Going to an in-tank heat exchanger.. Advice needed.

[deerefanatic]

Hi all. After my experience with rust in my open tank, I've decided to do away with the flat plate heat exchanger and go to either copper or Pex coils. My reasons are this:

1. 1 less pump to run which means electricity saved and longer run-time on battery power during a power outage.
2. No need for rest chemicals ($100 for my 1200 gallons storage) and no need for maintenance on said chemicals. Just an occasional ph check.
3. Ability to heat DHW for the house more efficiently in summer than the current setup of pumping water out of storage, through heat exchanger into system plumbing, then into house and through another heat exchanger and pump before heating hot water..... The current way uses 5 pumps vs using 1 pump to circulate DHW through it's own dedicated copper coil.
4. Better stratification as my current system mixes the water up too much and these old milk tanks produce some issues with remedying that.
5. Less heatloss from tank as the current setup thermosiphons and will reduce 120 degree tank water to 60 degrees within 4 days.... (And this is an insulated tank, with extra R38 rolled insulation piled on top of it.)

So, how much should tubing should I use? I'd like to go copper if I can swing it. I NEED to be able to sink the full output of my new EKO 60 into this as I'll be heating the storage during summer for DHW purposes, and there wont' be other heat loads. I realize that once the tank starts getting warmed up, that I'll have some idling going on, but I'd like that not to happen until the top of the tank is around 150 or so....

So, what's the recommendation? I assume 1/2" tube would be best, with multiple coils in parrallel as more of the water is in contact with the walls of the tube vs large diameter tube.

My gut feeling says 300 feet of copper or 1200 feet of pex would work. For the DHW loop, one 50' loop of copper or 200 feet of pex.......

Am I on the right track here?

 

[Duetech]

Copper will give you a better heat transfer. If you go half inch you will need four 1/2" to equal a one inch and six half inc to equal a 1 1/4" system line. You have to be careful with the size of pipe you use in the tank as too small of a diameter will not allow your system to dump the heat into the tank fast enough to prevent idling. I know that coils work but I do not think they are the optimum in efficiency for a one pump setup. Something like a four core truck radiator would work real well for charging the tank if they would take the pressure as radiatior caps bleed at 16-18#. Building your own would be labor intensive but put your "heater" in the best place to get the greatest long term delta. How you "flow" your water (conventional or counter flow) for charging with coils will effect the window for idling along with fuel used. Extraction is another story similar to the heater but with a wider array.

 

[Nofossil]

I'd really consider buying some copper flashing and bending up some clip-on fins. You can never have too much surface area.

 

[deerefanatic]

Ok. The coils will be piped with two pumps facing different directions. counter flow will be used to both charge and discharge the tank. Same coils both ways.

I'm leaning towards 1/2 inch refrigeration tubing (seems refrigeration tubing is MUCH cheaper)... Which equals 3/8 nominal. Really small I know, but if I use lots of loops, I can get the flow through the HX, while having more of the water in contact with the tubing walls..... Right now, 1/2" OD x 50' coils is $55 from pexsupply on sale. I was thinking 8-50' coils in parrallel. Plus 1 3/4" x 50' coil to heat my DHW (this would be a draw-off coil only)

Does this sound like I'm on the right track? The bill for 8 - 1/2 OD x 50' coils would be $440. The Bill for 1200' of 1/2 " pex would be $465. So in this case, copper is coming out cheaper.

 

[deerefanatic]

As for the idea with the fins, I could make a bunch and solder or braze them on... That'd be ideal for thermal transfer.

 

[deerefanatic]

Here's another possibility that would flow better than the copper coils too:

DoIt hardware has 1/2" (nominal) x 10' Type M copper for $7.49. That comes to $299.60 for 400 feet of copper. Place vertical copper fins on the tubes and braze them in a grid like Eric did. The only downside is lots of labor to braze all that, and having to do so inside the tank.. But, I'm willing to do that to save money. Plus, the thing will be self supporting in the tank too...... (I'll probably still use a soft copper coil for the DHW.)

 

[tom in maine]

I think fins need to be parallel with the downward movement that will occur as the tank stratifies, otherwise, they will obstruct water flow over the coils that occurs with heat transfer. (this would not be good!) They are not that simple to install and will not give you the performance that everyone wishes for.

Fins work good in air and with small coils. This means a LOT of fins, not the amount a DIY'er can install.

I would lose the concept of using fins and use more tubing, although you are on the right track. There is enough info on this site for fin-less coils that perform well.

Tom
americansolartechnics.com

 

[deerefanatic]

My thought is a grid of vertical pipes. Then take a strip of 3" wide copper and weave it between the pipes sort of like a basket weaver would do. Then solder/braze to each pipe. The fins that way would also be vertical and be running in the same direction as the water will be as it rises/falls. I'll have to see if I can figure somthing out for a diagram or picture.....

I'm leaning seriously about using the hard copper. Cut to 5' lenghts. To obtain 400 feet of heat transfer would make for 80 parallel runs of 1/2 inch pipe. According to my handy pocket ref, the head loss would be in the neighborhood of .0125 feet for the entire heat exchanger at 10 gpm...... In other words, my supply piping will be my limiting factor.

 

[deerefanatic]

I'll also throw out that 400 feet of 1/2" copper has a surface area of 65.5 sq ft. That doesn't include the surface area of the 3/4" and 1" pipes that will connect them all together.

My current flat plate had 23.54 sq ft of surface area. So, that means that my in-tank will have 2.8 times as much surface area.

 

[WRVERMONT]

Hi,
For your Boiler's output, I would use a minimum of 4 120' coils piped in parrallel using 3/4" copper. Having your coils stretch nearly the full depth of your tank/water is always good for improved stratification. Good luck with your set up.

 

[deerefanatic]

I agree, full depth is the best bet.....

4x120 feet = 480 feet of tubing.... WOW...... $$

I need to ask Eric about his system, and see what he's using for footage.... He's got an EKO 60 like me......

Oh well. Didn't say it'd be cheap. Just better.

Course, there's always the Pex option......

 

[Rory]

If it helps, I have a Tarm Solo 30 with 670 gallons of self-built storage. I just got done heating this week's water. My primary heat exchanger is two 120' copper coils plumbed in parallel, and my DHW is simply 180' of copper in the DHW line. The water is always plenty hot enough.

My boiler rarely shuts down for overheating, but perhaps your EKO 60 would be a different story.

 

[deerefanatic]

well, the difference between a tarm 30 and eko 60 is exactly double.... So I'd need double the heat exchange area. What size diameter did you use for your HX?

Plus, here's another thought. My tank is an old milk tank. Milk tanks have two sets of coils in the bottom to cool the milk. One set is bad in my tank, hence the reason I've got it now... But, makes me think I could use the bottom coils also when charging the tank to put a few more BTU's into the tank... After all, the water at the bottom is the coolest.... so the most transfer...... But, for discharge, these coils won't serve much purpose so I'd probably figure out a way to only pump water through them during charging... Just an extra thought.

 

[EricV]

I am pretty sure you can buy the aluminum fins that are used in baseboard heating and slip them 3/4 pipe. Check your local heating supply house. They slip on and stay in place with their own tension, no soldering.

 

[Rory]

What size diameter did you use for your HX?

It's all 3/4".

 

[tom in maine]

Please, do not use aluminum fins in a tank of water with copper tubing. This is a recipe for an electrochemical mess.
The fins will corrode and cause all types of issues. And it is a waste of money.

And, smooth tubing is a far better performer in an immersed situation.

Most everything that has been mentioned thus far in this thread (by way of coiled tubing) will work as desired.

Remember, even if you screw up a heat exchanger design, the worst thing that is going to happen is a smaller delta T on the space heat
exchanger or diminishing DHW temps on the DHW hx. This is unlikely, given what you all have mentioned as designs so far. And these are easily remedied by
modifying the hx.

Coils work and are the simplest way to do this. A bunch of Tee's and rigid copper may make you feel like a great fabricator, but lineal footage (copper or PEX)
with attention paid to pressure drop is what these installations need! Also--more joints can mean more leaks.

 

[deerefanatic]

Ok, here's my idea.... I plan to make 5 grids, 6 foot long by 4 foot tall, with 20 - 1/2 inch pipes each running vertically. These 5 grids will be manifolded together..... For a total of 400 feet of 1/2 inch copper, plus 60 feet of 3/4 inch for the headers. I will use pex to carry water to/from the grids so that I don’t loose too much heat from the tank due to the water going to/from the bottom of the grids passing through the upper layers of water (due to the fact that pex doesn’t have as good of heat transfer as copper.) Then the 5 grids will be manifolded together external of the tank with balancing valves. By using lots of copper in parallel, it will keep flow velocity down and also keep head pressure loss to a minimum.

While I like the idea of coils, they're just too cost prohibitive........ Pex though is tempting...... How much pex do you guys think I need?


And here's a new thought that just occured to me......being my tank is an old milk tank, I have a set of coils in the bottom shell of the tank that I can use when charging..... Flowing water through these makes sense when charging as they are at the coldest point in the tank..... They're designed for like 15,000 btus of coiling capacity in a refrigeration environment.

 

[deerefanatic]

Ok, I'm going to expand this out to 500 feet of 1/2 inch rigid copper running vertically in 4' sections..... Plus 3/4 inch for header pipes. Probably 6 grids piped in parallel.

Any good reasons why this won't work? I'm going to Menards tomorrow to buy my chimney and I'll be buying this pipe then, so I need to hear any last minute advice soon..

 

[tom in maine]

It will work.
Probably too big, but that is not a bad thing.
Pressure test it before sealing the tank.

Good Luck.

 

[Medman]

What do you figure the pipe and fittings will cost? I'm just curious; I am going with pressurized storage for my system.

 

[deerefanatic]

Well good... better too big than too small.

Well, 1/2" rigid copper is $6.49/10 feet........ So not too bad.... I'm not going to use fittings..... I'm gonna drill into the header pipes and braze everything together.... More permanent and much cheaper than the hundreds of fittings I'd need.....

Don't worry, many weeks of pressurized testing before filling the tank.... Plus, the tank has a manhole so I can get inside it.

 

[deerefanatic]

I now have a pile of copper on m shop floor.......... What a beautiful sight!

 

[hkobus]

Matt,

Do you have an update on this project? Maybe I missed the post, but I have the same delema between PEX and copper. Actually to be fair I should say that copper is out at this time and PEX is in on the cost of material.
The key again is what is the right amount of tubing to put in to sink the full load of the ECO 40 or even a bit more. I am looking at 4 coils of 1/2 inch PEX parallel to flow a full 1" copper feed and the same system for charge and discharge.
Do we have any other members here with PEX coils other than the big 5000 G concrete inground tank with some experience on the system? Or anyone that has the formula to convert copper performance to PEX.

Thanks,
Henk

 

[Gooserider]

I've seen some mentions in the past, so the search might help. I'm not at all positive, but I think the numbers being tossed around were that you needed something like 4x the amount of PEX as you did copper to get the same heat exchange. From that I would wonder how bad the cost of doing copper is in actuality. The other thing that I would see as a downside is the challenge of stuffing all that PEX into a storage tank and still having room for the storage water...

Gooserider

 

[hkobus]

The search didn't do much for me, I know we discussed this last year with no definite answer on the numbers. I had an engineer friend over tonight and we did some of the numbers and the closest we got was that if copper has a 70 Btu/hr/ Ft2/ *F (as per Nofossil's spreadsheet), the Pex has 23 Btu/hr/ Ft2/ *F. This than looks like I will need 3X the Pex and in order to get the 135,000 Btu I think I am looking for, I would need 900 ft 1/2" PEX :bug: or 295 ft 1/2" copper.
Hmmm, copper does not look that bad maybe, I just know I have no time for endless soldering or brazing...