Radiant Floor Heating System Pre-Install Advice Sought

RickMac

Hi guys



I stumbled across your great site yesterday, and I'm grateful to have found you.



I purchased a 'package' radiant heating system for installation in my new-build home. I have installed the tubing and manifold and just wanted to seek advice on whether the rest of the system as drawn should work OK.



I'd much rather do that before I install the P/S loops, expansion tank, wye strainer, pump, Isotherm mixer and controls, and then not have to ask for help troubleshooting it afterwards.



I have attached the drawing of the system to this post and would welcome any advice, suggestions, tips, landmines-to-avoid etc.



Heat source is a Rinnai 190,000 BTU tankless heater, and primary loop will be 1" PEX. My first question is, should the pex be barrier or non-barrier?



I did find out about the suggested 12" of straight pipe prior to and after the two tees between the primary and secondary loops, and wondered if there were any other nuggets like that that I should know about?



Thank you



RickMac   



     

Jean-David Beyer

I am not a professional, but have a couple of suggestions.

1.) Do not put anything between the two closely-spaced Ts. They should generally not be more than 4 pipe diameters apart.



2.) Should you use use a Y-strainer in the primary loop? The problem with that is what happens if you do not frequently check it for being clogged up. If you fear dirt in the primary loop will be a problem, you may wish to use a low-loss header such as this one instead of the closely-spaced Ts (sorry link is so long):



http://www.google.com/url?sa=t&source=web&cd=3&sqi=2&ved=0CCcQFjAC&url=http%3A%2F%2Fwww.caleffi.us%2Fhr_HR%2FTechnical_brochures%2F01076%2F01076.pdf&rct=j&q=caleffi%20hydraulic%20separator&ei=6L3nTN6XBML98Abrh7HvDA&usg=AFQjCNHZtS2ZtZA4YH87u9ax8_-O1A0W4w&sig2=im01wVsCRg5negr0pMEgFg&cad=rja



The advantage of something like this is that the dirt separater is very much less likely to clog, and is easily cleaned. It also has an air separator that may be enough for your system.



3.) Some professionals do not care for hot water heaters to be used as boilers. In the case of boilers, there must be no valves of any kind between the boiler and the pressure relief valve.



4.) I do not think you need flow-check valves after the closely-spaced Ts.



5.) Is 1" Pex enough for 299K BTU/hr? For my boiler, the manufacturer requires 1 1/4" for their boilers between 155 and 230 BTU/hr, and 1 1/2" over that. Come to think of it, did you calculate the heat loss for your home? Is 199K BTU/hour enough? Is it too much?



5.) It is not clear to me why water should bother to flow through the radiant zones. This may be misunderstanding on my part, since I do not know what some of the components in the secondary loop are. E.g., an "iso-therm" or the circles with T and F in them. I assume they are thermometers and flow meters.



6.) Why no purge valves in the secondary radiant zones?



7.) It is now 2010. Why did you not consider a mod|con boiler with outdoor reset for this system?



Have you had a professional look at this?

kcopp

barrier pex....

non barrier pex is cheaper but then you need to get all non ferrous parts... and still then it can cause headaches w/ water quality issues. Did you say the tubing is already installed?  How long are the loop lengths? Is this an "open" system? Did you do a heatloss on the system? 

kpc

CMadatMe

Changes

Pipe size is dictated by the heat loss. What is the heat loss? Radiantly I would say 3/4" would be sufficient. The primary pump is in the wrong location should be closer to the tankless return pumping through the heat exchanger not away from it. I would also then move my expansion tank and feed so I am pumping away from the point of no pressure change.



I also don't see how any water is going to get through the radiant. The radiant pump looks like it is pumping back to the return on the primary loop. Don't know whay the valve is on the manifold. Is is a diverting valve? If so, what controls it. I assume something is providing out door reset to it but I don't see it in the diagram.



I'm not a big fan of the tankless as a heat source. There is no way to modulate water temp based on outdoor temp so it sends a fixed water temp. What temp do you plan on running the tankless at? What is the radiant application, joist with plates, under slab?

Jean-David Beyer

Pipe size.

"Pipe size is dictated by the heat loss. What is the heat loss? Radiantly I would say 3/4" would be sufficient."



That is certainly true in the secondary loop. I tried to find out how many gallons/hour are required to transfer 299K BTU/hr and I could not. I defer to your judgement here.



The greater pipe size dictated by my boiler manufacturer is specified for the primary loop only. The manufacturer is silent about what is in the secondary loop. I imagine they are concerned with getting enough flow through the heat exchanger in the boiler. This may not be a concern with a tankless hot water heater, but at lower flow rates, I imagine the firing rate is automatically decreased.



My view, as a non-professional, is that the O.P. needs a heat loss calculation done on his house.  This would permit the determination of the proper boiler size. Then the size of the radiation needs to be calculated so as to run at the desired return water temperature. I did this for myself (and oversized the baseboard in my upstairs zone to get low enough return water temperatures), but I had my contractor review what I had done, because although I read a lot, I had no practical experience.

EricAune

Tankless or not

There are "land mines" with this drawing, the others have touched on most.



Rinnai does not approve their tankless units for use in radiant applications, this will effect the warranty. Aside from the piping concerns, I would seriously consider either a condensing boiler or a tankless heater rated for space heating.

NRT_Rob

isotherm

is a watts radiant product. supply manifold is the bottom, return is the top, water stays "in circuit" unless the tempering valve opens.



appears to be a fixed temperature mixing valve. Strike Number one, you want reset control in most cases.



Strike 2, Rinnai is a very poor choice here for both warranty, and unnecessarily huge boiler pump wasting $15+/mo in electricity costs. If you have a load that needs high output... which I really doubt you do, with only one radiant manifold... then a mod/con boiler is the choice you want. Ebay the rinnai and get that set up properly with a mod/con.



if you have a tiny heat load, as I suspect you might if you only have the one heat load, a tank water heater with a small plate exchanger would be a much better idea.



to answer your question though, if the tubing you installed in the loop field is non-barrier, Strike 3. If not, good, and your primary header needs to be barrier pipe too.



I have a pretty dim opinion of the system you were sold, I have to say. I would say this looks like a single zone system, probably slab, right? maybe a shop or a basement? If so, an on demand can work, but it shouldn't have been primary/secondary... it should have been pumped direct with high flow through the zones. what you have here is a short-cycle nightmare that will almost definitely overcycle this unit and kill it before its time, even IF it were rated for heating use in the first place.

RickMac

Reply to JDB

Hi JDB



Thanks for the reply.



Here are the answers to the questions you asked:



1) OK, I'm happy to not put a ball valve between the tees



2) I might consider the Caleffi separator, but can't find a supplier or price - it looks expensive!



3) Tankless heater was professionally and correctly installed with pressure relief valve closest to boiler and ball valve behind, unlike drawing..!



4) Happy to leave out flow check valves after tees



5) I didn't calculate heat loss, I supplied info to radiant system supplier who sized system and provided drawing I posted this morning. Info on quote from supply company is as follows:



Total Square Feet…………. 2,100

Loops………………….… 8

Total Number of Zones.. 2

Approximate BTU's/Hr: 34,394

Pipe Feet……………….. 2,275

Pipe Centers…………….. 12

Heat Source……………… Tankless

Wall heights: 9' & 20'

Winter Low Temp: 12 deg F



6) There is a manual air vent at RH end of manifold



7) System was designed and sized for us. I was not aware of tankless versus boiler pros/cons. Can I connect a remote thermostat to tankless heater to monitor outside temp and automatically adjust Rinnai temp?



A so-called professional company designed and sized this system - I removed their identifying info from drawing to avoid embarrassing them.



Thanks



RickMac

RickMac

Reply to kcopp

Hi kcopp



Thanks for your reply.



Yes, 1/2" non-barrier tubing is already installed in concrete slab floors of basement and Great Room, the two areas the system was designed to serve.



Non-barrier tubing was used because that is what the radiant heating supply company specified and supplied. I did not know any different then...



There are 8 loops each approx 275' long.



What is an "open" system?



Design parameters were decided/calculated by supply company.



Thanks



RickMac

RickMac

Reply to HVHEHCCA

Hi Chris



Thanks for the reply.



I don't know the heat loss. All calcs were performed by radiant supply company.



3/4 would be fine by me, as going to 1" makes it more expensive by way of the fittings, and more expensive still going up to 1 1/4" as has been suggested.



I'm happy to re-site primary pump to pump through the tankless rather than sucking from it.



Where would you move expansion tank and feed to in the drawing I posted?



The secondary pump is part of a very expensive Watts IsoTherm Mixing Valve unit, see attached diagram of how it operates. It is the IsoTherm that produces flow in secondary to supply zones.



It uses the Law of Tees and a Mixing Valve to control temp to the zones. It must be installed using a P/S system.



Check valves are needed on supply and return to prevent ghostflow.



Because the IsoTherm mixes cold return with some hot supply before the cold return goes back to the boiler, it helps prevent thermal shock.



Valve on IsoTherm is a manual mixing valve that can be set to required temp. Boiler temp must be minimum 158F for IsoTherm to operate correctly.



Application is slab on grade in Great Room and basement



Thanks



RickMac



  

kcopp

An open system....

would be one where the heating water and the Domestic hot water are one and the same,  no heat exchanger....

RickMac

Reply to Eric Aune

Hi Eric



Thanks for the reply.



Unit did not have a warranty. It was bought from wholesaler friend at cost to him, and hence no warranty. Besides, unit has been installed for 2 years without being fired-up, and hence warranty would probably have expired by now had it had one.



Condensing boiler or tankless rated for space heating are not viable for me from cost point of view. If I could afford the alternative, I would, but Rinnai is installed and ready to use, and use it I must.



Thanks



RickMac 



 

RickMac

Reply to NRT_Rob

Hi Rob



Thanks for the reply.



You're on the money with the IsoTherm. It is a mixing valve and secondary circulator arrangement.



It is fixed temp (although temp can be manually adjusted). Application is slab-on-grade, and temperature variations outside should have little effect on two slabs, one 1000 sq ft, the other about 650 sq ft, shouldn't it?



The slabs were installed with heatsinks below to store residual heat even after system shuts down.



As I explained in my previous post, I have no warranty on Rinnai tankless anyway.



Swapping out the boiler really isn't affordable for me right now.



Field tubing is non-barrier, because that is what supply company specified and supplied, but all fittings will be brass or stainless steel, including primary pump. Manifold is stainless steel too, but I don't know if the Rinnai heat exchanger is.



Your guesses are pretty much on the nail. Application is slab-on-grade in both 650 sq ft Great Room and 1000 sq ft basement.



Can you possibly provide a simple diagram of the layout you're suggesting? It sounds as though a big primary pump feeding straight to the supply manifold and forcing the return back to the boiler is what you advocate, but isn't thermal shock an issue when doing that?



It would be great if this system could be as simple as you're suggesting.



Thanks



RickMac   

Jean-David Beyer

professionally and correctly installed

Remember, I am not a professional. That said, I have my doubts about your present installation.



1.) "Approximate BTU's/Hr: 34,394" If your heat loss is only about 35,000 BTU/hr, it is absolutely crazy to have a "boiler' capable of supplying almost 10 times that. Your efficiency would be just awful.



2._) "pressure relief valve closest to boiler and ball valve behind, unlike drawing..!" There must be no ball valve, or any other restriction anywhere between the "boiler" and the place where the relief valve drains. That drain pipe may not even be threaded to prevent a knucklehead from putitng a valve or cap on it, or hooking it up to a drain directly with no air gap. Makes me doubt the competance of whoever installed it.



3.) "I might consider the Caleffi separator, but can't find a supplier or price - it looks expensive!" It does not have to be a Caleffi separator. In fact, in my system, there is no dirt separator at all. My contractor did not wish to put one in for fear it might clog and that would destroy the heat exchanger in the boiler. Of course, he did a very thorough flush of the system before installing the new boiler. Perhaps other brands of hydraulic separators are cheaper. Also, you could get just a Caleffi dirt separator. It is just the Y-separator that concerns me. The Caleffi separator may seem expensive (I do not know the price, and we cannot give prices here anyway.), but when you consider that it is a separator, an air remover, and a dirt separator all in one, and involving less plumbing (labor), its overall cost may not be unreasonable.



4.) "Can I connect a remote thermostat to tankless heater to monitor outside temp and automatically adjust Rinnai temp?" I have no idea. It has to be a continuous (thermistor) type sensor, not a typical room thermostat (on-off) to be of use here, and I have no Idea if the Rinnai can accomodate this. My only experience with on-demand hot water heaters is from one I had to use in France in 1950, and it had no such possibility.



5.) I hope you are not getting domestic hot water out of this system too. One of Murphy's Laws goes, "When you open a can of worms, to recan them takes a larger size can." This seems like such a situation to me. I would never want a system anything like this in my house.

RickMac

Hot Water Out Connection

Hi JDB



Again, thanks for your reply.



Perhaps I wasn't clear about how the hot water out connection is configured, so I have attached a pic of pressure relief valve arrangement.



No, I am not drawing DHW from this tankless. I have two identical Rinnai tankless heaters, one for DHW, one for the radiant system.



Thanks



Rick

RickMac

Not Open System

Hi again



I have two Rinnai tankless heaters, one for radiant, one for DHW.



Thanks



Rick

Jean-David Beyer

Code?

I do not think plastic pipe is allowed at the drain of a pressure relief valve.

CMadatMe

I'm Sorry

But this is a horrible design and you have been given some bad advice. The heatloss is probably only in the 25K range. Rinnia also recoemmends that this system run at 20psi min but really want to see 30PSI. Couple of more questions.



Is the slab insulated to include perimeter? Slab applications don't require a high water temp. Like Eric said in his post. Get rid of the pri/sec and even that "expensive" manifold set up and just direct pipe it. You only need one water temp and the tankless can do that.



What water temp does the design say you need? Remember you only need that water temp on your design day not when it's 20 or 30 degrees out.



The non use of barrier tubing is going to hurt you big time. Your going to eat up those circulators to oxygen diffusionm so start putting a few bucks away here and there and get ready to replace them. You must use bronze/stainless circs.



I'm still questioning how that manifold pump is going to move water through the radiant based on the way the diagram is. Seems to me you need an injection pump to get water out of that primary loop. You could then do outdoor reset for the radiant.

RickMac

Radiant is Secondary Heat Source

Hi again



Perhaps I should explain that my 2-storey home is heated by 2 heat pumps with backup heaters, one for the 1st floor and one for the second floor.



The radiant system I am installing is to make the concrete-floored cathedral-ceiling Great Room a little easier to keep warm and more comfortable, and to warm the basement floor which is also bare concrete.



I am not relying on the radiant system for primary heat.



Thanks



Rick

RickMac

Slab Insulated

Hi again



There are two slabs, both insulated with an open center to create a heatsink.



Design does not mention water temp...



I intended using a stainless circulator.



I'll install it per advice here, direct piped to manifold and return.



Thanks



Rick

CMadatMe

Oh Boy

Residentially not a good idea. You have to heat mother earth before the slab will even begin to heat. Your downward heat loss increases substaintially. Going to have some catch up issues "ala" chasing the night in the swing months. One way to control some of this is the use of slab sensors. This would allow you to maintain some type of minimum slab temp and keep mother earth from sucking all the btu's out of the slab when the zone is satified especially during the day when solar gain influences room set point.



Be prepared for a very uncomfortable system. Huge swings in temp.

kcopp

I have to ask....

what was used for insulation? How thick? kpc

RickMac

Slabs Have Stats Installed

Slabs have stats installed, one in each, to monitor temp of slab and shut off system when slab is at temp.

RickMac

Insulation is 2" Thick Polystyrene Slabs

Insulation is 2" thick polystyrene slabs with 2" thick perimeter insulation between slab and walls. 

NRT_Rob

simple

I'd ditch the isotherm, and run my loops wide open... but you might still need a bypass to avoid massive velocity issues if you have a zone that is only 2 or 3 loops, maybe.



If you own all this stuff I would simply turn up the mixing valve as high as it goes so it stays open, and set the rinnai to the temperature I want or so that the average temp is what I want. reset may or may not be a big deal to you here depending on what you need for a supply temp, how well insulated your spaces are, etc. I still think the rinnai will die an early death.



the 26-64 is not a bronze or stainless pump though. that will need to be changed for stainless or bronze pump if this is all non barrier.



the rinnai does not worry about thermal shock as it is built for domestic hot water.

Ironman

One More Option...

Rob and Chris are dead right about this thing short cycling itself to death (not to mention the other issues pointed out).



Pumping directly through that high head tankless, as pointed out, will have issues also.



You may want to consider putting a buffer tank between the Rinnai and the radiant. That would cut down on the cycling some and give hydraulic separation between them. Of course, then you'd need two non-ferrous pumps and another component has been added.



No really good solution to a poor design.

Jean-David Beyer

Did you have a geologist study this first?

"The slabs were installed with heatsinks below to store residual heat even after system shuts down."



Did you have a geologist study this first? If not, you may have had money sinks installed below by mistake. If the water table is fairly close to the bottom of your slab, your so-called heat sink will have the drain unplugged and all the heat going down there will go to heating the underground water that will then be carried away. In other words, why even bother to insulate the edges of your slab if you are letting the rest of the heat disappear out the bottom, never to be returned?



If the water table is very very deep, and not moving, during the heating season, this just might work. In my experience as a non-professional, I do not even know if my slab is insulated or not. But it is a huge thermal mass with a time-constant of over four hours. That is all the storage I can stand. It makes setbacks impractical. I have to start the night setback around noon to cool things down a few degrees by bedtime. And I have to start the recovery around 9 PM to get the temperature back up by 6 or 7 AM. Even that does not work out all that well because the changing outdoor temperatures confuse the outdoor reset as the hot water into the slab all night is too hot for the house in the daytime.



In a feedback control system, there are two parameters that are critically important: the gain of the system and the delay. If the product of the gain and delay is too high, the system can be unstable, resulting in wild swings of the controlled variable (the temperature of the house, in this case). Now the delay of the slab is enormous, so you have to keep the gain of the system very low. The gain of this system is related to the size of the boiler, the slope of reset curves, the amount of setback attempted, and various other things. The trouble with that is that the lower the gain of the system, the less accurately the controlled variable (the house temperature) can be maintained. If I were designing a house with radiant heating, I think I would not heat the slab at all, or just put a little heat into it to remove extreme chill, and heat with radiant wall or ceiling with much less thermal mass.



I have abandoned setbacks in my radiant slab zone. I have diddled the reset curve of that zone to deliver as cool a hot water as I can that barely overcomes the heat loss. Consequently the circulators run almost all the time when it is cold out.. The boiler adjusts its firing rate to keep the temperature of the water to ensure this. Since it is 50F outside right now, it cannot get down that low, so the boiler has not run at all so far today and only 4 hours yesterday. When it is cold out, it runs about 18 hours a day.

RickMac

I am a Geologist

Hi again



I am a geologist.



Perhaps what I have not made clear is that this is a secondary heating system only. Primary heating system in my home is heat pump with backup heater. This radiant system is only meant to slightly warm the basement floor and the Great Room floor.



The Great Room has a 20' high ceiling and it is difficult to maintain a comfortable temperature when all the fan-forced warm air wants to find its way upwards.



The in-floor heating is only intended to provide some warmth at low level, which radiant can do much more efficiently than a heat pump with backup.



Thanks



Rick 

NRT_Rob

what exactly

are "heat sinks" in this situation, and why did you think they would be a good idea?

RickMac

What Heatsinks Are

A heatsink is an uninsulated area of ground at the center of the slab which absorbs heat when the slab is heating and releases that heat slowly when the system is not calling for hot water.



It irons out any minor fluctuations in temperature and thus reduces the 'on' cycles of the system.



    

NRT_Rob

ah

that was a poor choice of design. the ground is not a good heat storage mechanism. Water is very good. Dirt, not so much.

Jean-David Beyer

reduces the 'on' cycles of the system.

If you say so. With my slab, it already held so much heat that it really made controlling the system quite a problem. When the thermostat is satisfied, the slab continues to heat the house for 4 or more hours. At the other end, you might say it stores cold, and when the thermostat calls for heat, it takes 4 or more hours to get appreciable heat. And this is not due to an undersized boiler. My former boiler burned 70,000 BTU/hr, and I calculated the heat loss of my house to be somewhere in the neighborhood of 35,000 BTU/hr when it is 0F outside. And the design day around here is 14F. So when the thermostat called for heat, the boiler would go on  for something like 45 seconds, off  for about 75 seconds, rinse and repeat for hours. It was a rather noisy oil burner and I would count the seconds while getting to sleep. The burner is in the garage that is attached, but with no connecting door. I could also hear the 3-piece circulator go on and off, but that was not very often. When the circulator threatened to quit, they replaced it with a Taco 007, which was good enough and I did not hear that from inside the house.



Now if I increased the heat capacity of the slab by some means, it would just make the control of the system more difficult. If I raised its time constant to 12 hours, I would be putting the heat required for the day into the slab all day, but this would not be enough for the nite time when it was starting to affect the house temperature. And at night, it would be putting the heat required for the night, but this would be too much for the daytime when it would take effect. Now maybe if I raised the time constant to 24 hours, it would be easier to control, but I would not wish to risk it.



As I said in an earlier post, I no longer try any setbacks for the slab zone. And I use outdoor reset so as to put the minimum temperature I can get away with into the slab. This works better than the bang-bang control system with the old boiler. Instead of getting temperature swings from about 67 degrees to 75 degrees, the swings are reduced to 68 to 70 degrees most of the time, and it goes up to 71 if the sun load is especially high. The outdoor reset allows for making only small changes in the water temperature delivered to the slab instead of 130 or so water being turned on and off at long intervals. The circulator for that zone runs most of the time in very cold weather. It would do that in warmer weather except my boiler is still a bit too large, and it will not modulate down far enough when it is 55 degrees or more outside.



From the point of view in reducing rapid cycling, instead of increasing the thermal mass of the system, which makes for instability problems in the control system, it seems to me that making sure the boiler is not oversized is the #1 priority, and outdoor reset is #2. But I remind you that I am not a heating professional.



P.S.: It might make sense to increase the thermal mass of my upstairs zone that is baseboard. That zone requires only 6,500 BTU/hr when it is 0F outdoors. And the boiler will modulate down to a minimum of 16,000 BTU/hr. So it is about 3x oversized if only that zone calls for heat. I have set the controller to cut the firing rate in half, widened the hysteresis between turn-on and turn-off, and now the cycling rate is better for just that zone than it was with the old boiler trying to heat the entire house. I can set the firing rates, reset curves, and hysteresis for each zone separately.  Around 8 minutes on, 10 minutes off on a warmish day. It would probably be longer cycles on a cold day. I think this is good enough, but if it were not, putting a storage tank in series with the baseboard (probably on the return side) seems like the way to go.  But I do not think I will do it.

Gordy

Heat Sinks, Thermal Mass, and Fly Wheel Effect

 HeatSinks are great when the energy used to heat them is free, or cheap. In your situation  you are using a paid fuel source to heat the structure, you are also paying to heat the heat sink.



  Rob is spot on when he said that dirt makes for a poor storage medium. It takes a lot of energy to charge it, and you are getting that energy back when you do not need it, and its not contained as far as where the heat is limited to travel or when you want it to travel.....back in to your house, and China.  Water is best, and more versital, and easier to control.



  This is likley to create another problem called fly wheel effect. Basically as previously stated you  are getting more energy back in the structure when you do not need it. Your concrete slab adds to the fly wheel effect due to its thermal mass.



 This is why the best thing you can do is insulate the whole slab. You will use less energy to bring it up to temp, and it will be quicker to respond which yields a lower energy bill. Plus the control strategy ends up being much simpler because you are not trying to prevent things from happening you are only making them happen how you want it, and when you want it.



Gordy 

RickMac

Pump Selection

Hi again



Would a Taco 009-SF5 stainless steel pump be a good choice? I need high-head low-flow, and this circulator will give me the required 15' head at a flow rate of 5 gpm.



Longest loop is 250' at 6' head-loss per 100', i.e 15' head required. Flow rate from Rinnai is 5gpm max.



Thanks



Rick

Jean-David Beyer

Have you considered reading this book?

There is a really good book on hydronic heating by John Seigenthaler available at this site:



http://www.heatinghelp.com/products/Books/5/96/Modern-Hydronic-Heating-Second-Edition-br-by-John-Siegenthaler



It will answer lots of your questions. It even includes a CD-ROM with the student version of some very useful design software. Requires Windows.