Rigid styrofoam under radiant slabs
Streamline
| Posted in Energy, Heating & Insulation on
I am preparing for a hydronic radiant floor installed in 4″ concrete slab. I want to insulate the slab properly. I am thinking that the right way to do this is to provide a 4″ capillary break with gravel, then visqueen for moisture control, then 1 1/2″ rigid styrofoam, then Pex tube. Concrete would be poured 4″ over the styrofoam. Questions:
1. Is the process correct?
2. How do I attach the 4X8 sheets of styrofoams together to close off the air gap? Duct tape fine?
3. Should I insulate around the slab perimeter edges also by lining up 4″ high styrofoam strips (because the slabs are 4″ high) against the foundation board so there is a clear thermal break between the slab and the foundation wall?
Thanks.
Replies
1. Is the process correct?
Yes - tie your tubes to your wire mesh (mesh first) with zip-ties
2. How do I attach the 4X8 sheets of styrofoams together to close off the air gap? Duct tape fine?
Just push them together - there will be no gaps that have enough effect on heat transfer. Duct tape if you like, or use thinner foam sheets staggered.
3. Should I insulate around the slab perimeter edges also by lining up 4" high styrofoam strips (because the slabs are 4" high) against the foundation board so there is a clear thermal break between the slab and the foundation wall?
I am assuming you are pouring inside a wall, not on top of one. I would indeed insulate. Some folks worry about termites when you do this.
Most people expect concrete to be terrible as a finished floor - not so when it is warm & dry - add a few throw rugs & you're golden.
and...make sure you have a perfectly flat surface under the styrofoam so it will fit together nicely.....and don't forget to pull the wire mesh up so the tubes are about half way into the slab.
Keep air pressure in the tubes while you are pouring so you know if you punch a hole in one of them
Don't let anyone stand a wheelbarrow on end when they are dumping it
Make sure your PEX runs are not over 250' to 300'
Make a sketch of where the individual PEX runs are located in relation to your floor
Glue the side pieces of foam at exactly the right elevation before you pour, so you can use the top of the foam to strike off the concrete
Good luck! You'll love your heated floor!
I used a pex stapler to attach the tubing to the foam and it is much faster then tying to wire. Skip the wire mesh and use fiber in the concrete.
Tom
Douglasville, GA
I'm with Tom. Unless It's a small area, foam staples are GREAT. soo fast..Generally I'd go with a full 2" of foam under the slab, but it depends a bit. If it's a very large slab especially 1.5" wouldn't make me throw fit.If you do use mesh, be sure to lift it, but the pipe itself doesn't really need to be lifted; doesn't do a lot for you to lift it, really. Doesn't hurt either, other than raising your risk of puncture or cutting it a bit.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
Wouldn't that put the tubing nearly 4" away from the living space? It seems the response to a call for heat would be awfully slow. I have radiant in 1 1/2" mortar on top of plywwod and tile over that. The swings are pretty high going as high as 4 or 5 degrees over the thermostat by the time the thin slab is warmed, wouldn't it be much worse burying that tubing so deeply in the slab? Wouldn't response time be much slower? Wouldn't it be better to have the tubing as close to the living space as possible without compromising the slab?Lee
Technically, yes, it will slow down response time a little. But not so much.. ultimately, you need to heat the whole mass up for effective heating, whether the tubing is at the bottom or not. If you have direct conduction to a cooler mass, that's going to draw most of your heat until it's equalized, generally speaking.If you are seeing temperature swings with your system, then it was not designed or controlled properly. Proper thermostats and/or water temperature control, and proper operation (constant, no daily setback), sometimes even floor sensors... these are the ways to address mass lead/lag. tubing position isn't going to create or solve a problem in that respect.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
We don't use it constantly, no matter how you slice it wood is cheaper than LP so we only run the floors when we want to run around nekkid or when it gets reallly cold.Lee
Gotcha. If it's truly periodic, that's fine. But if you're just, say, turning it on every evening, you are not helping your case.. or if you are, it's not by much at all.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
It still seems to me that heat would enter the room sooner if the tubing was only an inch or two from the room rather than 3 1/2" separated by concrete. It also seems to me that less heat would be lost through the bottom if the hottest part of the system was farther away from the ground. I suspect that I'm right and stapling the tubing to the foam is more for the benefit of the installer than for the benefit of the system.I have a programmable thermostat on the system and it keeps track of times the system run. It was a wash between setback and steady so I agree with you that setting the thermostat to a fixed temp is the way to go with these systems.Lee
Some heat will. There is some benefit to raising the pipe.. it's just not much, and not worth worrying about. Presuming adequate insulation is in place, and the system is controlled properly, it's really a non issue. If neither of those presumptions are true, raising the pipe won't help enough to rectify a problem.As long as part of the concrete is not up to temperature, it's going to draw heat to itself, and since it's in very good direct conduction with the rest of the mass, it'll be a significant draw until the mass is up to temperature. You may get some heat faster otherwise, but you won't maintain a comfortable room temperature until you heat the mass. Unless you are using inappropriately high water temperatures, perhaps. But then you're being inefficient in a much more serious manner.So again, technically speaking you are right. In practice though, it's not worth worrying about.It sounds like you are using a bang-bang strategy if you are watching run times. I suspected that, and that is most likely the cause of any inaccuries you are getting in temperature control. If you ever end up using this system full-time, I suggest using a variable temperature system based in indoor or outdoor feedback.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
Rob, I see you're in Gardiner--I'm a foreman for Fine Lines Construction in Yarmouth. We do a lot of radiant-heated floors but it's usually just our plumber laying it out. I'd like to see it done more scientifically. If you have any marketing materials could you send them to FLC, PO Box 1110, Yarmouth ME 04096?
Thanks--
Mike Maines
I installed this system in our new house about 8 years ago. I had the technical help of a guy that has been doing these systems for years, both here in the USA and in Scotland. He gave me loop lengths, all the parts, instructions and I assembled it. He then inspected it before fill. It was a win/win situation for both of us, he made a few bucks on parts and I saved a ton on labor. It's the best form of heat I've had. There is a For Sale sign in our front yard and if I build for us again I'll be looking into the sensors you're talking about.Lee
You are correct.But the concrete is a far thermo conductor (ie, poor insulator) and it has a large thermo mass that needs the energy to raise the slab temp anyway.So we are talking about difference in seconds, not hours.
The process sounds fine. Make sure to use XPS, extruded polystyrene foam, instead of EPS, expanded polystyrene foam.
The XPS has much higher compressive strength and absorbs less water than the EPS.
Stu
Guys, what I learned from this conversation is I should do the following (some are subject to discussion as I can see above): Use 2" XPS rigid foam, glue the perimeter boards against the foundation wall for thermal isolation, keep the tops level, raise the tubes to be center of 4" slab, push sheets together is fine with no taping required, keep air pressure on tube during pour, keep runs 250' - 300' (not more, not less). I will have engineered wood finish on top of the slab. Hope I got it all right. Thanks a bunch everyone!
Don't be alarmed if the tubing takes a while to get up to pressure - we had 3500' of tubing and thought we had a leak, but after 10 minutes or so, the tubes filled and equalized.
The inspector required a 100lb pressure test.
Treat every person you meet like you will know them the rest of your life - you just might!
My experience was that the tubing took much more than 10 minutes to show a stable pressure. This was using a hydrostatic test pump that I borrowed from a plumber. My theory is that the tubing slowly expanded in response to the extreme pressure. After about half a day I had a steady 100 lbs.
Yeah.
We used a compressor and made up a fitting for 3500 feet of 1/2" pex - and it took forever - I have heard 45 minutes is not uncommon.
Of course the sun comes out later and pressure rises etc...
Treat every person you meet like you will know them the rest of your life - you just might!
Y'all ever see a "wild hose" with 100psi spurrin' it?
Be safe.
Be much faster.
Fill with water, then pressurize with air.
Filling with 50# water supply . . . 2 minutes? Pressurize in seconds. Air flush with 30# air. . . 4 minutes? Turn off the valves and . . .ready for pour.SamT
At the last fire I went to someone lost control of a 2-1/2" hose with about 150 psi in it. There was a Y-fitting installed with one gate open, so the whole affair took off sideways at a high rate of speed. The Y is cast iron and probably weights about 10 lbs. The pump operator saw it happen and immediately shut down, so no one got clobbered.
I agree a water test would be better, & easier to find any leak.
What do you use to get water into the lines when you are carrying water to new construction?
(our inspector requires 100lb water test, but let us skip the water b/c there wasn't any close enough)
Treat every person you meet like you will know them the rest of your life - you just might!
SamT
I very much disagree. Air can escape holes water can't.. it's better for pressure testing. Save the water for when you know it isn't leaking.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
I see your point - we've only done air testing to date anyway.
A few questions I have always had:
What pressures do the typical radiant heat system you design reach? I am thinking 30-60 psi tops, but I'm relatively new to radiant.
Do tubes encased in concrete really test out to higher pressures because of the concrete? Or is that a myth?
Treat every person you meet like you will know them the rest of your life - you just might!
12 to 15 psi is pretty typical for an operating pressure, 80 for the test.I guess technically pressure would increase in a slab.. when the tubing is heated in concrete, it expands inward because it can't expand outward, but not much. I don't know how much you'd notice it in actual operation though.. the expansion tank would probably take up at least some, if not all of the increase.. I think?-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
NRTRob: your experience is showing again!
Imagine you're pouring a radiant slab. You've filled the tubing with water and then pressurized it to do your "test". Someone does something stupid and punctures a tube during the pour. You see the pressure gauge go down- but how do you tell WHERE the problem is so you can fix it?
Except in rare instances, tubing buried in slabs should be continuous lengths without joints. The tubing itself is very unlikely to leak as supplied or installed. So you're really not "testing" anything.
You're filling the tubing during a pour for two reasons: to provide the tubing with more rigidity to resist kinking etc. which might occur during the pour, and to provide easy means of leak location in the unlikely event that you puncture the tube during the pour. PEX tubing is far from fragile, but concrete finishers are far from gentle- there's still some risk.
Ultimately, you need to do an assembled system pressure test so you can verify that all your joints are tight. A LOW PRESSURE air pre-test, below 15 psig, can help you to find major leaks (or worse still, drain valves you forgot to close, or ports you forgot to plug!) BEFORE doing the filled system test. You still need to do a higher pressure liquid filled test after that. In the absence of codes which dictate a higher test pressure, this test should be done at 1.5x the relief valve setting. It's between you and the boiler manufacturer whether you include the boiler itself in this test by blanking off their factory-installed relief valve. They've already tested their unit per the applicable codes, and re-testing their work represents some risk of damage to the unit if you apply over-pressure during your own test.
Since compressed gas is the thing that stores energy and represents a hazard should something fail during a test, a water-filled test using the supply water pressure is better than a static air-over-water test.
Easy. Fill it with water then and look for the wet concrete, or heat it, wet the top and look for the big blotch that dries first. You can also use water later.So what happens when you fill that tubing, and construction gets delayed past the first frost?I stand by my statement. The system sees no water until it is ready for fill and fire, period. For tubing installed, manifold-based pressure testing at 60-80 psi is best, IMHO of course. In dry systems, nothing stinks more than unnecessary water damage from a hit pipe during construction either. The hissing air is pretty hard to miss too.For the final test, do what you like, but during construction I think it's a bad idea to fill pipes with water.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
Rob: I agree with you in almost everything you've said- I guess I haven't written things clearly enough! Your skill and experience as an installer of these systems comes through clearly in your posts.
Since the tubes are continuous when you use them in a slab pour, you fill them with AIR. In the unlikely event that someone damages a tube during the pour, you'll find the leak right quick that way! There is no point to filling the tubes with water during a pour since you're testing nothing- and leak location becomes impossible until it's too late to fix it without a demo hammer!
As far as 60-80 psig for a pressure test for the assembled system, that's up to you. To me there's little point in testing beyond 1.5x the relief pressure which of course varies with the type of heat source you're using, but for many boilers is about 30 psig. We agree that with all new equipment and a skilled plumber/fitter doing the installation, 45 psig of air pressure during a test probably doesn't represent much of a risk. Most of the plumbing components used in these systems are rated for pressures beyond 100 psig, which means they should withstand a 150 psig test without damage. And much of the stored volume in the system is in tubing which is well restrained in the structure so it cannot whip around and hurt someone if it lets go.
But if a used expansion tank or an old cast iron rad were to let go during an 80 psig air test, someone could be killed. The stored energy inherent in significant volumes of compressed gases, even at pressures as low as 60-80 psig, can be devastating. I'd rather risk some mess from a water leak than risk a catastrophic rupture. If a system is filled with water completely (i.e. no unfilled dead legs, partially filled expansion tank etc.), and since water is nearly incompressible, losing a few drops of water will drop the pressure from 80 psig to 0 and nobody will get hurt. A water-filled test can therefore be very sensitive, and represents zero risk of injury if a rupture were to occur. That's why hydrotests are specified for most piping and pressure vessels- they're safe and effective.
Gotcha.. as long as water is not in the pipes during the active puncture risks, I'm with you. For the record though (hangs head in shame), I'm not an installer; I just work with a whole lot of them, and get the benefit of their trial and error over the years. To be frank I'm personally not sure of the end-of-project pressure test methodology, that's after my partner takes over the project. My only experience is with distribution testing (testing freshly laid pipe, and through the overall construction process).Thanks for the clarifications!-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
Rob, Brian,
Most, if not all, residential HW heating boilers have a 30 psi relief valve installed at, or provided by, the factory. Hydronic system pressures of 12 - 15 pounds is typical, 20 pounds and greater is, IMPO, indicative of poor design or of a problem. Portions of system that are in a "pumping away" configuration can exceed that but again, only with poor piping design (if your total system pressure drop due to flow/friction is greater that 20 to 30 ft, you may want to consider resizing something). Large commercial/industrial systems, high rises, and non-comfort heating systems excepted here. This is why, when some confused plumber installs the circ pump on the return in a primary piped system, there is usually trouble with reliefs lifting.
Pressure increase in system, due to thermal expansion, is uniform throughout the entire system, for all practical purposes. The tubing material expands very little compared to the concrete in the temperature ranges of normal radiant heating systems. The water expansion will be on the order of about 5% of the sytem volume (heating from 40 to 160 degF), absorbed in a properly sized and located expansion tank. The pressure gradient in the system is also relatively constant from the high pressure point of the pumps discharg to the low pressure point of the pump suction. There are no "local" pressure increases in the system aside from the pump(s).
I wasn't really referring to local pressures, just theorizing that pressures in system with a significant proportion of slab encased tubing might rise in pressure slightly as the tubing expands inwards as it's heated.. pex wants to expand a lot more than concrete does.But since I haven't seen any large variations in pressure, I guess even if technically there is an increase, it's small and that's an esoteric point.. and the expansion tank probably just offsets any overall pressure increase it may give. I wonder what the actual effect is on frictional loss/flow rates though? Must be very small indeed.. but curiousity, you know. I'll have to dig up how much that inward expansion really is one of these days.-------------------------------------
-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply
http://www.NRTradiant.com
I agree that the expansion of the PEX in relation to the concrete will cause some constriction of the tubing ID. During heating, there always flow (hopefully, at least) through the tubes. While the pressure drop through the slightly constricted tube might rise slightly (very, very slightly), I am guessing it would be impercievable outside of a laboratory. I haven't crunched the numbers, but I believe the change in the specific volume of the water mass far exceeds tube constriction. In a very large slab, it would be interesting to run the numbers and see the effect. Might be enough to upsize the expansion tank a little
" a water test would be better, & easier to find any leak" I disagree with this, one too. If it'll hold air, it'll hold water. Leaking air makes no mess. If you can read a guage you can tell if it has a leak. Locating a leak might be more difficult, might not. But for testing, use compressed air.
"our inspector requires 100lb water test..." Is this a code requirement?
"our inspector requires 100lb water test..." Is this a code requirement?
Its required by the inspector - I don't know the code on this one - he let us go with an air test only.
Treat every person you meet like you will know them the rest of your life - you just might!
Just curious. I sometimes see inspectors inventing their own criteria, that is neither a code requirement nor a good idea. Knowing the code, if you are an installer or designer is good insurance against unreasonable "inspection" criteria. It is also just plain good practice. However, fighting an inspector that is "confused" on where their authority ends is seldom beneficial either.
Tim: just to be clear: I recommend filling the tubing ONLY with air during a pour, to provide rigidity and leak detection. But this is not really a test.
Air testing of an assembled system is riskier. Limited to 15 psig, an air test represents minimal risk, but at 80-100 psig the stored energy in such a system can be quite significant and can lead to significant injury if something were to fail.
A water-filled test is safer because it represents less stored energy.
A test which extends only to the relief valve setting really isn't much of a test. Most piping is tested to at least 1.5x the relief valve setting. But the required test pressure is set by the applicable codes.
A leak test at 1.5 x relief pressure is sufficient in all residential cases. If it doesn't leak at a pressure it will never see, why stress the system with a higher test pressure?
A 100# test of a 30psi system is a destructive stress test. Plan for a violent failure. Use water with minimal air volume.
Trust me (|:>) I used to test cast iron devices at 3000psi. That was 1.5 times designed gas pressure. I used water and a piston pump. Absolutely no air.
At =>50psi, where all connections were visible, I'ld use air. Air will leak where water won't and it's very easy to bubble test an air fitting. 100# of air will create leaks where 30# of water would never even dream about seeping out.SamT
I think we agree here.
Stu, EPS is more than adequate for a residential slab.
As to water absorption, from where? If your slab is sitting in water, the type of foam isn't going to make any difference.
Why pay twice as much for something that makes no difference?
Joe H
Also, you can get EPS with borate in it.When he posted this I tried to find it on Dow's site, but couldn't.I have looked it up in the past and EPS is available in different densities.But the common stuff is 15 psi compressive strenght, IIRC.That works out to over 1 ton/sq ft.
http://www.r-control.com/eps.asp
These are the guys. You can buy direct with free shipping for (used to be) $1500ºº worth.
I think Type II is the low density stuff, the more dense the more $$ with not much of a change in R value. Type II works under a slab, as you said over a ton per square foot. A 4" slab weighs about 50 lbs sq ft.
Joe H
View Image
All I have on file is this SI (metric) version.RBean
web: http://www.healthyheating.com
blog: http://wonderfulwombs.typepad.com
Hi Rob - good to see you're here too!
Things not mentioned in this thread: control joints & nails.
We work closely with the GC and their concrete crew to determine where control joints will be sawn or zip-stripped. If possible, we route our tubing around them or dip below or cover the tube with a short piece of split flexible plactic corrugated conduit (like you find under the hood of your truck or car that's used to house wiring bundles.)
We also lay out walls and doorways so we can use the doorway openings for routing tubing. That keeps us out of the GC's nails where tubing might otherwise be within puncture distance for shot nails.
Same goes for supports nailed into the floor to hold up block walls during backfill.
If you really feel a need to raise the grid-bed & tubing, we use cribs & that avoids having the concrete guys, who really don't want this task, stabbing at the grid with pointy rakes! We also do not permit any rough treatment of tubing during a pour & that includes a no smoking policy during the pour. Hot glowing stoogies and PEX do not play well together with 100 PSI in the tubing!