Radiant ceiling

keyote

Am about to start installing rad ceiling on lower floors having been persuaded they were a better bet than floors which i did upstairs. but I based my judgement on scant evidence besides several recommendations and readings from here. i know uponor supports it and siggy touches on it barely but there not a lot of ceiling info out there i have found. So i thought id just ask for any last minute voices.
My plan is to either use 1/2'' tubes in omega stamped thin plates between ply rips 8'' OC under the sheetrock and although the radiant (homemade warmboard) floor above required me to insulate between the joists with fiberglass i may also add a foil backed styrofoam board under the plywood and plates for good measure. But actually I hope to make my own routed foam panels laminated with graphite. i took Gordy's advice and used a product like that for a small wall zone upstairs and kinda liked them but the quality and especially the price put me off and I have been experimenting and think i can make my own in which case that will be the panel for those unfamiliar graphite actually performs better than aluminum. For those interested the difficulty is the foam density must be dense enough to hold the tubes from popping out and of course either routing or hot wire is something that takes a bit of research and set up i have 3000 sf to go and could save like 12k so im motivated and so far have it working on a 4x8 size though had to go to a denser more expensive board.
anyway im really asking am i crazy to do radiant ceilings, at all? will the plates and ply thing be ok if i get cold feet on the foamboards?

jumper

Cheaper less expensive will be fine. But are you going to use same temperature of water for both floor and ceiling?

Rich_49

I have done radiant ceilings for no less than 8 years and I stand by the thought that ceilings provide better comfort than floors

keyote

Jumper I hope so. Im aware ceiling is more tolerant of high temps and conversely also usually need less loop or lower temps because less cover. But the floor above i didn't use plate i only laid aluminum flashing under tubes, i thought this was causing higher SWT need but once i insulated improved a lot, but still not as much as loopcad seemed to predict. and was afraid since loopcad didn't have plug on for non typical construction my manual inputs were not right. But then realized loopcad intended almost constant flow and i was running shorter cycles when i lengthened cycles i got close to design predictions and suspect after an additional attic blow in i will be dialed in at 127 swt design day as predicted or at least close. obviously dont want to get into a mixing station though i could.
now that top floor had highest load but also highest solarization at times and has 12" OC ,and my ceiling systems I was going to do 8: OC so it does have me worried but not as worried as I was when i thought i under supplied the top floor with my lame DIY warmboard without plates. another thing is currently the top floor although joists are insulated is unconditioned old heating system removed. i think when its not only conditioned but has a ceiling radiant on the other side of joists as well as the attic blow in mentioned it will dramatically improve and swt lowered more.
Im 3 wye brick, spray foam insulation in metal joists, triple pane krypton windows, four stories surrounded by similar building so loopcad and other calculators are not set up for such atypical ( and of course my diy warmboard) so i kind of trust past two years experience over loopcad heat loss though as outlined its starting to look like the numbers i substituted are actually working as things are dialed in. To be honest from what i have learned i think i ought to spread loops or make smaller ceiling panels but am just afraid of being short like i thought i was.

MikeG

This is just my experience on a 14X12 utility room. I do have some copper under the concrete in the stone. I figure this will fail some day but probably not in my lifetime. So I put radiant in the ceiling. I had to work around some can lights but I ripped some 3/4" plywood strips and went across the 2X10 joists used 9" spacings because that is what worked out. I spaced the strips so the thin AL plates would fit between so I could staple the thin AL plates to one side.
There is a bathroom above part of the room. I used the 1/2" Uponor PEX AL in the thin plates. I could have used non AL pex but I didn't want any noise. I put the drywall right on top of the strips. Just had to be careful of the screws. I did put fiberglass above most of the ceiling since some of the area is currently a non insulated garage, but where the bathroom was above I did not so I could get a bit of upward heat. Not as much as I hoped. I run 110* water through the ceiling and then it goes through the floor. Best room in the house. The rest of the house is finned BB and that works fine. I am planning to do a 16x24 family room with extruded plates under the floor with pex this summer. If I knew then what I know now when I added on that room I would have done it in the ceiling. Less cost and easier to do. No obstructions, no worry on floor coverings, furniture etc. Just my small experience.

keyote

yes rich not to lay too much on you but it was you an gordys very positive reaction to my original suggestion that convinced me it was the way to continue the project. I also have read everything else on here about ceilings most by you two and a couple others so i guess im really just about to commit to them in the next day of construction and was double checking and looking for any comments on the method im contemplating

hot_rod

A router with a vac is one way to groove foam. what thickness foam? What about 3/8 tube to save another 1/8" OD.

I worked with Watts a bit to develop those early graphite plates.
I did a prototype home here with it. Tough to staple up and keep tight to the floor. They added the plastic stiffener later.

If you need any I have a whole roll of 24" graphite with peel and stick on the back side. I could mail a sample.

I've been building a RV with SIP panels. I modified my wood burning stamp to relieve the foam for the joining splines. It is much, much less mess than cutting or routing foam. The white beadboard foam melts easily.

Way back in the Heatway days before Watts Radiant they built some router foam with aluminum channel glued in, pre Warmboard days.

They used a 2" high density 60 psi foam. It was almost like working with wood. It held grooves much better compared to the 15 psi.

Here is a sample of it, it got to be crazy expensive to produce.

No need for aluminum faced board IMO.

keyote

MIKE G thanks this first room is 12x 14 but really well insulated its a 17 degree design day here but think we have an ice age coming up so theres that. you dont do ODR huh? well i am having problems with my boiler so im not odr yet either but loopcad predicted 127 swt on design day and at first i found it wasnt keeping up and added a wall zone in the worst room but then realized loopcad was probably assuming near constant flow as i tried longer run times it seemed lioke i could very well get to the predicted curve especially since i can add some crawl space blow insulation (even though i have some open cell between ceiling joists already) above and the floor below were are discussing rad ceilings is now unconditioned. so your numbers confirm what i think that the two systems will both have nice low swt. i just hope they are well matched.
think worse case top floor would maybe call for longer runs than lower floors but with 4400 sf i think there will always be lots of calls at any one time

keyote

Professor HOT ROD! good to hear from you. figures you'd be tinkering with this too.I agree hot wire is the way to go ultimately looked into cnc hot wire systems and plans for foam prolly cheaper to buy one than build but think its prolly not worth it for my short order.though i have half a mind to go into the business if this works the panels i bought are insanely expensive for the cost of the raw materials even at amazon and home depot retail prices. That said ceiling application i figure require a robust omega profile to hold the tube without help. I achieved it with a ball router bit in cheap foam but felt though it held against the pex memory on the workbench i would spend the money on heavier extruded dow board for at least this one room and then see if i could find a cheaper source for expanded in a heavier weight. The panels i used are sunboard brand they are a denser exp foam than HD brand and actually only 7/8 thick for 1/2'' tubes but are a bit fragile for ceiling work I was going to use them and glue them to a osb backerground to make screwing the sheetrock less precision intense. but the cost for the whole job got me thinking about alternatives.and i found graphite rolls 60" that works out with the omega shapes perfect for a 48" width 8'' OC. however I havnt tried the graphite part yet so let me know about your leftovers might be interested i dont think my source has the sticky backing just some sort of backer. was going to spray glue or something. i worried a sticky backing might make getting into the channels hard but they could be blocked before spray glue did you have difficulty one idea was aluminum tape but its even more expensive than the thin alum stamped plates. yes those i think could also be used with foam about the same cost as the graphite. if you rather private email me about graphite and need an email let me know though i think you can do it through heathelp.
what would be really nice is a graphite on a plyable backer but not sure if thats possible the graphene molecule is practically 2 dimensional and so dont think its stretchy but graphite is different so maybe they have graphite paints im thinking about the turns the product i bought solved that by simply slitting the graphit over the turn and i suppose the idea is the tubes pushe the cut edges into the groove somewhat but of course only some of the groove is covered that way.

hot_rod

The tube just needs to stay in the groove until the sheetrock is in place. The graphite will not make noise like aluminum if the tube moves, but you want a tight of contact as possible.

You will never home-make the tight fit of the Radiant Design plates, they have that down pretty well.

Working with glue or adhesive overhead for the tube would be a mess.

Or maybe put the tube up first, wrap the graphite over it

Maybe a Rehau solution with their velcro pex.

keyote

Hot Rod- Bob
I bought a ball router both 5/8 3/4 and 11/16 even in the cheap HD foam they cut a nice omega profile but essentially what was happening is i was setting the bit to cut through the top of the foam and that thin edge worked ok but not great i realised i needed a two step operation first a 7/16 cahnnel cut then the ball router set deep enough to leave about 3/16 channel into the ball groove.5/8 works nice now though with the graphite i might have to step up the ball diameter.
glues. no overhead glue. idea is manufacture panels complete with graphite on ground run em all through router then use spray glue or something to adhere the graphite film to them.I thought trying to glue the film inside the channel was going to be impossible but unneeded. its only need to leave the correct amount of slack over the grooves and adhere the film to the flat areas.one way is to just pop a piece of pipe into the groove when yo get to it then spray the next flat and repeat. as i said the sunfoam guys dont even do that they just fly tight over the groove returns and then slit the film, but they do manage to get the film into the grooves on the straight. I havnt used the radiant design plates but know they are heavy extrusions, i was only looking for the omega shap of good stamped plated and the sunboard foam board i bought for one zone that held the tube on a wall without needing to screw any temp hold downs which would be a nightmare on a ceiling. They seem to call this an omega shape after the greek letter.it works it could be hot wired even easier but too much setup expense for one job even this 3000 sf job. I was going to use a glue on the back of the panels to hold them to the osb sub ceiling. decided 7/16 osb sub ceiling would be worth the .50 sf to have the ability to screw sheetrock anywhere but through a tube instead of having to miss the tubes and also find the studs or lath. aluminums really noisy huh? you get so many opinions on that i wondered if noise was bad install and if only one side stapled no noise of course on ceiling can staple only one side

Gordy

I still say Roth panels are the way to go. You have to make the plates have excellent contact, and the tube to plate contact with the drywall for what ever method you choose. Roth panels, or sunboard near guarantee this. Plus the whole sheet is a conductive surface which also has a bit of mass.

Loose fitting tubes in plates, and floppy plates on drywall no good.

Gordy

I would also not trust screwing 5/8 drywall to only osb, you need to hit a joist with grabbers. You’ll find the grabbers (coarse threaded drywall screws) may strip out in osb.

keyote

Gordy it was your introduction to the sunboard that got me going they are not too far from me and i used them for a small wall zone pretty much the same as the roth except graphite instead of alum which actually transfers faster and better than copper or alum. amazing graphite is the future of everything on earth.
"loose fitting tubes in plates" you mean using plates between the ply rips imagined them tight to hold the tube before rocking but i see what you mean that was my second choice i think i really can make the graphite foam boards well i have made several and ones nice and tight and strong and will cost 2k not 13k.
I have heard others worry about drywall screws no reason i cant also hit the studs but wanted the ability to hit wherever i wanted and do it without thinking about anything but the tubes. could even throw some glue in there. Ill know where the 2x3 sleepers i screwed to the joists are just didnt want to be laying out a subway map on every board then confuse marks and hiss.

any thoughts on 1/2' tube 8'' OC and pretty much entire ceilings

Gordy

Tighter centers lower water temps, and increase tube density for even heat.

keyote

thanks i thought 8 OC would work better than the 12 upstairs though thats working better than i thought once i stretched the runs and will work better still when its no longer un heated on floor below and i add some attic insulation above

hot_rod

Radiant floors are still my preferred emitters. When I come in from the cold with cold feet, standing on warm floors warms me instantly. Remember conduction is the strongest heat transfer

Ceilings are mainly radiant transfer, some convection, not much conduction. Depending on how hard you party

Gordy

I think both have their merits. You can feel that radiant energy while sitting in the recliner. Like the sun not as intense though. The colder OAT the more intense.

Rich_49

Keyote ,

Why do we always make life more difficult ? You should just bite the bullet and use SunFoam panels , they have the graphite , are 8" centers . You could choose 100% coverage or as low as 60% and the rest would be blanks .

keyote

LOL rich youre right of course i used them at gordys suggestion on a wall zone i added to the top floor. I liked them but when i realized the cost to do the two floors it got me wondering how hard making them could be. so i googled graphite sheets and did a little experiment that went so well i decided to make enough for one room and see how well that works. actually thinking of going into business with this i see ways to improve whats out there

Mark Eatherton

I LOVE my radiant ceilings, and the floors, are NEVER cold. Just right, even in bare feet.

Most over looked opportunity today. It's also NOT necessary to put it in every square foot of available space. Place only what's necessary, and put those near the exterior to counter the cold sucking effect of the exterior walls. Put in only what's necessary.

Regarding partial output by convection, now THAT baffles me. Where exactly does the hot air rise to? Hmmm?

Now if you have a reversible ceiling fan, then YES, convective out put to help in recovery from deep set back.

ME

keyote

Thanks mark
the load loopcad calcd was about 12- 15k per ea 1100 sf floor and at this point i think it turned out fairly with correct the design day 127 swt, Currently the top radiant floor has no plates under finished 3/4 oak and no attic /crawl space insul (but does have 8" open cell in ceiling joists) and its unconditioned floor below under construction, so its still hard to be sure, but i think when all that is corrected it will be close to prediction or better. Interestingly i found on a severe below design day it struggled to maintain 69' but was actually very comfortable at 66 i think because it was so tight with walls closed cell and triple panes. That happened because i am not yet doing continuous runs w ODR because boiler is a lemon and don't want to mask its startup problems with continuous runs rather get to the bottom of its problems. My amatuer theory is better to use the full ceiling when possible with 8" OC ( tube is cheap) and lower swt curves if need be than put in less tubes and wish i had more. I dont think with a 10-1 mod con and 4400 sf 4 manifolds and 16 loops its going to be a problem if loads turn out low. any thoughts

Mark Eatherton

keyote said:

Thanks mark
the load loopcad calcd was about 12- 15k per ea 1100 sf floor and at this point i think it turned out fairly with correct the design day 127 swt, Currently the top radiant floor has no plates under finished 3/4 oak and no attic /crawl space insul (but does have 8" open cell in ceiling joists) and its unconditioned floor below under construction, so its still hard to be sure, but i think when all that is corrected it will be close to prediction or better. Interestingly i found on a severe below design day it struggled to maintain 69' but was actually very comfortable at 66 i think because it was so tight with walls closed cell and triple panes. That happened because i am not yet doing continuous runs w ODR because boiler is a lemon and don't want to mask its startup problems with continuous runs rather get to the bottom of its problems. My amatuer theory is better to use the full ceiling when possible with 8" OC ( tube is cheap) and lower swt curves if need be than put in less tubes and wish i had more. I dont think with a 10-1 mod con and 4400 sf 4 manifolds and 16 loops its going to be a problem if loads turn out low. any thoughts

That is what we in this industry refer to as "Futurization". If you can afford to do it now, should you ever change fuel sources, like to a air source heat pump, you are set for low approach temperatures. Good choice on your part.

You'll enjoy the "glow" of the radiant ceiling.

ME

keyote

yeah been thinking about cooling with the panels instead of minisplits

Gordy

Keyote, I hate to see you go through the work, and a substandard product. These panels use manufacturing processes that can only be half **** done in a workshop. Plus you have a manufacture backed product if something goes wrong. I see Roth is using EPP foam for wall panels.

keyote

Thanks gordy im thinking about that since i didnt do a good job on the floor although it seems to work fine it could work much better. but i used sun foams product actually went out to his shop and i gotta say its fine but its pretty rudimentary i could actually make something as good in my shop and pretty easily.In fact it only took an hour to copy the first attempt the expanded foam from HD i found worked but was not dense enough i felt for holding the pipe more than one insertion like say it popped out it and you tried to put it back. so i realized he used a 7/8 foam but it was a bit denser so i coughed up for the dowboard which is extruded and much denser than sunboards even. this routed much better as well as held very strong.it was so easy and id done it on a lark investing just in some cheap router bits i decided to see if sourcing graphite on rolls was a crazy thing involving china and alibaba but it wasnt ebay and amazon had storefronts from an importer and i started talking to their guys and found the exact stuff on the sun boards and probably other boards is about 40 cents a sf and the dow board about 60 cents. so i played around with the scraps of his stuff and found it peeled right off and was able to reapply it to my board pretty easily. I know what youre saying when i first got the idea i thought as rich said why mike do you always have to do things the hard way. but it turned out so easy without even getting serious about it that i started thinking there's even a business opportunity here. to really produce these a cnc hotwire is the way to go but i actually think i can jig up my router table to knock enough out in a day and save 10k and see if im right about the business idea. My total investment so far is about $70 for a foam board and set of ball router bits and an hour playing around plus internet research. So now i need to commit to a roll of graphite film. its going to cost $450 for a 60" x 200' roll witch will make 200'of 48" panels so graphite and dow board combined is about $1 sf compared to close to $5 . i feel its worth the $450 risk to make enough for a couple rooms and then decide if im crazy. Im telling you gordy it routed easy but the bit on the first attempt barely poked out the face of the board which made a wide enough channel to pop the tubes in even in the light board but it was a ball bit so it made the edges pointed and they kinda crumbled popping it in. but it worked well enough i decided to try again and waste another half hour. this time i used his heavier board in a blank spot and set the bit deeper but first run it through with a dado so when the ball bit went through lower the entrance to the channel had square edges. wow it was a piece of cake and i realized he was using cheap foam only 7/8 and a full 1'' of dow board would work better still. so Im thinking Im on to something and start thinking about jigs and well now i gotta at least try it.the other thing was my research into graphite and graphene has me thinking I want to be in that industry cause a lot of things are going to be happening there. i retired a few months ago so gotta stay busy with mad science stuff

keyote

he gordy about my question on putting a tee in a loop did you see it whats the best way to do that or i it not a good idea even for the peculiar situation i have on that one loop?

keyote

EPP foam
no doubt if you really get into this there's all sorts of ways to go i bet they are trying to make a board rigid enough it can go right to studs Im sure styrene could even heard some do but i bet it worries some that it might not. today i was thinking of ways to fortify foams i have built with a cement and styrofoam block that was pretty cool we cut it with regular tools. you could infuse the foam beads with an epoxy before compressing or back it with something more rigid for forms they form iot with plastic and metal reinforcing a mesh back lot of way to make a 1'' board really ridgid cheap roth when i looked them up seemed to be using aluminum not graphite

Rich_49

Since you have a bunch of time on your hands , see what you can find out about the foam / cementitious material that Wedi systems uses , that is probably the ticket since it goes right to stud walls and you tile right over it . I have used it for showers and tub walls before and the stuff is as tough as it gets