Beams off center are refracted towards the center. This is why the PAR around the center of the tank goes up when you add water. Near the glass at the front, back, and ends, there is considerable reflection of light back into the tank from the glass-air interface, which increases the PAR near the glass. But, if the glass is less than perfectly clean on either side of the glass the amount reflected back into the water is decreased. As I recall I did those PAR measurements in a 10 gallon tank, which I didn't take pains to clean thoroughly. The glass must have been non-uniformly dirty. This was a few years ago, so my memory of that isn't perfect.

I don't have either a 30 inch long tank, nor a 24 inch long light to do PAR tests with. Most of the data I have comes from others who either post it or PM it to me. Someone with the right tank and light, and access to a PAR meter, needs to do that testing. One thing I do know is that if you raise the light far enough above the top of the tank a short light can light up a tank well enough for our purposes. In fact one good way to light a tank is to use a short, but very bright light (4 bulb T5HO, for example) suspended a foot or more above the tank. That greatly reduces the difference in PAR between the water surface and substrate level, and gives relatively uniform light over the substrate.
 
#6 ·
When light shines into an aquarium the air-water interface slightly focuses the light. Beams off center are refracted towards the center. This is why the PAR around the center of the tank goes up when you add water. Near the glass at the front, back, and ends, there is considerable reflection of light back into the tank from the glass-air interface, which increases the PAR near the glass. But, if the glass is less than perfectly clean on either side of the glass the amount reflected back into the water is decreased. As I recall I did those PAR measurements in a 10 gallon tank, which I didn't take pains to clean thoroughly. The glass must have been non-uniformly dirty. This was a few years ago, so my memory of that isn't perfect.
Thank you very much for the explanation. I understand what you are saying. The air-water interface focuses light somewhat like a magnifying glass does (maybe an odd shaped one) and light reflects off of the sides, front and back somewhat like a mirror. The disparity between the front to back measurements are understandable.

I don't have either a 30 inch long tank, nor a 24 inch long light to do PAR tests with. Most of the data I have comes from others who either post it or PM it to me. Someone with the right tank and light, and access to a PAR meter, needs to do that testing. One thing I do know is that if you raise the light far enough above the top of the tank a short light can light up a tank well enough for our purposes. In fact one good way to light a tank is to use a short, but very bright light (4 bulb T5HO, for example) suspended a foot or more above the tank. That greatly reduces the difference in PAR between the water surface and substrate level, and gives relatively uniform light over the substrate.
I don't have a 30" tank running either, but I read a lot of posts from people that have them. 30" is somewhat of an odd length when looking at lighting options. It seems that lighting is basically designed in one foot or 12" increments, like 24", 36", 48", 60" and 72". This makes 20" and 30" length tanks having fewer lighting choices. 60" fixtures and bulbs for them may be somewhat of an oddball length too.
 
#8 ·
Definitly sticky. Thanks for the time you put in on this Hoppy. It is a great help.
 
#10 ·
PUR = photosynthetically usable radiation, a concept I haven't yet seen a good reason to dig into. Plants use light from the whole spectrum between about 400 and 700 nanometers wave length. Unlike what many people believe, they also use green light, just not as efficiently as they use blue and red light.

If it was hard to get enough light to grow our plants, we would want to get every advantage we could, and look for bulbs that wouldn't waste any light in the green wave lengths, which, if you think about it, would leave us with very drab looking plants. But, the bigger problem we now have is avoiding too much light, not getting enough light. So, why bother with PUR?
 
#13 ·
The main argument that Daniel suggested in a debate I've had over the years was that with higher PUR, we get more growth per watt.

The other issue really is what coloration and aesthetics a particularly color spectral curve imparts to the plants. This is far harder to quantify(thus matters little in terms of growth itself), but PUR and RGR (Relative growth rates) can be quantified. This is particularly true when you get down closer and closer to where P=R, or the light compensation point for lower energy input for PUR.
 
#14 ·
It is also noted by some people who have measured PAR for their light, that bulbs with different color temperatures, 6500K and 10000K for example, produce different PAR. For now I think it is hard enough to characterize what a given light fixture produces in PAR vs distance, that the finer points, like color temperature, will just have to wait a few years. The difference in PAR from a 10000K and a 6500K bulb is probably within the uncertainty range for most lights. And, all of this data is from lights operated out in the air, not in a water filled tank, where the difference is also probably within the uncertainty range for most lights.

Tonight I was measuring PAR from a screw-in CFL bulb in a dome type reflector. I noticed that on turning on the light the PAR slowly increased for a minute or two, then even more slowly dropped part of the way back. It was an old bulb, so this may be a problem with the bulb's built in ballast. Imagine the error this could introduce if you didn't allow for it.
 
#24 ·
This is based on my own measurements, using brand new GE CFL 6500K bulbs, in 8.5 inch diameter reflectors. Larger reflectors will give lower PAR and smaller will give higher PAR.
Based on the threads I've read here the 8.5 dome lamp seems to be much more popular then samller one.

Is that because the 8.5 inch dome will give you a wider light spread then a 5 inch at the same distance?
 
#27 ·
Yes, when the same amount of light is spread out over a larger area it has to be at a lower intensity, and vice versa.

Excellent post and thread. My only suggestion would be possibly lowering the scaling on the CFL chart. It goes very high, and most people using CFLs are intetested in a PAR range of 10-100 or so ( from what I can gather).
Good suggestion! I changed it.

Great information! What is missing is a graph on LEDs. Thanks for all the time you spent gathering this info and then sharing it with us. I didn't realize that a less than $100 par meter option existed
LED lights can't be treated this way. The variety of configurations is almost endless, so no simple charts can ever show much of value with them. However, it would be possible to put the most popular, most usable on planted tank ones on one chart. Keeping that chart up to date as more and more different LED lights become available would be a major chore. Maybe someone wants to convert the information in http://www.plantedtank.net/forums/showthread.php?t=160396 to this format? It takes a lot of time.
 
#31 ·
OK, I just spent most of the day making a chart as you requested. It is best used just to help filter out the LED lights that won't work for you. Once you do that you can do more research on those that look like they might work. See the first post for the chart.
 
#30 ·
PAR values are subject to change and based on the posters (Hoppy) experience... in this case Hoppy has a lot.

Another thing to consider is the PAR values are at the substrate, so if you have 50 PAR at the substrate you could easily have 150 at the water surface.

Lastly, it has been said by many people that if you have 50PAR at the substrate there isn't a plant you can't grow with proper CO2 and ferts.
 
#36 ·
Hi Hobby! Just a quick question and hope you can help me clarify on this. According to the chart for T5HO, At 21'' above the substrate the Catalina 4 bulbs give 80 micromols ; the Catalina 3 bulbs gives approx 62 micromols. Does this mean turn all bulbs on will give that certain PAR value or we multiply the PAR by number of bulbs ? Thank you
 
#37 ·
The PAR numbers in those charts are for the actual light fixtures, with a full complement of bulbs. The exceptions are the ATI 4 bulb fixture, which had only 3 bulbs in it, and the calculated PAR shown for one bulb for Tek and Catalina lights. That last one for the Tek lights is usable by multiplying the number of bulbs by the PAR per bulb. But, Catalina uses different shaped reflectors in various light fixtures, so the number is only good for those with similar reflectors (the one bulb Catalina lights appear to use the most different reflector shape.)
 
#39 ·
You said in your data for t8's that you were recording par readings on t8's with basic white reflectors. Aqua one has two 36" fixtures "flouro grow" and "al30d" (specifically speaking of the one I want). With "polished aluminum" reflectors which looks like it is one reflector spread out for two tubes. What would you recon The par would be with this in consideration as well as the new data collected from Thr diamond plated t8's.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top