Idea for geothermal air conditioning, cheap

[Squankum]

Have a buddy who has moved into a home in the country. He has a healthy creek running on his property, close to his house. This is the Southeast, hot and humid is what we've got. His attached garage has no A/C.

My idea: run tubing upstream in the creek, have cold water flow into it (hopefully no pumping needed, depending on elevation change), have the tube enter the garage, flow through a large aluminum SUV/truck radiator ($40ish at Pull A Part), have a box fan blowing air across it (ducting, of course), and another tube taking the water back to the creek. If one radiator isn't powerful enough, you can always get more junkyard radiators and put them in series?

This is not a swamp cooler, there is no heat of evaporation being used. That reduces the power, of course, but then again, this system introduces no humidity into the building, either.

Dig this: loaned my buddy my infrared temp gun and he went home after work yesterday, and after a long, sunny day, temps in the mid 90's? This creek coming out of the hills, by his house, it is 43F. It's fed by springs and shaded by trees all the way. Looks promising to me.

I've pondered other issues, like a sediment trapping barrel and valves and pressure gauges and maybe running the fan directly off the water flow so zero electricity is needed.

Heck, I don't see how this couldn't do the whole house, if this idea is worth a darn. A puny output running 24/7 could add up. A drip pan and drain tube under the radiator (now promoted to the fancy title of "heat exhchanger") could even take some humidity out of the air.

Anybody have any comments/suggestions on the basic concept? My googling hasn't found much about this, but I haven't done my usual full google frenzy.

 

[vartz04]

youre going to need pumps. The restriction of the radiator will kill the flow of the water. And how are you going to get the water back to the creek.

Pumps might be way cheaper than you think though. Could even solar power them off of some HF panels.

 

[HoosierBuddy]

From an engineering standpoint, I'd say it's worth a shot.

From a practical standpoint, it may be marginal.

From a legal perspective, you may run afoul of various environmental rules. Unauthorized discharge into a stream or waterway...etc, etc.

Good Luck!

Phil

 

[sneezer41]

While it would take a large heat exchanger, I imagine it would work to some extent

You would probably need a pump.


And a wetlands permit......


You might be surprised how much water flow you would need.

 

[venturesomerite]

DEP will probably shut that down if they got wind.

As far as doing it, The house would have to be significantly lower than the creek to get it to flow through one or more radiators without the aid of the pump.

I think practically speaking, the hardest part would be keeping the inlet and tubes from clogging with debris.

 

[Squankum]

The most tempting thing about this is the 43F water and the low cost setup.

Thanks for the heads up re. restriction in radiator. This site is, I strongly suspect, also worth some kind of minihydro. Running this tube just gets him halfway towards realizing that.

Returning water to the creek? Near the house, the creek is lower than the heat exchanger and every other thing in the house, like most houses near creeks.

Don't think we'll need thundering flow to this heat exchanger. The real challenge is extracting the cold from it with ... air.

As for discharge to the creek, hey, we're just borrowing that water, nothing added! (Once that radiator's cleaned out in the first three minutes.) I realize the laws and regs may not see it that way...

 

[daveroy]

cool concept... go for it (don't forget to post pics!)

 

[Squankum]

Sediment: it's a rocky, mountain kind of creek, so that might help a little.

Swirl chamber?

Made in India, looks worth ripping off with a 55 gal drum, heh heh.

http://www.jains.com/irrigation/filtration equipments/jain hydrocyclone filter.htm


.

 

[Cougar]

What about a closed system.
A copper coil in the creek to cool the water.
Pump water from the coil up to the radiator, then back through the coil to be cooled.
Wouldn't be pumping water from the creek.

 

[240sxguy]

^ I was thinking the same thing.

 

[Jagmandave]

Or running the creek water into a barrel big enough to get some head pressure, then tap for you radiator far enough off the bottom of the barrel to still get pressure without the sediment. No pump needed.....

 

[pseudorealityx]

Does the creek every dry up when we're in drought?

If the water truly is at 43 degrees, you'll need a condensate pan, and a way to drain it.
I also vote for a closed system, otherwise sediments are going to foul that radiator in a hurry. That said, you're likely to foul up the water side heat exchanger if it's sitting in a creek too.

A box fan doesn't give you much static to get through the radiator and then out to wherever you're trying to go. If it's just a small garage, it might be ok. Otherwise, I'd look at a fan that's got some more oomph.

 

[Squankum]

Creek has not been dry once yet, and there's been some dry here this summer. It's always running strong, he says.

I was thinking $15 box fan because... $15. I was prepared for more fan being needed. Gotta try it out first and see what it delivers. Don't want to spend too much electricity on any of this, right? The goal is to get away cheap, investment and operating costs, compared to pumping freon.

 

[rlitman]

With the closed system, you can get around the fouling of the wet side heat exchanger, by using plastic. Just use black poly sprinkler tube. If it isn't exchanging well enough, just use more. Make a 50' loop that goes around and around. Heck, just buy a 100' coil of it, spread it out a little bit under water, and connect the two ends to your feed lines.

But, I'd verify the actual water temperature first. It's not that 43F is unreasonable (depending on the source of the water), just that an infrared thermometer is fooled by many things. Dip a real thermometer in there.

Don't try to use creek water in a copper radiator. They make monel radiators for this purpose, but they're mighty expensive. Stick to the closed loop. Oh, and consider filling it with a glycol mix. It will be a little less efficient, but you won't have to drain it in winter.

 

[CrashTestDummy]

What about a closed system.
A copper coil in the creek to cool the water.
Pump water from the coil up to the radiator, then back through the coil to be cooled.
Wouldn't be pumping water from the creek.

^^This. Or some black PVC tubing like they use in the real geothermal systems. Bury the tubing between the shop and creek, and you'll want to run some antifreeze in the tubing to keep it from freezing in the winter. Yep, it'll take some pumps, but should be cheaper than an A/C condensing unit.

Gene Beaird,
Pearland, Texas

 

[Dick in Wisconsin]

Worth trying. Let us know how it turns out.

Another thing you could try is finding a working window A/C unit and try to figure out a way to put the coil on the back of (the one that gets hot) into a pan of water. Then run 43 degree water from the creek through the pan of water. It will use electricity ... but the A/C portion will be really efficient!

 

[deter]

just to throw a couple things out there:

A 1/3hp sump pump uses around 800watts, a box fan uses around 150watts.... a window air conditioner has a surge usage at startup, but typically run about 900 to 1000 watts

ymmv

 

[HoosierBuddy]

Worth trying. Let us know how it turns out.

Another thing you could try is finding a working window A/C unit and try to figure out a way to put the coil on the back of (the one that gets hot) into a pan of water. Then run 43 degree water from the creek through the pan of water. It will use electricity ... but the A/C portion will be really efficient!

I like that. I know of a couple of people that put their geo-thermal loops in the bottom of a lake. I guess it's the berries.

Phil

 

[Squankum]

just to throw a couple things out there:

A 1/3hp sump pump uses around 800watts, a box fan uses around 150watts.... a window air conditioner has a surge usage at startup, but typically run about 900 to 1000 watts

ymmv

There's flow here; any pumping that's needed can be done with simple direct mechanical hydropower. Why pay money/juice?

Heck, a simple paddlewheeel and a junkyard water pump...

And that's only if there isn't enough head from the creek itself. I've not been to the site yet, but from what I hear, there probably is enough elevation change to feed this thing, no pumping at all. (Plus, the longer upstream that pipe runs, the more this turns into a dual use tube -- geothermal A/C now, minihydro electrical generation later.)

The goal here is cheap. My buddy's into thrift and self-sufficiency, and he's handy. Spending money to reduce carbon footprint is not his bag. And I really think this can be done pretty cheap. Junkyard radiator is $40ish. He already has some tubing. With some head in the creek, we're managing pre-existing pressure.

 

[Squankum]

These guys are up my alley: dirt cheap DIY stuff. I like alternative energy, but I especially like it when it pays itself off sooner rather than later. Here they are turning a Volvo McPherson strut/hub into a windmill:

http://www.otherpower.com/bartmil.html

They used to have a nice little writeup about a guy who turned an automobile blower fan into a waterwheel in a creek behind his house, using a large log as the dam/sluice to it. That page isn't around anymore, not in the original form, but while its output was meager, its cost was ridiculously low. Cheap financial input plus 24/7/365 electrical output, I'm not one to quibble so much about the amount of ouput.

 

[pseudorealityx]

There's flow here; any pumping that's needed can be done with simple direct mechanical hydropower. Why pay money/juice?

Heck, a simple paddlewheeel and a junkyard water pump...

And that's only if there isn't enough head from the creek itself. I've not been to the site yet, but from what I hear, there probably is enough elevation change to feed this thing, no pumping at all. (Plus, the longer upstream that pipe runs, the more this turns into a dual use tube -- geothermal A/C now, minihydro electrical generation later.)

The goal here is cheap. My buddy's into thrift and self-sufficiency, and he's handy. Spending money to reduce carbon footprint is not his bag. And I really think this can be done pretty cheap. Junkyard radiator is $40ish. He already has some tubing. With some head in the creek, we're managing pre-existing pressure.

The creek does not have any head above it's own elevation. You WILL need a pump, even if the coil is at the same elevation as the creek.

 

[vartz04]

you will more than likely need 2 pumps, even if they are just very small pumps. one to pump into the area of the garage, and one to pump it through the radiator.

Even if the creek is 20 feet higher than the garage you would only have 8psi. On top of that youll have to add the resistance of the length of the pipe which would probably drop it down to 6psi or so. I highly doubt if this creek is close to the garage you would have the 20 feet elevation change. On top of that if you run at such low pressure the water is going to move slowly, and then the temp rise is going to increase between the creek and the radiator. Even if its underground you are still talking about 10 degrees warmer than the creek water.

Now if you are talking about creating a funnel so the creek is "forcing" water through the tube you might get 10 psi after calcing the pressure losses.

A cars water pump puts out between 3-15 psi depending on engine speed. Just so you can get an idea of what its going to take at a min to get the water to go through the radiator properly

The way to do this without the electrical outlay is a small 12 volt pump. This runs at 110 watts. http://www.harborfreight.com/12-volt-marine-utility-water-pump-94639.html

then you would put 3 of these in out by the pump and hook them up to a couple of deep cycles http://www.harborfreight.com/45-watt-solar-panel-kit-68751.html you won't need the batteries if you are only going to use this during the day. the 3 of them together should power the pump. I would put one battery on it anyway to handle the cloudy days or what not where the panels cant keep up.

Off the grid, under $1000 outlayed for the whole project, and nearly free a/c.

Its a good concept I just don't see how you are going to do this for the couple hundred bucks you seem like you want to spend.

 

[Thruxton]

Good ideas here. If you want to see this idea exploited on a very large scale, google Cornell University Lake Source Cooling. One thing, I would be surprised if the water temp is 43, but I would count on the low 50's for sure. Small difference, not important to the concept.

 

[Sureshot]

I think I would try it with the paddle wheel pump drive concept before anything electrical. IMHO the warmup would be minimal between the creek and garage. You also have the siphon effect if you run the discharge farther out to a lower point as long as you are airtight. It may take a pump inline to get the flow going but once started it may be very easy to get sufficient flow.

As far as freezing I am not sure of your weather but if left running it likely will not freeze. I would use it as a heat source here if it is in the 40s or 50s during winter.

I love the idea of using thermal mass and natures own power. We have a 2 million gallon pond here that is 15' deep. I have thought about laying lines in it to cool the house then in winter pump the water on top the ice and try to freeze it solid for summer use. I would have a 2 million gallon ice cube to cool my house.

 

[tdkkart]

Start with a large, 8-12" pipe, open end in the strong water flow, reduce it down to 1-2" for the pipe to the shop. Don't remember what it's called, but you can multiply the creek' head by several time with the right size pipes.
If you're worried about debris etc, dump it into a tank then pump it from there with a filter on the pump inlet.

 

[jkeyser14]

just to throw a couple things out there:

A 1/3hp sump pump uses around 800watts, a box fan uses around 150watts.... a window air conditioner has a surge usage at startup, but typically run about 900 to 1000 watts

ymmv

1 hp is about 700 watts so a 1/3 HP pump will only be ~240 watts.

As far as everyone saying you will need multiple pumps, you will not. My geothermal system has 1000' of pipe (2 vertical loops) and a 1/3 hp groundfos pump has no trouble pushing the coolant through all of it.

As others have said, your buddy could get into large legal trouble. Your best bet would be to make it a closed loop system which also solves a lot of other headaches. And yes, you'd also need a condensate pump under the heat exchanger inside.

 

[lilredex]

The city of Tor. has a closed loop sytem in place to use that cold (drinking) water brought in from Lake Ontario, for cooling buildings.

http://www.toronto.ca/environment/initiatives/cooling.htm#system

 

[gatchel]

If using the coil in the water method, with Geothermal cooling you need about 500 feet of pipe in the water to get 1 Ton of cooling or 12,000Btu.

 

[5lima30]

A lot of older places around here have "springhouses or boxes" where they used to store dairy products prior to refrigerators. They are typically made out of native rock or even brick and were built right on the spring. Maybe mount a squirrel cage blower from a plenum (inside the spring box) and duct the cool air into the garage. Just thinking aloud....hmmm!

 

[where2]

Start googling: burkhardt turbines or Harris turbines. I think you'll like the results, especially if you run across the home power PDF that explains how to rebuild one of these little gems.

 

[bobscogin]

Water temperature increase from thermal pollution of the creek is going to have an affect on the eco system. Just sayin' --

Bob

 

[Riley]

Water temperature increase from thermal pollution of the creek is going to have an affect on the eco system. Just sayin' -- QUOTE]

Ya know this kind of generalizing really gets me; I'd wager, if I was a betting man, that logically, you couldn't prove any affect on the eco system short of adding a requirement for a reactor level of cooling.

You'd have to know the flow rate of the stream, the volume of water re-routed and returned to the stream through the system, the change in temperature specific to the system opposed to any solar or environmental temperature gain added during the time the water was removed then returned to the stream. Then, figure the impact of the re-route temp increase to the total volume and flow rate of the stream. Then compare that with the natural warming that occurs during the course of the streams flow.

You'd be hard pressed to convince me that it would be significant or likely even measurable more than an insignificant distance from the outlet of the return. Kind of like peeing in the ocean.......

Broad, sweeping generalizations are not always accurate when you actually consider the problem at hand.

 

[whatuusay1]

Water temperature increase from thermal pollution of the creek is going to have an affect on the eco system. Just sayin' -- QUOTE]

Ya know this kind of generalizing really gets me; I'd wager, if I was a betting man, that logically, you couldn't prove any affect on the eco system short of adding a requirement for a reactor level of cooling.

You'd have to know the flow rate of the stream, the volume of water re-routed and returned to the stream through the system, the change in temperature specific to the system opposed to any solar or environmental temperature gain added during the time the water was removed then returned to the stream. Then, figure the impact of the re-route temp increase to the total volume and flow rate of the stream. Then compare that with the natural warming that occurs during the course of the streams flow.

You'd be hard pressed to convince me that it would be significant or likely even measurable more than an insignificant distance from the outlet of the return. Kind of like peeing in the ocean.......

Broad, sweeping generalizations are not always accurate when you actually consider the problem at hand.

Simply wanted to say

 

[armorer]

I just assembled a test setup in my shop using the same idea. I have a well about 30' from my shop I used a 1/3 HP well pump plumed into the shop with 1” PVC to a radiator and returned back to the well. I made a shroud for the radiator and installed a squirrel cage blower. My well water is 58 Deg. and I have been getting 68 Deg. air out of the radiator pulling in 98-110Deg. shop air.
The final system will use a bigger radiator with a diffuser in the shroud as the temps on the outer 1/3 of the radiator are 15-20 Deg. cooler than the center where the airflow in concentrated.

 

[uniballer]

well, in my area(va) we can pull all the water out we want. but to put it back it cannot go in the creek, directly. there is something called a french drain, two foot square gravel pit surrrounded by landscapers cloth. top 12" down, 12 feet or more away from creek. the rocks are natural filters, and it is code, well here. I have been doing plumbing for 13 years, and done this a d-day memorial where i live, for the outside drinking fountains. and for seasonal cabin users.

 

[kbs2244]

I like the overall idea.
But I think the radiator is the problem.

You need higher flow rate at a lower pressure.
An old cast-iron hot water or steam heat radiator would be ideal.

3 inch gravity feed to a header tank, with an overflow outlet feeding around to the radiator output. 2 inch flow through the radiator. You want an over sized header tank to feed the radiator to keep the system running with out a need for pumps.

If it is a high humidify area you will need some kind of condensate collection under the radiator.

 

[pseudorealityx]

I don't understand how people keep mentioning that there is no need for pumps.

In the 2nd post by the OP, he says the creek is BELOW the heat exchanger.

 

[Hades12]

look up a water ram pump it uses the water as the power for the pump.

I did a quick setup like this with city water. going into the rad was 81 comming out was 96. It was a real hot day.

 

[mrobins297aaa]

Sounds like a good idea on paper.

Forget the box fan save your $15 it'll never move enough air and if you add any ductwork to it you'll be lucky if it moves any there just not made for that. You need something like a used furnace blower, there cheap and you can fin them on CL for not much more than your box fan.
you didn't say how big the garage is but lets just say it needs a ton of cooling (12,000 btu's) that means you need to move 400cfm which you could do with a small furnace blower on low speed.
now thats what you need for the air.

as for the water: 43 deg water has a certain amount of btu's in there and there are ways to calculate how much water you need to move thru the exchanger, which I can't help with but I think your going to need to move a lot of it and your defintely going to need pumps.
the worst thing going for you with chilled water is it can't absorb any heat without getting warmer itself which is a real hinderance........now if you had ice that would be a big difference because it can absorbe heat without getting warmer as it changes state to water.
Anyway maybe someone with chime in with the amount of water you need to flow thru your exchanger to cool the garage.......assuming its heat gain is 1 ton.

 

[pseudorealityx]

Sounds like a good idea on paper.

Forget the box fan save your $15 it'll never move enough air and if you add any ductwork to it you'll be lucky if it moves any there just not made for that. You need something like a used furnace blower, there cheap and you can fin them on CL for not much more than your box fan.
you didn't say how big the garage is but lets just say it needs a ton of cooling (12,000 btu's) that means you need to move 400cfm which you could do with a small furnace blower on low speed.
now thats what you need for the air.

as for the water: 43 deg water has a certain amount of btu's in there and there are ways to calculate how much water you need to move thru the exchanger, which I can't help with but I think your going to need to move a lot of it and your defintely going to need pumps.
the worst thing going for you with chilled water is it can't absorb any heat without getting warmer itself which is a real hinderance........now if you had ice that would be a big difference because it can absorbe heat without getting warmer as it changes state to water.
Anyway maybe someone with chime in with the amount of water you need to flow thru your exchanger to cool the garage.......assuming its heat gain is 1 ton.

BTU = gpm * 500 * delta T (of the water)