Maximum temp for suspended hydronic PEX floor heat

[Plumber69]

But the manifold is feed by half inch PEX.

 

[StephK]

Yes, 3 circuits. The furnace sits about 30 feet from the manifold/floor heat piping. That is 1.5" copper. Then it converts to 3/4 pex. Then from the manifold is 1/2 for the 3 circuits. As you see in the pic.

Dana - Sorry, that 170/145 is the temps just at the floor heat. So 145 is immediately upon returning from the floor circuit. I don't know what the water temp is by the time it gets back to the Munchkin. But something less than 145. I understood that the floor input/output differential should be about 20 degreese, so seems like we are right on.

Our Munchkin is heating the rest of the house radiators and I believe it's set to 180 fixed temp. I don't think we have an 'outdoor reset'. My thought for setting this floor heat system up, was that since we have this boiler heating up all this water, why not use it for the floor heat. And since the floors only require 125 (supposedly), then the floors shouldn't make the furnace burn much more than it already is. But, whenever the floor circulator runs, the furnace fires. I'm not sure if that's b/c the floor is taking all the hot water so it validly needs to heat up? Or b/c it's wired to burn when the circulator calls it. But since the floors circulate all the time, the furnace runs all the time.

As far as heat load. The exterior walls are made of 3 courses of brick. About 660 square feet of exterior walls. (60 linear feet x 11 foot ceilings). And 5 big windows for a total 75 square feet. Right now the windows are plane old single pane glass. We'll eventually be getting good storms, but for now they are 1910 wood sash, single pane. And i'm sure they leak alot of air.

BUT, all that being said .... shouldn't the floors still feel warm??

If I was sure the issue was the 6" gap b/w the insulation and the metal shields then yes, we could blow cellulose in. But that would be a heck of a lot of holes. I'm guessing 15 joists pockets, but there's a bunch of recessed lights and I think we'd have to blow in between each one of them.

 

[Sponsor]

 

[Jadnashua]

I'd take the time to try to tighten up the windows. Then, I'd look for some low-e film to install on the insides...it's fairly cheap and if done right, almost invisible. It comes in various light-transmission densities. It doesn't need to be dark. Before I ended up changing out my doors and windows, I put some film on my slider that faced west and got the direct summer sun. It made a big difference in the summer, and sitting by it in the winter was warmer, too as less heat was transmitted out and was bounced back into the room. NOt as efficient as having the film(s) between the glass and it can be scratched, but it can help.

Put a folded blanket on the floor with a thermometer under it. Leave it for say at least 15-minutes and see how warm the floor is. That will give you an idea of if there's much flow and actual heat transfer. If you have or can borrow an IR camera, it might be illuminating to use it to view the floor. YOu can use it also to check the walls and locate insulation deficiencies and leaks.

 

[StephK]

Thanks for the tip on the windows. Like i said, good quality (and historically acceptable, since the n'hood requires it) storms are in the plan for the not too far future. We've used them before on another house and they do make a huge difference.

I did the blanket thing on the floor. It got to 80-85 degrees.

 

[Dana]

But the manifold is feed by half inch pex

Looks like 3/4" between the pump & manifold to me:

 

[Dana]

Yes, 3 circuits. The furnace sits about 30 feet from the manifold/floor heat piping. That is 1.5" copper. Then it converts to 3/4 pex. Then from the manifold is 1/2 for the 3 circuits. As you see in the pic.

Dana - Sorry, that 170/145 is the temps just at the floor heat. So 145 is immediately upon returning from the floor circuit. I don't know what the water temp is by the time it gets back to the Munchkin. But something less than 145. I understood that the floor input/output differential should be about 20 degreese, so seems like we are right on.

Our Munchkin is heating the rest of the house radiators and I believe it's set to 180 fixed temp. I don't think we have an 'outdoor reset'. My thought for setting this floor heat system up, was that since we have this boiler heating up all this water, why not use it for the floor heat. And since the floors only require 125 (supposedly), then the floors shouldn't make the furnace burn much more than it already is. But, whenever the floor circulator runs, the furnace fires. I'm not sure if that's b/c the floor is taking all the hot water so it validly needs to heat up? Or b/c it's wired to burn when the circulator calls it. But since the floors circulate all the time, the furnace runs all the time.

As far as heat load. The exterior walls are made of 3 courses of brick. About 660 square feet of exterior walls. (60 linear feet x 11 foot ceilings). And 5 big windows for a total 75 square feet. Right now the windows are plane old single pane glass. We'll eventually be getting good storms, but for now they are 1910 wood sash, single pane. And i'm sure they leak alot of air.

BUT, all that being said .... shouldn't the floors still feel warm??

If I was sure the issue was the 6" gap b/w the insulation and the metal shields then yes, we could blow cellulose in. But that would be a heck of a lot of holes. I'm guessing 15 joists pockets, but there's a bunch of recessed lights and I think we'd have to blow in between each one of them.

Even though they are more expensive up front, a Low-E storm window will "pay back" more than twice as fast as a clear glass storm, and will lower the cooling load as well.

A wood-sash single pane has a U-factor of about 1.0 BTU/hr per square foot per degree F temperature difference. So in a 65F room it's losing about 35 BTU/hr per square foot of window when it's 30F outside. Code min indoor temp for heating systems is 68F at the 99% outside design temp , which in Richmond is +18F a 50F delta. So at code minimum that window is losing 50 BTU/hr per square foot of window. A tight clear glass storm would cut that to about 25 BTU/hr per square foot, but a tight L0w-E storm reduces that to about 17 BTU/hr per square foot.

With 75 square feet of window that's 3750 BTU/hr of load in that zone/room at code min design condition for just the windows, and that would be reduced to 1275 BTU/hr of window load (not counting the infiltration savings) with a low-E storm.

A 10" thick common brick wall has a U-factor of about 0.4 BTU/hr per square foot per degree F, if plaster directly on brick, about U 0.3 BTU/hr if it's furred out with lath & plaster with a 0.25-1" air space between lath & brick. With 660' of wall, less 75' of window you're looking at 585 square feet. So at code min design condition your wall losses could be as high as...

U0.4 x 585' x 50F= 11, 700 BTU/hr

...or as low as...

U0.3 x 585' x 50F= 8775 BTU/hr (and probably not a lot lower.)

If it's foot-thick brick it'll be between U0.25- U0.3 depending on how the interior plaster is set up.

So, thinking optimistically on the wall you're currently looking at a total load of 8775 + 3750= 12,575 BTU/hr + infiltration. Assuming it's pretty leaky but not an open barn, let's call it 10cfm per window, or 50 cfm, or (x60=) 300 cubic feet per hour. At a 50F delta that would be about

0.018 BTU/cubic foot per degree-F x 300 cf/hr x 50F= 270 BTU/hr.

If it's really drafty it could be 500 BTU/hr but probably not more than 1000 BTU/hr when all leakage is added up unless the floors and ceilings also leak air like a tennis racquet.

So call it ~13,500 BTU/hr, assuming it's conditioned space above, and an unventilated basement below at a 50F delta, or (13,500/50F=) ~270 BTU/hr per degree of indoor to outdoor temperature difference.

It looks like you have about 16'x 25' or 400' of foundation footprint (?), less about 70' of stairwell, for a net of 330 square feet. After partition walls and cabinets etc you're looking at only 300 square feet of active radiant floor exposed to the rooms, best-case.

13,500/300'= 45 BTU/hr per square foot of radiant floor to meet design condition. That would take a floor temp of about 95F, which is a bit tough to do even with a plate-up and 180F water even when using extruded rather than sheet metal plates. See the nomograph on p 4. With 7/8" subfloor + 5/8" CDX alone you're already at about R2. With 170F water better class extrusions 8" o.c. will deliver about 25 BTU/hr per square foot, which is only half what you need. Best-in-class sheet metal heat spreaders are somewhat lower performance than that, call it 20 BTU/hr per square foot with 170F water, so with 300 square feet of active radiant floor it's only able to deliver 6000 BTU/hr. It'll deliver more than that if you boost the temp to 180F, but not a lot more.

In a zone with a load that increases by 270 BTU/hr for every degree of indoor to outdoor temperature difference it starts losing ground whenever the outdoors is (6000/270=) 22F colder than the indoors, which would explain why it's unable to maintain the 63F indoor temp when it's dropping below 40F overnight, but probably does keep up when it's 45F and above.

If you peel 2500-3000 BTU/hr of load off by installing l0w-E storms your load will be about 10-11K. Let's call it 10,500 BTU/hr, at a 50F delta-T (21o BTU/hr per degree F) or 10, 500/300'= 35 BTU/hr per square foot of floor. With a floor maxed out at ~20 BTU/hr per square foot of floor it still won't keep up at design temp, but at 210BTU/hr per degree the 6000 BTU/hr it can deliver would still support a 6000/210= ~29F difference, maintaining the 63F indoor temp all the way down to 35 or a bit lower before it begins to lose ground.

Any chance of insulating the exterior walls? An inch of foil faced polyiso would cut your wall losses by more than half and with low-E storms the floor would likely be able to keep up even at design condition. With outdoor reset it would also be able to run in condensing mode at least half the time, as opposed to 0% of the time the way you're running it now, delivering better than 90% seasonal efficiency compared to about 85-86% running at a fixed 180F out with a 140F+ return.

 

[Plumber69]

still 3/4 would feed 2 zones not 3

 

[StephK]

Wow, thank you Dana for all the time you took! Above and beyond! I need to re-read this information a bunch of times before it sinks in. But, I think I summarize that the floors do not feel warm nor heat the room sufficiently due to heat loss elsewhere. IE. windows, unheated downstairs, open stairwell (we do have it very well closed off with several layers of blankets.... but still), wall R factor (which we can't change). We had a company do the design for us, so it's frustrating that their design isn't work as hoped. (Although I do think we told them to design it assuming we had the storms, b/c we have always planned on that).

We can't insulate the walls. They are plaster on brick, in good shape. We will get the Low-E storms when we are ready for that expense. We have discussed blowing more installation into the attic so perhaps that will help too. And the furnace outdoor reset. I'll look into that. I know the Munchkin is starting to have issues so one of these days we will have to cut the cord and replace it. Just not too soon I hope.

One additional option is to add back a radiator. We very specifically did not want to do this, but we left the pipes in the basement just in case we needed to.

Thanks to everyone for all your comments. Much appreciated. My final thought is this: We were planning on doing the same format of pex pipes for the first floor kitchen floors (below this area). We have already completed the copper/manifold/pump layout. But NO WAY are we running it under the subfloor!! I will have to do the research to see what would change if we ran the pipes on top the subfloor, under the tile. Back to the drawing board on that.

 

[Dana]

Note: There was a order of magnitude error on the infitration WAG: 50 cfm is 3000 cf/hr, not 300, so that would be 2700 BTU/hr (not 270 BTU/hr) at that leakage rate. (Some days I skip my coffee breaks, which probably isn't a good idea! )

Before replacing the Munchkin, take the time to run a more formal heat load analysis (or a fuel use heat loss calculation) to size it correctly for the load. Using the methodology of my above example would get you close, but analyzing the U-factors of the walls more carefully would still be in order. Even though you have fair amount of thermal mass to work with in the water volume of your radiation, oversizing leads to more cycling, more wear & tear on the boiler, lower efficiency. There are several fire-tube boilers out there with 10:1 turn-down ratios, but even with those you can't hang a 200KBTU boiler on a 30K design load and expect optimal results.

 

[Dana]

One additional option is to add back a radiator. We very specifically did not want to do this, but we left the pipes in the basement just in case we needed to.

Since you have an approximation of the design load and how much of it the floor would support it gives you a good idea of just about how much radiator it would take to cover the difference. If you oversize the radiator the room would hit the target temps, but the floors would be colder.

Yet another option to consider is running the radiant off a floor thermostat, but maintaining the room temperature with a modulating ductless mini-split heat pump. There are several mini-splits that would cover the full heating load on their own, which isn't necessarily what you want. A mini-split with a wide modulation range and not ridiculously oversized for the cooling load would be a better choice.

My go-to choice would be the Mitsubishi FH09NA, which is capable of delivering over 12,000 BTU/hr at your 99% outside design temperature, yet can throttle down to as low as 1600 BTU/hr @ +47F.

Another would be the LG LSU/LAN090HSV5, which puts out about 11,000 BTU/hr @ +17F, but can modulate down to 1023 BTU/hr @ +47F.

There are cheaper 3/4 ton mini-splits than these, but the key to getting the best comfort and efficiency out of them is being able to add just a tiny bit of heat when it's dropping into the 30s and lower rather than cycling on/off a lot.

Between a decent 3/4 ton cool-climate minisplit and the radiant floor you should be able to bump the room temp up to 70F when it's 18F outside and still keep up.

With either mini split, in your house it will be worth getting a wired or Wi-Fi wall-remote/ thermostat, or it won't track the temperatures very well at all. In their native control mode mini-splits sense the indoor temperature by the intake air temperature at the head, and mounted high on a low-R wall the air entering the head won't be anywhere near the average room temperature. For a couple hundred USD (overpriced for what it is, but in your case "worth it"), setting it up to use the temperature sensed at the wall-remote/thermostat allows you to put the sensor in a more reasonable location, such as 5' off the floor on a partition wall where it won't get direct sun.

Reliability & efficiency-wise the Mitsubishis are pretty good, with the reputation to match. The LGs have a somewhat spottier reputation, mostly due to manufacturing defect issues (and sometimes installer training), but once they're fully commissioned and running they're pretty good too. Local support is important for either of them.

Fujitsu & Daikin are first-tier vendors, but their 3/4 tonners don't modulate as low as you would like in this application. Fujitsu 3/4 tonners are more like 1-ton cold climate units, and only modulate down to ~3000 BTU/hr. Daikin 3/4 tonners don't use vapor injection scroll compressors, and as a result have quite a bit lower capacity & efficiency at +17F , yet have an even higher minimum output at +47F.

 

[StephK]

Well, funny you should mention it, but we did put in a Trane mini split pancake unit for this same area of the house. It hangs in the linen closet with 3 attic ducts, one to each room. I’d have to look up the model # and BTU, but we mainly put it in for AC in the summer.

We do have a wall thermostat (wanted wifi, but apparently Trane doesn’t make one for this unit yet, they say next year). We haven’t quite figured out when/why this thing runs. Sometimes the air seems to be blowing all the time, and then it’ll go hours without coming on. We’ve never had a mini split before, so it’s new to us.

We also don’t know how we’d run both at once. If the Trane controlled the room temp, would we set the floor heat higher? But then it would come on all the time too. Right now the stats are right next to each other, so seems they’d be constantly battling it out.

I had one more question, sort of hypothetical, but not really. So, considering all other things equal with your heat load calculations. Would it change substanctially if we had laid the pet pipes on TOP of the subfloors? Directly under tile? I ask b/c downstairs in the kitchen, it’s the same layout, same windows and everything. We are now wondering if we can lay the pex on top and if that would change the performance substantially.

 

[Jadnashua]

If you can stand the floor buildup, look into https://www.schluter.com/schluter-us/en_US/Modular-Screed-Systems/Schluter®-BEKOTEC-F/p/BEKOTEC_F . Putting the heat closer to the surface will speed up response, and embedding it in the mortar gives things more thermal mass and helps spread the heat well, too. This has been used in Europe for ages, so it's not new.

 

[Dana]

Well, funny you should mention it, but we did put in a Trane mini split pancake unit for this same area of the house. It hangs in the linen closet with 3 attic ducts, one to each room. I’d have to look up the model # and BTU, but we mainly put it in for AC in the summer.

We do have a wall thermostat (wanted wifi, but apparently Trane doesn’t make one for this unit yet, they say next year). We haven’t quite figured out when/why this thing runs. Sometimes the air seems to be blowing all the time, and then it’ll go hours without coming on. We’ve never had a mini split before, so it’s new to us.

We also don’t know how we’d run both at once. If the Trane controlled the room temp, would we set the floor heat higher? But then it would come on all the time too. Right now the stats are right next to each other, so seems they’d be constantly battling it out.

I had one more question, sort of hypothetical, but not really. So, considering all other things equal with your heat load calculations. Would it change substanctially if we had laid the pet pipes on TOP of the subfloors? Directly under tile? I ask b/c downstairs in the kitchen, it’s the same layout, same windows and everything. We are now wondering if we can lay the pex on top and if that would change the performance substantially.

Look up the model number- the number plate on the outdoor unit may be more accessible. Even the smallest & least efficient Trane minisplits , the 4TXK6509A10N0BA is good for 9000 BTU/hr @ +17F outdoors, 70F indoors (the AHRI test temperature for heat pumps.) If they say it's rated for 9K it has to be able deliver at LEAST that much at +17F in/70F out to rate it there, and it will have slightly more capacity at 65F indoors. Trane has put their label on a few different Asian mini-split vendors- not sure which series was build by whom. The more recent "M-Series" Tranes are Mitsubishi (and even say Mitsubishi on the outdoor unit) , which would be pretty good, but I suspect you'd know if it were one of those.

Because they modulate the output the SHOULD run extremely long cycles- the ideal right-sized mini-split would have nearly continuous operation whenever there is a heating or cooling load. With your low-R walls and high-wall location the setpoint it's operating to won't look very similar to the temperature you punched into the remote, but that's not a big deal- just bump the temp up when you're too cold, bump it down when you're too hot, and just forget about the absolute meaning of the number. The high thermal mass of the plaster-on-brick will also keep the active cycles pretty long, even if the thing is oversized.

Odds are pretty good it'll keep up, and the floor will still give you that barefoot-cush if you turn on the mini-split any time you expect it to drop below 45F outdoors. Try setting the mini-split to 70F- it'll probably run full speed most of the time. You may have to back off to find the sweet spot. As the room temp rises, so will the floor temperature, as long as the radiant's thermostat is still calling for heat.

Trane doesn't publish a lot of info on their mini-splits. We don't really know what the turn-down ratios are for the non-Mitsubishi models, or what the extended temperature capacity tables look like. Do you have the full printed user manual at least? (The online third party bootleg versions of the manuals for that series on the web look too much like virus contagion for me to feel comfortable downloading anything.)

jadhashua tells it true about going above the subfloor. There are several above the subfloor systems that really can deliver 40-50 BTU/hr per square foot through tile or even engineered wood flooring.

 

[Jadnashua]

Most electric, in-floor systems are designed for floor warming, not space heating, although, depending on the heat load, MIGHT be sufficient to heat the area. You generally can't supply more than 12-15W/sqft or maybe 240w/m^2 or about 800BTU/hr. If you try to get more than that, the floor can become uncomfortably hot.

What's common in the UK, can't say, most of us here are from the USA.