Questions about radiant heat system

[mx842]

Question

The tank is connected to the stove via thermo-syphon flow , correct? When you see 140degree water into the 20 foot line leading to the manifold, WHAT IS THE WATER TEMPERATURE at the bottom of the tank ? That would be the supply TO the wood stove. Is that also 140? It would be a good sign if it is.

I'm not sure, I can't remember if I ever checked the temp there I know it always feels cold when compared to the hot line going upward. I guess as that hot water leaves the tank out the hot supply line it is drawing cold water into the tank via the return from the manifolds keeping it cold at the bottom and hot at the top of the tank. Is that even possible? I often wondered what would happen if I left the pump off longer just to see how far the hot water would travel down the line to the manifold on it's own but never did try it. I think I did try once cutting both of the valves off at the supply and return at the top of the storage tank and let it run for awhile as I remember I did have hot water at the bottom supply line to the stove. That was awhile back and I think I was messing with different valves trying to figure out how to get the return temps up then it got really cold and I decided the sending 110 degree water down the line was better than nothing once I figured out how to do that I just left it alone.

 

[mx842]

mx,

One way to measure how many BTUs you are able to get from the fire would be to put a small circulator on the HX lines to the water heater. This would stir the tank enough to make it's entire temperature measurable over time and make it the same on the bottom as on the top.
Then you can see how long it takes to rise 40 gallons 100 degrees, for instance and calculate the BTUs you collected. The main floor area should see a minimum of about 10 BTU per sq ft to make it rise in a useful way. 25 degrees per foot would be better. 50 would be better yet.

Here's the suggestion: Install a circulator on the HX lines. Determine your BTU availability from the fire with temperature over time and gallons in the tank. Calc how many BTUs you need based on floor area.

Then you will know if you can make it work or not.

Then increase the flow rate through the floor so you can evenly deliver the energy.

That may be worth a try I don't have a small pump I could use. Would closing the supply and return valves going into the storage tank do the same thing by just letting it circulate by thermo-syphon flow and see how long it takes to get the temps at the top and bottom about the same?

I need to get a pump anyway but hate to buy something that's just to experiment with.

 

[Raspy]

mx,

You need a different pump for the floor, so you could use the one you have there now (B&G Series 100), on the HX. Shutting off the tank and letting it sit, if that is what you meant, will not equalize the temp. It will allow it to stratify and be at different temps top and bottom.

Do don't have to equalize the tank, but it was a suggestion to find out how many BTUs your system was producing so you could determine if it's capable of heating the slab.

 

[Raspy]

mx,

One more thought. I'm assuming your loop lengths are all about the same. So, if you get the manifold loop metering valves and shut off valves all wide open such that you are getting a gallon per minute flow, or even 3/4 gallon, per loop, that pump will be OK. Then you could find the smallest Grundfos or Taco pump for instant water recirc systems and put it on the HX circuit.

That would be cheaper, but you must prove first that the B&G pump can deliver to the loops.

The thermosyphon setup you have obviously works, but is not set up in the most affective way. You'll have to maximize everything in the system before you can realistically hope to heat your floor.

 

[mx842]

mx,

You need a different pump for the floor, so you could use the one you have there now (B&G Series 100), on the HX. Shutting off the tank and letting it sit, if that is what you meant, will not equalize the temp. It will allow it to stratify and be at different temps top and bottom.

Do don't have to equalize the tank, but it was a suggestion to find out how many BTUs your system was producing so you could determine if it's capable of heating the slab.

Ok, I see what you are saying, I was just thinking that with the two valves turned off at the top of the storage tank and with the hot water going in to the top and the cold water being drawn out the bottom that sooner or later it would come close to being the same temp......that is if it didn't blow up first I guess

I was checking around yesterday to see if I could find out how to figure out just how many gallons is in the whole system. I had all that stuff along with pictures of the layout of the pex throughout the building on another computer but it crashed and I lost all of that information. I could maybe recover it off the old hard drive but I just haven't had the time to mess with it. I know it is 8 loops and they all are from between 240 and 285 feet I do remember that. I was also trying to figure out what kind of head pressure that the supply and return loops plus the manifolds and floor loops came up to but I couldn't wrap my head around it.

 

[mx842]

mx,

One more thought. I'm assuming your loop lengths are all about the same. So, if you get the manifold loop metering valves and shut off valves all wide open such that you are getting a gallon per minute flow, or even 3/4 gallon, per loop, that pump will be OK. Then you could find the smallest Grundfos or Taco pump for instant water recirc systems and put it on the HX circuit.

That would be cheaper, but you must prove first that the B&G pump can deliver to the loops.

The thermosyphon setup you have obviously works, but is not set up in the most affective way. You'll have to maximize everything in the system before you can realistically hope to heat your floor.

One question how would I check to see if I'm getting the flow you suggested?

 

[Raspy]

mx,

The total gallons in the whole system doesn't matter. For calculation purposes, all you need are the gallons in the water heater tank to figure out your BTUs. Then the square footage of the slab to get the BTUs required (for rough numbers to determine if it will work).

The system pressure doesn't matter too much and doesn't affect flow rates or how much energy is delivered. Normal pressure for a hydronic system is 12 PSI. 10-20 is a fine working range. Always include a 30 PSI relief valve and an expansion tank to stabilize it. As log as it is full with no air, the pump will be happy. Normally you'd have a pressure regulator to fill the hydronic system from your street system and an air vent to get all the air out.

If you don't want to get that fancy, you could manually fill the system and leave a trapped air space above the pressure relief port on the side of the water heater. Don't use either of the top fittings and let the air accumulate there above the upper side fitting. This would become an air separator and expansion tank all in one. Then pump off the bottom of the tank toward your floor system and return through the upper side port. This upper port can also tee to your HX hot line. With this setup you'll have stabile pressure, good air management, the fewest parts, and the ability to separate the heat source from the delivery side to calculate BTUs. You can fill the whole system with gallon jugs of water and a funnel in the old "hot" port on top of the tank. Fill it to just above the upper side port and then cap the top two pipes (H and C). All air in the entire system will gather there which will get it away from where you don't want it, and it will stabilize the system pressure. Then you'll start out with atmospheric pressure in the whole system and it will work fine. It will probably climb to about 5 PSI when hot and the slab warm. Before capping the upper fittings, run the pumps for a while to get all the air to that area. Then add more water as needed to get the level above the upper side fittings. Then cap the "H" and "C" ports and you're done. Oh, one more thing. You can use the top "H" port to to install the 30 PSI relief valve, it will act as a cap and give you the safety you need.

Done. Fire it up! Get some number to see where you stand.

Review what I said about your PEX layout and direction of flow and make sure all loops are flowing as much as possible. You need another pump for the HX loop and, possibly a stronger one for the floor. Can you include a picture of the floor tubing layout that might show the design and tube spacing?

 

[Raspy]

One question how would I check to see if I'm getting the flow you suggested?

If you mean the flow through the floor loops, look at the flow meters on the manifold and adjust to fully open. The flow through the HX is less critical but should settle in to be nearly the same on the supply and return.

 

[quicksandfarmer]

Ok, I see what you are saying, I was just thinking that with the two valves turned off at the top of the storage tank and with the hot water going in to the top and the cold water being drawn out the bottom that sooner or later it would come close to being the same temp......that is if it didn't blow up first I guess

I was checking around yesterday to see if I could find out how to figure out just how many gallons is in the whole system. I had all that stuff along with pictures of the layout of the pex throughout the building on another computer but it crashed and I lost all of that information. I could maybe recover it off the old hard drive but I just haven't had the time to mess with it. I know it is 8 loops and they all are from between 240 and 285 feet I do remember that. I was also trying to figure out what kind of head pressure that the supply and return loops plus the manifolds and floor loops came up to but I couldn't wrap my head around it.

What is the diameter of the tubing in your loops? With that it should be easy to calculate the volume.

Half inch PEX is 0.00961 gallons per foot, so 2000 feet is 19 gallons. If you go up to 3/4 it's 0.01894 gallons per foot or 38 gallons for 2000 feet.

You can estimate flow rate using the flow curve of your pump.

 

[quicksandfarmer]

You can calculate the BTU/hour by multiplying the temperature change of the water times the gallons per minute times 500 (The 500 comes from the fact that BTU's are degrees times pounds, a gallon of water weighs 8.3 pounds so one GPM is 500 pounds per hour). This should be the same whether you measure it at the heat exchanger or at the floor tubing, heat is neither destroyed nor created when it transfers.

The question you're looking to answer is, what is the weak link? If the system isn't giving satisfactory performance, why not? My suspicion is it's going to come down to the stove. If it's not producing enough heat to warm the space, no amount of fiddling with the plumbing is going to change that. If it is, then you can make adjustments that will do a better job of distributing that heat in a more usable way and keeping your stove operating temperature up.

 

[mx842]

mx,

The total gallons in the whole system doesn't matter. For calculation purposes, all you need are the gallons in the water heater tank to figure out your BTUs. Then the square footage of the slab to get the BTUs required (for rough numbers to determine if it will work).

The system pressure doesn't matter too much and doesn't affect flow rates or how much energy is delivered. Normal pressure for a hydronic system is 12 PSI. 10-20 is a fine working range. Always include a 30 PSI relief valve and an expansion tank to stabilize it. As log as it is full with no air, the pump will be happy. Normally you'd have a pressure regulator to fill the hydronic system from your street system and an air vent to get all the air out.

If you don't want to get that fancy, you could manually fill the system and leave a trapped air space above the pressure relief port on the side of the water heater. Don't use either of the top fittings and let the air accumulate there above the upper side fitting. This would become an air separator and expansion tank all in one. Then pump off the bottom of the tank toward your floor system and return through the upper side port. This upper port can also tee to your HX hot line. With this setup you'll have stabile pressure, good air management, the fewest parts, and the ability to separate the heat source from the delivery side to calculate BTUs. You can fill the whole system with gallon jugs of water and a funnel in the old "hot" port on top of the tank. Fill it to just above the upper side port and then cap the top two pipes (H and C). All air in the entire system will gather there which will get it away from where you don't want it, and it will stabilize the system pressure. Then you'll start out with atmospheric pressure in the whole system and it will work fine. It will probably climb to about 5 PSI when hot and the slab warm. Before capping the upper fittings, run the pumps for a while to get all the air to that area. Then add more water as needed to get the level above the upper side fittings. Then cap the "H" and "C" ports and you're done. Oh, one more thing. You can use the top "H" port to to install the 30 PSI relief valve, it will act as a cap and give you the safety you need.

Done. Fire it up! Get some number to see where you stand.

Review what I said about your PEX layout and direction of flow and make sure all loops are flowing as much as possible. You need another pump for the HX loop and, possibly a stronger one for the floor. Can you include a picture of the floor tubing layout that might show the design and tube spacing?

Are you making a suggestion as to how to re-pipe the HX and holding tank setup permanently or is this just to check the BTU's the stove can put out? If it is to just check the BTU output of the HX and storage tank I can do that already by just shutting the two valves coming into and going out of the storage tank. That would isolate those two parts from the rest of the system completely. As far as filling the system I don't have a pressure reducing valve......well I do have one but just never saw the need to install it. When I fill the system I have a 1/2" valve piped into hot output port on the HX. I open this valve and the water back feeds through the HX and out of the cold side into the bottom of the supply tank. I have a fill valve teed into the cold water inlet pipe on top of the tank that I open a little to let the air out as this part of the system is filling. When only water comes out I shut off this valve at the top and open the return line valve and this sends water down the return line to the manifolds. At this point I have already flushed out any air in the manifold loops so I crack open the connection where the return line connects to the manifold and let it run until only water comes out. With the return line full I go back to the storage tank and shut off the return line valve and open the supply line valve so water can fill the supply line over to the manifold and I remove any air that is in that line the same way I did the return. Air has never been a problem.

I'm working on getting a small pump for the HX that's why I was wondering if this is how you are suggesting to re-pipe the tank. Also where would the cold water return to the HX go?

 

[mx842]

OH, I forgot to mention that my loops are 1/2" pex tube. I had a bunch of pics of the pex layout before I poured the floor but as I mentioned before all those pics were on my old computer that crashed. The loops are pretty much laid out on 12" centers with the ones that travel on the outside edges being 6" apart.

 

[quicksandfarmer]

I'm working on getting a small pump for the HX that's why I was wondering if this is how you are suggesting to re-pipe the tank. Also where would the cold water return to the HX go?

There's no reason your system can't run with a single pump, most boilers just have the one pump.

What you really need is two separate thermostatic controls -- one for the hot loop (heat exchanger) and one for the warm loop (floor). They can be mixing valves, aquastats, separate circulators, whatever, there's lots of ways of doing it. But your system's never going to give satisfactory performance if you don't have a way of keeping each part in the right temperature range.

Then, as Raspy suggested, fire it up and see how it works. You can tweak from there.

 

[Raspy]

There's no reason your system can't run with a single pump, most boilers just have the one pump.

What you really need is two separate thermostatic controls -- one for the hot loop (heat exchanger) and one for the warm loop (floor). They can be mixing valves, aquastats, separate circulators, whatever, there's lots of ways of doing it. But your system's never going to give satisfactory performance if you don't have a way of keeping each part in the right temperature range.

Then, as Raspy suggested, fire it up and see how it works. You can tweak from there.

This type of system cannot be left with no circulation when the fire is lit. If he depends on thermosyphon to warm the tank, it will eventually get too hot and get into trouble unless there is a load applied to deliver the heat. Or if the HX lines are restricted further, there will also be problems. The rise rate of the floor is very slow and will never be very precisely controlled. The simple and best solution is to simply run the pump or pumps when the fire is lit. This can be done by plugging them in when the fire is started. Regulating different temps with controls or adding restrictive parts in the HX lines will eventually lead to a problem and does nothing to improve the flow rates through the system. The floor is a very large load compared to the available energy, so there will never be runaway temperatures. The fire can be allowed to die out when the room temp gets near where he wants it. Later, if needed, he can make it more sophisticated.

 

[quicksandfarmer]

This type of system cannot be left with no circulation when the fire is lit. If he depends on thermosyphon to warm the tank, it will eventually get too hot and get into trouble unless there is a load applied to deliver the heat. Or if the HX lines are restricted further, there will also be problems. The rise rate of the floor is very slow and will never be very precisely controlled. The simple and best solution is to simply run the pump or pumps when the fire is lit. This can be done by plugging them in when the fire is started. Regulating different temps with controls or adding restrictive parts in the HX lines will eventually lead to a problem and does nothing to improve the flow rates through the system. The floor is a very large load compared to the available energy, so there will never be runaway temperatures. The fire can be allowed to die out when the room temp gets near where he wants it. Later, if needed, he can make it more sophisticated.

Running the pump when the fire is lit is what MX is doing now. In the very first post, this is what MX wrote:

"The biggest problem I have is that even dumping 110 degree water in the slab the return is always too cold when it gets back to the stove and this causes problems with creosote build up. I have to rod out the stove pipe every couple days to keep the keep it drawing."

The solution to this problem is to keep the heat exchanger temperature up. The way to do that is to have some sort of thermostatic control which can limit the amount of heat taken out of the heat exchanger if the temperature drops below the desired operating temperature. I recommended an aquastat on the circulation pump, but there's lots of ways it can be done. It's true that if a component fails the heat exchanger can overheat, that's going to be a risk no matter how you plumb it and that's why you have a pressure relief valve.

Properly sized controls won't limit the ability of the system to provide heat at all.

I also think he should abandon the idea of the thermosiphon and get rid of the storage tank, it's not contributing anything. The water volume in the heat exchanger should be just enough to avoid dramatic changes in temperature when operating. Oil-fired boiles seem to be around five gallons so I'd start there. More water just means longer warm-up times and more creosote production.

 

[Raspy]

quick,

Couple of problems with the design you are promoting.

This system is not a boiler and should not be compared to one. A boiler can be turned on and off with a simple controller or thermostat and is designed to do so. It also has internal controls to manage it's temperature. Not so with a fireplace. The only reason to keep the HX temp up is to drive the thermosyphon, but it has no value otherwise. Boilers must heat up to survive, if they are non-condensing, but not fireplace heat exchangers. If he reduces the water volume to the minimum, there would be no buffering affect and it would lead to boiling or explosion hazard in the HX. That is why the large tank is valuable, it stabilizes the system. It's not there to store heat for later, just to stabilize. Tempering it is not useful because there is no reason to try to deliver a certain temperature. The idea is to get as much from the HX to the floor as possible and not worry about temperature. It's not temp we want it's BTUs. Then the floor temp will come up as efficiently as possible.

Since he cannot get much heat to the floor and has no idea how much energy the HX is producing, I recommended a layout that would give him a chance to measure what he has and decide if he's on the right track or not. Tempering adds restriction, adds expense and gives no functional advantage. Just because they are sometimes used on hot water systems elsewhere, with gas heaters, is no reason to think they would be useful here. He wants an inexpensive system and he wants to identify why his is not working. His thermosyphon setup is not the most affective design and he needs much better flow. His heat source is small and will always be operating in an overloaded condition, thus, cool be design.

If a component fails, or the power goes out, this system can fall back on thermosyphoning to keep itself under control. He should never rely on a poor design that has to save itself with a pressure relief valve, even though he does need one for overall safety. It must not be designed to ever allow a dry HX sitting in a hot fire. A large tank, sitting above the HX, helps ensure this. An added value of the tank is that it can easily act as an expansion tank and air separator, both of which he does need. It also eliminates the need for a fill system. All while stabilizing the temp.

Heating of large mass slabs is done with boilers that start out at slab temperature, for the most part. The slab and the boiler heat up together in a long process that is limited by the boiler output, controller and the wall thermostat.

So, it doesn't matter what is "normally" done with gas boilers or that they are normally five gallons (not true). And it should not be designed that it has to be shut down quickly to save itself (wood fires die down very slowly). It needs a buffering mass to be safe and stable.

Simple is beautiful and simple is difficult to achieve. Testing at home with limited parts requires a simple design. Conventional boiler design sometimes goes out the window with experiments like this, and that's why it's fun.

It appears that his pump is only circulating the floor. Thermosyphon does the HX loop. He won't prevent creosote just by raising the water temp a bit. It would have to get up to 300 degrees or so to reduce the creosote.

 

[quicksandfarmer]

He won't prevent creosote just by raising the water temp a bit. It would have to get up to 300 degrees or so to reduce the creosote.

I think this is where you and I fundamentally disagree. And I'll say that if you are right about this, then it doesn't make sense to do any modifications to the system. It is already working at maximum efficiency. The water from the floor loop is returning at slab temperature, so 100% of the available heat is being extracted. The heat exchanger is as cold as can be, so maximum heat is being extracted from the fire. So all of the available heat is being used.

But I believe that it's possible to reduce creosote buildup by increasing the heat exchanger temperature. The 300F number you cite is the chimney temperature. There is going to be a temperature gradient between the chimney and the water, there's going to be at least a layer of metal for the chimney, an air space, and a layer of metal for the heat exchanger wall. I believe it's possible to have 150F water in the heat exchanger and 300F temperature in the chimney, and I believe it can make a big difference if the water is 150F or 50F. But I'll concede that so little is known about the system that it's impossible to know without trying.

MX gave us two design parameters: first, he wants to cut down on creosote buildup. Second, he wants to limit the water going into the slab to 110F to protect the slab. The only way to do the first -- if it's possible at all, and I'll admit it may not be possible -- is to raise the temperature of the heat exchanger. If you do that, then to achieve the second parameter the water coming out of the heat exchanger needs to be mixed. It may be possible to come up with a completely passive system that achieves that, using the physical properties of water, but MX has tried that already and isn't happy with the results.

In a wood-fired system, unless it is completely gravity run, you run the risk of boiling over if power goes out. (I guess you could have some sort of fail-safe that quells the fire if power goes out but that's getting complicated. Maybe the PRV valve could go into the stove.) With more water in the system, it will take longer to boil over, but when the boilover happens there's that much more water boiling over.

 

[mx842]

Man!!!! This is great and probably the best thread I ever started. That is, with regard to possibly coming up with some sort of positive result. I love it when two or more people who know about a certain subject can discuss different ideas without a big pizzing match blowing up the whole thread. I want to thank you guys for trying to help me get my system running the best that it is capable.

I loved it when Raspy brought up the fun aspect of this project because if this was just about heating my shop I would just move the stove into the main part of the shop and let her rip. The stove part puts out a lot of heat from the sides, in fact too much because I have to open the door to the boiler room sometimes because it's just too hot in there. Cutting those two openings in the walls helped a lot but it is still too hot in there for comfort. I like that room though, you can go in there after being out in the cold for awhile and it sure feels good. Nothing feels better than a nice wood stove on a cold day.

I haven't had time to mess with it much these past few days but hopefully I will get some time by the end of next week. That will give me some time to gather some parts and think about what needs to be done. You guys have given me some really good ideas and I have a few of my own so hopefully I'll get this thing going again before the cold weather sets in; today it's 85 degrees and sunny

 

[GLyford]

A long story to make a couple of small points.

My Dad had a heat exchanger in his woodstove in the 80's, before they stopped being allowed by code and the insurance. He wasn't using it for heat but for domestic hot water, and pumped it into parallel with his rooftop solar water panels, and tied into that exchanger and controller with a separate pump circuit. Out put calculation was at most recorded on electricity cost savings on the main water heater, the solar was a used as a preheater.

The heat exchanger was on the side of the firebox, and was not connected to the flue at all. The exchanger regularly had a thick coating of creosote on it, even with a hot fire to clear the flue at least once a day. By keeping the creosote buildup away from the flue, even though he lost an opportunity for harvesting heat he at least made the flue and fire easier to manage. I am not sure that you can afford that loss of collection area, but it is one way to reduce your creosote issue.

He found out that he was dumping more heat than he thought into the exchanger. He had his relief valve at the back of the stove plumbed into the basement, across the ceiling (maybe 15-20 feet) and out the wall, where there was an elbow pointing down. Near as we can tell, the relief was opening fairly often to vent steam. The steam as it was crossing the ceiling re-condensed, and drained out the elbow as water...until we had a cold snap (-10F) and the elbow froze solid. At this point, one of the sweated joints in the line popped open and vented steam. Once we figured that out, we removed the outside elbow to reduce the restriction, which fixed the issue as far as we could tell. So that's my other point: don't put a long line on a relief valve out to where it can freeze. Down at floor level is wise. A long run to an outside wall with an elbow, less wise. (And if it is opening often enough to be annoying, you are losing heat.)

 

[Raspy]

GL,

Yours was the way it used to be done. The heat exchangers were sold at fireplace stores and solar outlets. Problem with that setup is that it was plumbed to run when the solar ran. But you don't have fires on the same schedule as the sun, so it will stagnate and make steam. It also requires a pump to move he water through it and that can fail too. I've done them with a thermal switch near the fire to turn them on. Some of the old heat exchangers were a box type setup and they were more dangerous. Later ones were a loop of stainless pipe where they could not have a large volume of overheated steam.

These systems are fun to play with, but they always seem to have a weakness.

The creosote buildup on the heat exchanger in the fire box isn't a problem because it will only get so thick before it burns away. I think it's a better design than cooling the chimney.

Some systems are simply a coil of copper tubing wrapped around the outside of the flue pipe for several feet. This works to.

mx has the thermosyphon tank on his and it seems like it won't get into too much trouble. But the creosote buildup may be the deal breaker, unless he has very dry wood.