Let's talk liquid cooled lights.

[Uncle FREEDOM]

I want to open up this thread for discussion as to using liquid cooled light/tubes.

Let me fire off the first question...


i am looking to try some out, and here is what i want to do. I want to have a small control rez (to fill and unfill the water, plus add anti algae, and monitor temps) But what i am interested in is if anyone could work out how much copper (or other kind of line) i would have to run and bury 4 feet in the ground to cool say per 1000 watter. Since i have the land and the tools i would rather let mother nature do the work for me...

is there any idea's as to what i would need?

thanks

 

[Guest]

Lets not, and say we did. Whats the point?

Edit.. ya looked at every water cooled thread, and came to the same conclusion. I really don't think it is a good idea to be playing with fire like that... especially when you come here of all places for engineering advice..

 

[1love1earth]

OK, here's the deal, exactly how many watts are you running/going to run?

Water cooling involves alot of equipment and really only makes sense for a major commercial op, and even then...

You would need to make sure 100% that it is leak safe, and even then, leaks can happen.

The premise behind water cooling is that water cools far better than air. However, you have to cool the water, either with a radiator and fans, or bury something in the ground like you said, which is costly, and kind of sketchy looking.

Another big concern, even if you do use the "under ground" method to cool your water, now you're talking sub ambient temps, and you potentially are going to have condensation, which is deadly. The bulb housing is going to be the warmest area, and here you are introducing sub ambient water, could be deadly.

Watercooling has been done with computers for some time now, and without issues at that.

However computers are not illegal, and don't produce nearly as much heat.

Also, don't forget that we've got a pump which is going to have to be sized for your loop, and I'm going to guess we're gonna have to measure in HP and not GPH.

Point blank, I'd stick with air.

 

[Stealthy]

Water, Hot bulbs and electricity just don't mix.

 

[Uncle FREEDOM]

well first off yes i DO want to talk about it, so please no more let's not comments.... sheesh

second water that has all it's impurities removed is a very POOR conductor of electricity, and steps will be made to make sure things are tight.

Now the reason for the ground cooling is so i do not need a big chiller to keep the temps down... The chiller will be set to keep the water temps within the dew point range. I have a secluded place and quite frankly would have no problem tearing up my lawn to install said ground cooling system.

As to the size of the set up, that is secret for now, but it will be bigger then average hehe... that is why i am looking for ranges in the 1000 watt range and extrapolate from there.

The purpose of using ground cooling and water cooled lights is to minimize my heat signature (even when using air cooled shades and a/c you still have to vent the hot air some where).

It seems to me with some work and prior planning i could have a mostly silent, cool running, grow.

plus i get to experiment at the cutting edge which i enjoy.

Now with that out of the way, i am looking for people that like to put their heads together and figure out some basic designs/physics and i will be the lab rat! lol

If i fail, fine. i will share my fail points. If i can make it work i will show the world, and we as a community grow with knowledge.

So if there is any positive comments/ suggestions, or even careful words of caution, as to what points to look out for, lets discuss. but...

please no more it can not work stuff.... i will be the judge of that positive or negative, to the betterment of the community.

p.s. Just to start us off... ground temperature at 4 feet below the surface is a constant 50-55 deg f.

 

[1love1earth]

well, one big thing is going to be condensation, whether on the tubing/piping, or the actual bulb housing itself.

Regardless, are you talking about fabbing your own stuff, or buying a premade unit?

I mean, digging the hole is quite a project in and of itself.

Not saying it can't be done, just saying I don't see it as the absolute best option out there.

If you're just doing it because water cooling your lights is slick as hell, and not for easy of use bang for the buck, I'm all for it, just as long as that's what it's about.

Conductivity really doesn't matter.

If you have a leak in the actual "bulb cooler", it might eventually just fill with water (as it'll be a pretty good container) and it'll probably at least short out and burst the bulb itself.

Maybe that's a stretch, but a leak will at least end up all over your floor, as the ENTIRE res slowly empties into your growspace. Even if you had a cutoff for your lights, you still might lose the crop, and the growspace would at least be pretty flooded.

Just things to think about, and plan against.

How many lights are you talking about?

You need to size your loop.

I would do research on "underground PC cooling", and look at how this has been done before, you're doing the exact same thing, except cooling a light instead, and maybe a lot more units?

 

[smoketrichs]

Well if you're only going to cool 1 1000w HPS, why not just cooltube that bitch and be done w it? from what i've read, water-cooling is very inefficient. Not only does that light have to travel through two or more layers (glass + water) to reach the plants, the light that does come through is mostly unusable. So yea, for the time and money just cooltube that bitch and put a dedicated 6" inline or blower on that badboy.

 

[gregor_mendel]

I support you, Uncle

I support you, Uncle

Though it may me true this is not the best place for this advice. As you can see, there are many here shooting down your idea without having any experience. I have thought of doing exactly what you propose, but I don't have the space and privacy you do.

I do think it would be good to get just one unit, and make it work before you start trying to water cool 10K watts (or whatever you will be running)

Six feet under, the ground is 55 degrees F in most places.

The manufacturer can give you a specification for length of tubing per 1000 watts when using just a fan. I would use the same amount underground. I would also backfill with sand most of the way, as it is easy to place your coils in, and to remove them.

I understand your desire for quiet. I am cooling a 2600 watt vertical setup with a fan, and it is loud as hell. An AC is just as bad. A water cooler is quieter than an AC or fans, and the ground is silent.

I know it won't dissipate heat as well, but PEX, or some other plastic pipe may make your job easier than coppr coils underground.

Before you look anywhere else, give the manufacturer a call or email. You aren't the first person to think of this, and they can probably tell you exactly how to to this, or steer you sway from it.

 

[ixnay007]

Your idea is interesting, but the heat exchange you can obtain from underground transference isn't all that good, you need to read up on the carnot cycle, and the law of diminishing returns Unless you're doing huge runs, the water probably won't lose all of it's heat, and the whole thing will probably start to get gradually hotter (which isn't good)...


I'm not a huge math guy, so I'd hate to make estimations about how much tubing you'll need, but I'd imagine it would be at least 15 or so feet..

There's lots of info on these pages.

http://en.wikipedia.org/wiki/Thermodynamic_cycle

http://en.wikipedia.org/wiki/Carnot_cycle

http://en.wikipedia.org/wiki/Vapor_compression_refrigeration

 

[Uncle FREEDOM]

leaks will be contained in the following manner...

moisture sensors on the floor of the room(s)

if sensor tripped then flow will be interrupted.

once flow is interrupted then lights will be hooked up to automatically shut down.

once the above happens i will be sent a e-mail.

so that part is taken care of.


condensation....

the lines will be wrapped in insulation just like you would in a house. the temp or the control rez will be set to a point with a chiller/heater as to avoid reaching the dew point.

The land used for cooling the tubbing...

i have over 3 acres of land that i could use, but i figure 1/2 a football sized plot running back and forth should be more then enough...I have read that once you get to below 4 feet the temps remain constant 50- 55 degrees (my well water tap temperature confirms this) Also i am lucky as i live in a sandy soil area. (i could sell my water as bottled)

humm i am hesitant to say how big i am planning ... but in the interest of science lets go big.... lets plan for 10,000 watts on a flip/flop cycle (so a constant 10,000.

b.t.w. it is for a friend not me (wink)

Also i live in canada, where i have all 4 seasons. I doubt that if i spread it out over a large area, and the shear size of the the ground involved that it will not change the soil temps much. I could be wrong, but this is why i need your help, as i do not propose to be a expert on everything.


Thanks again for all the replies... let's constructively put or heads together and i will see if i can return the favor by showing you guys the results of my testing.

thanks

p.s. if this helps you guys out ... to cool 10000 watts of lights unshaded you would need 4000btu's per light (4000 x 10 = 40000 btu) how many tons is that??? So now we need to figure out the ground loop size that they use for cooling houses. Once that is figured out we need to extrapolate that info to cool 40000 btu's worth of heat.

let the googling begin! LOL

 

[green_tea]

this is my idea:

you have 2 loops.

one loop is for your lights (possibly more than one depending on how many lights etc...)

this loop will circulate cool water from a res around the lights and back to the res.

Second and important loop:

this is the loop that you use to cool the res. this is where you dig the hole, and put coiled copper tubing in the ground. length to be determined based on how many W of lights you have going.


the reason you keep them separate is so that you can monitor the res tank, and make adjustments accordingly.

If the res water stays too warm, you can add a water chiller to the res.
if the res water gets too cool, as in goes below the ambient temp of the room (probably around 80 if you are running CO2) you have a heater set to turn on and make sure the water ALWAYS stays say 5 degrees above ambient temp. (ambient temp of where the lights are)

THis can all be done electronically if you know what you are doing. id suggest multiple temp probes, in case one breaks. Id also suggest go by the average between the too, but then when one breaks and goes outsids say 10 degrees of the other probe, you stop using it. (this is going to far, and just getting into all the programming specifics, so Ill stop here)


designing it is another issue, make sure the water is distilled and has 0PPM, and also add water additives to basically make the water 100% non conductive.

you could probably modify a cool tube to run water if you want to, though that will mean you should have a REALLY huge pump as those things are huge. (you really want to make sure that there isn't too much water surrounding the bulb, maybe an inch or two at the most. This water will also have to be going around the loop extremely fast so that you are creating as much turbulence as possible.

ideally you'd have bumps and shit inside the tube to create more turbulence, but that is going to screw up the light output.

you should still create a T; this is basically to help you get out all the air bubbles in the loop. (basically just a T break in the water loop, with the 3rd fitting going to a pipe that is made to be the highest point in the loop; like an upside down T)

 

[Outlaw Grower]

leaks will be contained in the following manner...

moisture sensors on the floor of the room(s)

if sensor tripped then flow will be interrupted.

once flow is interrupted then lights will be hooked up to automatically shut down.

once the above happens i will be sent a e-mail.

so that part is taken care of.

dude if your water cooled light leaks, before the moisture hits the floor to set off the sensors your house may already be on fire. so what if it emails you, whats that going to do if your house is on fire? does it also put the fire out and hide the grow?

edit, i hate saying shit like the above but these water cooled lights are asking for trouble bro... thats like water cooled electrical sockets..

 

[Uncle FREEDOM]

dude .... who said the grow was in a house, or for that matter even above ground.???

want me to install a fire suppression system??? fine....

next problem

thanks for the concern though... sure i might run into problems, but if no one is willing to try and iron out the kinks in the new technology then we never advance... i have the time and money to do both.

p.s. lets all relax... take a deep breath and repeat after me.....

PURE WATER IS A POOR CONDUCTOR OF ELECTRICITY....

 

[pHaroaH]

My biggest problem is in the sacrifice of lighting when it passes through two layers of glass and the water. I think green_tea is on to something. I propose this... why not make the underground water loop and have it pump through a radiator that you pass the light's air through. A chill box. Just like a car uses a radiator to cool a car engine. This way you could control the pump speed based on temperature. This would also allow you to use the light as a heat source when it is cold. This way you could heat and cool your grow room air, and not even have to exhaust it if you wanted to use CO2.

P.S. Do some searching on Hydronics and Radiant Heating. You will find rules of thumb about BTUs exchanged, recommended length of runs, expansion tanks, material types, and etc. For example, to get the best exchange of heat you want the water to have turbulence, not laminar flow. Tubing size, tubing run lengths and pump capacity need to be properly calculated to ensure this.

 

[1love1earth]

don't forget about CONDENSATION!

I realize you aren't stupid, but please please please think about this one.

Condensation occurs (basically, i'm no physics expert) when the temp of any given object (in this case all of your piping/tubing that is indoors, and maybe even the light coolers) is below the ambient air temperature.

Now lets say the ground temp is 50F, thus your water stays around 50F, before it hits the first light.

Now lets say it's a hot day, and you've got an ambient air temp of 70F, or even 80F, all of your tubing is going to start dripping little water droplets.

So like it was said above, you'll need a heater running, which sort of defeats the purpose, to some degree.

Also, you could just use a radiator, but then you've got the dreaded "hot air" that you were worried about being picked up by thermal imaging.

Though, lets be frank, if someone is hitting your house with thermal imaging, they're probably already onto you.

I mean, again, if you're doing this because water cooling your lights is cool as hell, that's good.

If you're doing it because this is a great functional idea, please please please please reconsider other alternatives.

Hell, you could aircool the lights, and pipe the hot exhaust air through a chimney.

Plus, unless someone is sitting there 24/7 with thermal imaging (again, if they're bothering to do that, they're already onto you anyways) you could just vent the hot air out, and make it look like the vent for a clothes dryer.

Plus, do not underestimate the size of the pump you will need, it will be fairly large.

Not only do you have the issue of how many meters of tubing/piping, but also "head", or feet up and/or down the loop travels over its course. The difference between the lowest and highest points in the loop represent "head", and effects how powerful of a pump you need.

Again, I'm going to guess you'll need one beefy ass pump.

And unless you want to go returning pumps, and digging up your yard twice, you'll need to get all this right the first time.

IE, how big an underground resevoir, and how many feet of tubing do you need to properly exchange the heat? How many GPH (or likely horsepower) do you need to not only actually pump water, but pump it sufficiently for proper heat exchange. How large does your heater need to be?

You're talking a good number of 'basic' physics equations, and a solid bit of math, and it all has to be more or less spot on, you can do too much, but too little and the whole shebang ain't gonna work right.

Nevermind the whole leak problem. You could easily spring a leak every month or two, even IF a moisture sensor picks it up, and your pump cuts off, it'll still leak (though you'll be notified by email, and able to fix it quick enough) somewhat, but more importantly, your lights will cut off, and potentially the entire crop is compromised.

The reason I'm beating this to death, is that this seems like a more or less commercial op, or at least an op that requires a decent bit of extra capital compared to normal, which I assume you're trying to recover through the grow op.

Again, look into the underground cooling that has been doing with watercooled computers, it's all DIY.

BTW, just curious, you aren't planning on digging this hole by hand are you? When I saw it done with PCs they used a claw digger, not that it's a big deal, easy enough to rent one, but it does call some attention (potentially).

The sub ambient temps are something you REALLY have to watch. You will need a heater in the loop, and it will have to be sized appropriately.

You're talking quite a bit of electronic hardware as well, moisture sensors, variable water heater, pump cutoff, light cutoff, email notification, if you pull this off, it's gonna be one helluva thread.

 

[DIGITALHIPPY]

id like to see one of these thigs in action...hf makes one...
i bet itll work for small growers 1-3k...
distilled water and there will be no minerals to conduct.
you could use a old car radiator and just put one of theos 14-20$ window fans on it(you know the square "BOX FANS")

 

[sinickal]

www.ebay.com

search for "water cooled lights" without the quotation marks. it should bring up a couple of housings for this.

I have quite a bit of experience watercooling computer components and game consoles, maybe i could help explain it to you. make a list of everything and i'll gather some of my knowledge and explain in deatil for you piece by piece. I'm just too stoned to look back and see what all you said on the first page.



2nd edit

But i found this one to be better for the application you are thinking of i believe this . You'd want that being connected to a radiator to dissipate the heat, with fans blowing through the radiator to cool the radiator/water. I am a firm believer you need to have a large reservoir for your water cooling system, since it makes the flow, keeps water back, faster like a dam. And the reservoir cuts some of the heat down in the water also, it allows it to sit for a little bit before being put back into the water flow. Then obviously you'd water quite a good pump, unless you could mount all of this equipment (reservoir, radiator, and the pump) above the lighting fixtures height: this in turn would make the pump not have to work as hard. I'd still look for a good pump anyway, and try to get the stuff above the light.

If you have anymore questions post em up.

 

[sinickal]

I keep so i'm thinking deep... lol

if you wanted to get hardcore, you could put the reservoir in a mini fridge, and have the water lines running through the side of the fridge. people have tried to fridge cool computers, it sorta works, but i'm 99.9% sure this wouldn't comprimise anything.

 

[1love1earth]

even a normal sized fridge probably couldn't even handle a single 1000 watt light, they just aren't built to fight heat loads like that, way too many watts to reject.

Again, sub ambient water means condensation.

I fought this when water cooling my PC with a recirculating lab chiller, and a beefy one at that.

Again, your pure water may not be all "that" conductive, but your condensation will be.

Also, using a radiator eliminates the whole "thermal signature" advantage.

 

[sinickal]

There's always going to be advantages and disadvantages to doing something in this setup. if you were to not have a radiator, then the water would just continue to be at the same higher operating temps. I'm not sure at what you mean by "even a normal sized fridge probably couldn't even handle a single 1000 watt light, they just aren't built to fight heat loads like that, way too many watts to reject", but the whole point of the fridge idea, was to cool the excess water (the reservoir), not the the light bulb it's self.