Heat Exchanger Update

[Eric Johnson]

I finally got some of my old 1" copper runs replaced with pex-al-pex, so now I have the raw material to build the heat exchangers I've been planning. I put one of the header units together tonight. After getting all the initial measuring and figuring and prep work done, the actual assembly went pretty fast.

There will be 8 in total, out of which I plan to make two heat exchangers for my storage tank--one for heat storage and one for recovery. This project is the result of my being too cheap to buy soft copper coil. So in total, my tank will have 200 feet of Type M 1/2-inch copper and another 80 feet of 1" Type M copper to get the heat in and out.

The plan is to connect four units with a 1" manifold arrangement so that they can be positioned vertically in the tank. The heat storage hx will sit on the bottom and extend up about 3/4 of the height of the tank. The heat recovery hx, on the other hand, will hang from the top of the tank to 3/4 of the way down. They will be nested together. Basically, the top hx will be attached to the bottom one in the right position, so the whole thing sits on the tank floor and extends to near the top.

That's the plan, anyway. Any reason to think it won't work?

 

[Nofossil]

I finally got some of my old 1" copper runs replaced with pex-al-pex, so now I have the raw material to build the heat exchangers I've been planning. I put one of the header units together tonight. After getting all the initial measuring and figuring and prep work done, the actual assembly went pretty fast.

There will be 8 in total, out of which I plan to make two heat exchangers for my storage tank--one for heat storage and one for recovery. This project is the result of my being too cheap to buy soft copper coil. So in total, my tank will have 200 feet of Type M 1/2-inch copper and another 80 feet of 1" Type M copper to get the heat in and out.

The plan is to connect four units with a 1" manifold arrangement so that they can be positioned vertically in the tank. The heat storage hx will sit on the bottom and extend up about 3/4 of the height of the tank. The heat recovery hx, on the other hand, will hang from the top of the tank to 3/4 of the way down. They will be nested together. Basically, the top hx will be attached to the bottom one in the right position, so the whole thing sits on the tank floor and extends to near the top.

That's the plan, anyway. Any reason to think it won't work?

Awesome soldering. If you pass leak test with no problems, I officially hate you forever

Your arrangement makes sense, except I'd be inclined to have one horizontal just below the surface for the last part of the heat extraction set - If you have a layer of hot water at the top, you want your exiting heated water as close to that temp as possible.

 

[Eric Johnson]

Thanks. As I've said before, you can inspect these joints in detail. When you put that pipe into a conventional fitting, you can't see the solder take to the joint, and some problems are inevitable. So you're really flying blind. Here I inspect and clean each connection, so if the silver solder holds, it won't leak.

I'll post some pics of the sequence of events once I get into production mode. I had to buy a drill press and an expensive drill bit to cut the holes, but that's something I've always wanted, so it's nice to be able to "write it off" on a project.

 

[WRVERMONT]

Good to know there is some experimentation going on out there!

 

[Eric Johnson]

I got four of the eight units built. What you see here is one heat exchanger--either the storage or recover hx. They're both basically the same. The heat recovery unit will nest inside of this one.

I'm curious about the best way to pipe it. Obviously, it's just stuck together now to show what it looks like. I have to introduce both a 1" and 3/4-inch line into the storage hx, but the recovery unit will be only 1" in and out.

Once I got into production mode, making these sections got pretty easy. I developed a few tricks to speed things along and add some consistency to the project. Interestingly enough, the soldering is a relatively minor part of the workload.

Anyway, this is where I'm at with it. Gotta buy some fittings once I figure out how to pipe it.

 

[WRVERMONT]

Have you thought about using both heat exchange loops for Charging the tank as well as Recovering heat from the tan, at different times(obviously).

 

[Eric Johnson]

That's the way it's usually done with a coil. I prefer the simplicity of having dedicated units for each function. You can position them in the tank for optimum functionality and the way I see it, draw from the tank at the same time you're charging it. Two separate pumps, each doing their thing, as circumstances dictate.

 

[Nofossil]

That's the way it's usually done with a coil. I prefer the simplicity of having dedicated units for each function. You can position them in the tank for optimum functionality and the way I see it, draw from the tank at the same time you're charging it. Two separate pumps, each doing their thing, as circumstances dictate.

I considered that, but I couldn't come up with a scenario where I'd want to put heat into the tank and take it out at the same time. I ended up putting all my copper in one coil. Of course, I have two others for other purposes....

 

[Eric Johnson]

It's probably obvious that I don't have a great technical mind, so I tend to go with what I understand. And sometimes understanding relatively simple concepts takes some effort.

Anyway, it seems to me that if you have one coil that flows both ways, you can't really optimize both input and output functions, since at times your water is either going to be entering or exiting at the wrong place in the tank for what you're trying to achieve. Not a big factor in the real world, probably, but something to consider.

Another consideration is the tank's role in the system. Is it just another zone off the wood boiler or is it the primary concern of the wood boiler? If you have dedicated heat exchangers, then each works at the convenience of the part of the system that it's serving. You can be charging up the tank with the hx on the bottom, while your house is drawing the hottest water from the top--simultaneously. No compromises.

I guess I don't understand how a two-way coil works. Seems to me when there's a call for heat from the house, you either have to bypass the tank or you have to stop using the coil for storage for the duration of the heat call. Bypassing the tank makes the most sense. I imagine that's what you do.

It would be no trick for me to pipe these two heat exchangers together and do the same thing, and maybe I will. But at the moment I can get my mind around the two hx concept with enough confidence to proceed.

BTW, one of my hvac friends on another website told me how to do the piping. Reverse return is what he recommends. Here's the diagram. Beautiful, no?

 

[leaddog]

I put my hx from the boiler in the bottem of the tank and have my hx going to the house on the top. That way I can have 180f water at the top and the water at the bottem could be 150f and being heated with the boiler. as I call for heat from the house it pulls the heat from the top and the return water then will flow down towards the bottem. seems to work well. I also have both ex piped with a manual valve that I sometimes open when it is warm outside and I need more hx capacity going into the tank . It seems to work well
leaddog

 

[Nofossil]

In my system, I only have one heat source active at a time. If I'm heating with wood, I'm not heating from the tank and vice versa. If I'm heating with wood, my first priority is any house zones that need heat. If they're satisfied, I dump heat into the storage tank.

When I'm heating the tank, I'm using the pump on the wood boiler, and the tank's pump is bypassed by a zone valve. Flow is top-to-bottom. That way, the top of the tank gets as hot as possible, and the return water is as cold as possible.

When I'm drawing heat from the tank, I'm using the tank's pump, and flow is bottom-to-top. That way the water coming out of the tank is as hot as possible. Took me a few tries to noodle it out, but it works great for me. Looking forward to data on yours. One thing for sure - your joints look about a hundred times better than mine.

 

[Eric Johnson]

One thing for sure - your joints look about a hundred times better than mine.

Not so fast, bro:

http://forums.invision.net/Thread.cfm?CFApp=2&Thread_ID=51427&mc=11

 

[Eric Johnson]

So, assuming that your coil extends from the very top of the tank to the very bottom (which would be required for the max Delta T in both storage and recovery functions), then when you're charging the tank, the return water is passing through the hottest part of the tank as it exits, pulling heat back out of the tank. And conversely, when you're recovering heat through the coil, the return from the zones is entering the coldest part of the tank. Wouldn't it be more efficient for those entry and exit points to be closer to the center of the tank, say 2/3 or 3/4 of the way from either extreme? Seems to me you're compromising efficiency for convenience and economy.

That about sums up my understanding of heat exchange theory.

 

[Nofossil]

So, assuming that your coil extends from the very top of the tank to the very bottom (which would be required for the max Delta T in both storage and recovery functions), then when you're charging the tank, the return water is passing through the hottest part of the tank as it exits, pulling heat back out of the tank. And conversely, when you're recovering heat through the coil, the return from the zones is entering the coldest part of the tank. Wouldn't it be more efficient for those entry and exit points to be closer to the center of the tank, say 2/3 or 3/4 of the way from either extreme? Seems to me you're compromising efficiency for convenience and economy.

That about sums up my understanding of heat exchange theory.

When I'm charging the tank, it's top-to-bottom. The exiting water is passing through the bottom (coldest) part of the tank as it leaves.

You're right - when I'm recovering heat, the zone return is entering the bottom, which may be colder than the returning water. In that case, I lose stratification (but not heat). I intend to add a variable speed pump and a radiant zone in series with the return water to ensure that doesn't happen as much- discussion for another thread and another day.

You're also right that for either function, the ideal coil location is different. My analysis says that for any given amount of copper, the compromise that I settled on works better. However, that's analysis, not data. As I said before, I'm eager to see data from your configuration.

Any engineering project is chock-full of compromises - that's half the fun.

 

[rreihart]

Just curious, how much temperature difference do you see top to bottom.

 

[Eric Johnson]

When I'm charging the tank, it's top-to-bottom. The exiting water is passing through the bottom (coldest) part of the tank as it leaves.

Many different ways to skin a cat, I guess. The designs I'm familiar with all put the hottest water for storage into the coldest part of the tank for the highest Delta T. But that obviously involves sending warmer water back to the boiler.

I'm getting two different suggestions on how to pipe the top hx. One school of thought draws water for the zones from the top of the tank. The other advocates a counterflow arrangement, whereby the flow in the bottom hx is bottom to top, but the top hx flows top to bottom, i.e., water for the house zones comes out the bottom of the top hx.

Is this all academic, or does it really make a difference with the temps and tank volumes we're dealing with? Another school of thought says: Just get enough copper surface area in the tank and forget about it.

 

[Nofossil]

It does make a difference - much more so for extracting heat than for putting heat in. In general, you want the highest possible heat out of the tank. During heat input, a coil at the bottom will set up convection and the heater water will rise, at the expense of stirring up the tank and messing up thermal stratification.

I get 25 - 40 degrees difference between to and bottom. Of course, I'll get a lot less if I push enough heat into it, as the bottom will get heated as soon as the top starts to approach the temperature of the incoming water.

Here's a graph from a recent fire - the tank has sensors near the top, middle, and bottom. Only 20 degrees stratification in this one, but you can see what's happening.

 

[Eric Johnson]

I never doubted you, nofo, but thought the old RR "trust but verify" approach seemed in order. My guru at the Forest Products Lab assures me that you're right about the top to bottom flow on the heat storage hx. He's also the guy who convinced me to go with two instead of one. I assume the top hx for recovery one would still flow up, right?

BTW, what do you recommend for rudimentary instrumentation? My inclination would be to put meat thermometers on the lines going in and out of the tank, but something a little higher-tech might be more fun. I don't think I need data ports--just a few good digital thermometers with remote probes that operate in the right range.

 

[slowzuki]

I feel quite spoiled for my hx design. Using in floor heating I don't need very warm water so the coil can be designed for charging the tank and will work fine for extracting heat. I'm actually going to try the coolant coils that are attached to the inner liner of my milk tank. If I pull from the shell of the tank artificially reducing the dT across the insulation I should have more of the heat loss delivered to my floor.

Ive been doing some really on zone valves too and it seems most solar places want more pumps and less zone valves to reduce the electric demand from boilers.

 

[Nofossil]

I feel quite spoiled for my hx design. Using in floor heating I don't need very warm water so the coil can be designed for charging the tank and will work fine for extracting heat. I'm actually going to try the coolant coils that are attached to the inner liner of my milk tank. If I pull from the shell of the tank artificially reducing the dT across the insulation I should have more of the heat loss delivered to my floor.

Ive been doing some really on zone valves too and it seems most solar places want more pumps and less zone valves to reduce the electric demand from boilers.

I'm gradually switching to the Taco EBV - virtually no power consumption. It's a big deal for me because I need to be able to run on battery power. I only ever have one pump on at a time as well.

 

[Nofossil]

I never doubted you, nofo, but thought the old RR "trust but verify" approach seemed in order. My guru at the Forest Products Lab assures me that you're right about the top to bottom flow on the heat storage hx. He's also the guy who convinced me to go with two instead of one. I assume the top hx for recovery one would still flow up, right?

If it was mine, it would ;-)

BTW, what do you recommend for rudimentary instrumentation? My inclination would be to put meat thermometers on the lines going in and out of the tank, but something a little higher-tech might be more fun. I don't think I need data ports--just a few good digital thermometers with remote probes that operate in the right range.

Rudimentary? What's that?

They make digital meat thermometers in wireless and wired versions. They cover the right range, and there's nothing like knowing that you're tank is 'well done'. Low cost digital thermometers use thermistors as the sensing elements. That's good and bad:

Good: you can extend the wire to really long lengths without affecting accuracy.
Bad: The sensing element must NEVER get wet.

Of course, you must have a non-contact infrared thermometer as discussed in other threads.

 

[slowzuki]

Eric we haven't heard much about your coils lately! Is it just a case of she is working well without the tank?

I've got that Jetstream just about all together, I just bought a new aquastat and the fittings for my coils. Hopefully the weekend will be nice and warm out for working on it.

 

[Eric Johnson]

It's not coils, Ken, it's a rigid copper header design. Two heat exchanger units, actually.

I haven't got it going yet. The boiler has been working really well without storage, so I haven't gotten around to hooking everything up yet. There are about 3 major things that could go wrong, but that's nothing new. Once the warmer weather sets in, I'll be back at it. I'm going to need the buffer and storage, especially in the summer for the DHW.