*
Has anyone out there ever used an 1.5″ sand layer as a thermal mass with PEX tubing instead of a thin poured slab over a plywood subfloor? Any luck or pitfalls?
Discussion Forum
Discussion Forum
Up Next
Video Shorts
Featured Story
The Paslode cordless siding and fencing nailer drives fasteners reliably without the hassle of hoses or compressors.
Featured Video
Highlights
"I have learned so much thanks to the searchable articles on the FHB website. I can confidently say that I expect to be a life-long subscriber." - M.K.
Replies
*
I wouldn't know, but won't it be a little challenging to put flooring on top of this?
*I presume you would place the sand between studs (atop the subfloor) and attach the finish flooring/subflooring to the those studs.Thermal mass occassionally has benefits. Mostly when the sun heats the mass which then carries the heat into the night. Also for the occassional power failure (my house takes 24 hours to notice a power failure, heat-wise; and 36 hours to cool down). But mostly, thermal mass is a pain. It makes your heating system less responsive and requires more sophsicated controls ($$$). The reason for gypcrete in a radiant heating system is as a HEAT EXCHANGER, not as thermal mass. Aluminium fins can be used be as well, but nothing other than concrete is available for $100/ton. And therein lies my concern about sand (versus gypcrete). Sand is somewhat insulative because there is a lot of dead air space and the sand grains are only connected by small points of contact.I would suggest, if you don't want a gypcrete pour, that you consider aluminium clip-on heat fins in a staple-up installation. Light, responsive, easy to install. Not cheap for a bunch of recycled beer cans, but not bad. -David
*That is a good point, David ,but why not just cut out the middle man and drink the beer then cut the can open and bend to fit your pipe . After a couple of cases you may want to wait till the next nite to make more fins.
*David,Today is the first day I've ever had occasion to disagree with you. Usually I find people disagreeable on _something_ w/in 24 hrs, but it's taken 2 years w/ you--some kind of record! >But mostly, thermal mass is a pain.My experience is that thermal mass is a pleasure, as long as it is incorporated into an intelligent design. Then it doesn't take costly controls--sometimes barely any. The biggest benefit is that it has a moderating influence. We don't get the wild daily swings that our neighbors do. On days they've varied >15 degrees inside (w/o any HVAC systems on), we've varied 3 degrees. To me, this translates into comfort. No control systems more sophisticated than a vanilla thermostat (my PLC is to control elec rates, not the HVAC effectiveness).I won't argue your answer on sand, though. And I'll agree on gypcrete as transfer rather than storage--that matches our experience.Jim
*Jim: I'm curious what kind of set-up you have. Gypcrete with tubing in it? What floor coverings? Something I didn't get into in my previous post is the insulating effects of carpet or hardwood over plywood versus tile, linoleum, and/or bare concrete. I still maintain that you can't move a big thermal mass as quickly as a small one, so if conditions suddenly change (outside temp drops 20F, sun stops blasting through the south windows, etc) then the slab will not be as warm as needed and will take some time to catch up. Or the reverse and be too warm.But this effect is worst if 1) flooring is insulative, 2) the mass is large (4-5" slab versus an inch or two of gypcrete), and 3) you can't hit it very hard with more heat due to small heater size or long tubing length.Definitely can be a problem with a small heater, mostly carpet and hardwood, 4" slab, 300-foot tubing length. Then it is helpful to have smarter controls that not only look at where you are temperature wise (proportional) but also where you've been (integral) and where you're going (differential).Another effect of thermal mass is that, how to put this?, your body doesn't average the flucuation out. 67F is fine for 10 minutes if the next 10 minutes are at 71F and then back again (typical forced air system). But 67F starts feeling pretty cold after 3 hours. As 71F is pretty toasty after a long time. Again, this is probably only noticeable with an insulation-covered thick slab. Not with gypcrete.So I agree with you mostly and retract my concern of thermal mass on the scale of a gypcrete pour or 1-2" of sand. But your confortable, even temperatures during the heating season are not a result of thermal mass, exactly. Rather that you are able to add heat at an appropriate rate (not too fast, not too slow) due to very large surface area are being heated. Consider that radiant systems based on heat fins work well also and weigh very little.Something off my radar screen is the non-heating season (outside temps above 45-50F for me which is not much of the year). With the heater off, thermal mass definitely helps. And reduces your heating season length because you can carry the heat of the day into the night and the cool of the night into the day.And it is the high confort level of radiant that lets us nit-pick these more minor details. Consider that many people have forced air systems with temps going up and down all the time, drafts and then no drafts, and dust being blown all over the house. One doesn't need to have a perfect radiant system to avoid those problems.I'll try say something else diagreeable sooner than another 2 years. :-) -David
*I agree that we're nit-picking on details (but it's damn fun!) and that it's RFH that allows this. Forced air would be a whole different discussion--I've always had a problem with blowing 70something air around and wondering why my body ain't warmer!!!!!!Back to thermal mass. I'm focusing mostly on large amounts of concrete. An inch or two of anything (sand, gypcrete, etc) will not hold enough heat to have an effect that lasts more than a few hours. If I heat the gypcrete in dead of winter from 8-midnight, it'll heat the rooms till morning, for example, but no more.What you say about the non-heating season is a large part of my positive experience. Here in Western NC, we have long springs and falls with daytime temps 40s-70s and nighttime 20s-50s. Lots of sun. Heat transfers to our mass in the day, and into the living spaces at night. My HVAC contractor stopped by on a 25 degree morning in 4 layers of thermals, and I'm barefoot and shorts with no heat on. That's typical.What helps, I'm sure, is that the mass is not just our floor. It's concrete walls and roof, too, all inside 3" of urethane foam. So the sun beating down all day will transfer some amt of energy into the shell of the house, which helps us at night. A result is that the big temp swings you mention simply don't affect the living space--the mass has that much of a moderating effect. So I've never needed a fast response system. I can have a system sized 1/3 of typical for this size house, be/c I don't need the high temps or high volume to fast-load or continuous-load the mass. Also, my main floor is 8" concrete, which takes about 8 hrs to start giving off heat, but can do so for days once it starts. We don't use real high temp water--112 or so, so it never gets so hot that we'd feel unfomfortable. I've never had the situation of being too hot and wishing I could "drain" the mass. All that mass, spread out rather than concentrated in one area, and placed on the proper side of the insulation, has worked for us like a thermal flywheel on autopilot. Low cost, low maintenance, low hassle. Jim
*All that and I didn't even answer your question. Sheesh...Walk-in basement floor: 4" concrete. Not heated.Main floor: 8" slab w/ RFH. Bamboo T&G glued to slab; Tile in kitchen and bathroomTop Floor: 1 1/2" gypcrete w/ RFH T&G bamboo glued to gypcrete.
*Jim: Wow. That's a lot of concrete inside the sytrofoam cup. No wonder you get such temperature balancing effects. And by gluing the bamboo down, you minimized the insulation on the slabs. My 3/4" T&G birch is toenailed into two layers of CDX. What tubing length did you use? I let me concrete guy lay it out in the 300-foot rolls that come in the box. When I'm sending out 100F water, I get back 75F water for a long time. Next RFH I do is going to be in 100-foot lengths. It will need a bit larger heat source to take advantage of it, but should allow the heat to be delivered faster without using high temp water and therefore should heat more evenly. A bigger manifold, but laying copper pipe is fun.Sometime I kicked around but didn't do (wife got pregnant during foundation pour so I only had 9 months to finish): Putting a fin-tube radiator (or auto radiator) with fan up high. Or connected by ductwork to the top of the cathedral ceiling. And circulate water from the slab through that. To move heat from the top of the house to the bottom in the shoulder seasons. For cooling, run unmetered 39F city water through the slab and radiator. Opening windows has sufficed for cooling but that is harder to automate. In late spring and early fall, before I have turned the heat on for real, I'll turn it on for 4 hours or so to take the chill off the slab. Do that every 10 days or so in the shoulder season.Both our masses are inside of a lot of foam. But yours are exposed on two sides and less insulated by flooring. An easier to control approach."It's concrete walls and roof, too" Worried about neighbors shooting at you? I've always thought that tilt-up slab, poured in forms to ressemble an old Victorian would be the best housing for inner-city gentrification. A safer place to be during drive-bys till the neighborhood improves. -David
*David, I attached a recent pict. Follow my profile for more info & pix. Being from AK, you may appreciate the thermal efficiency of an igloo. :) Got about 1.25 MM lbs of mass inside the foam.I use a lot of 300' lengths, 1/2" tube, just like Wirsbo speced. :) Actually, don't have a problem with them or with losing temp across the loop. Not anymore, anyway. They had 8 such loops off one manifold w/ a 3/4" feed. Never heated no matter what. They swore the engineering was right. I said theory be damned, I'm cold. Started shutting off loops. At 5, got almost instant response. Hmmmmm. We split it into two 4 loop manifolds (had another feeder nearby by sheer luck) and it works perfectly. They still say it works in theory, but in practice, the hot water wasn't moving through the pipe. Knew that any other flooring would up my temp requirements. Hoping the low-ish water temps will keep the bamboo from contracting too much and showing big gaps in winter.Only remaining challenge is finding a digital slab sensing thermostat (not a set-point controller). Only one I know of isn't UL approved yet (it's primarily for elec radiant) and so not sold in the states.>Sometime I kicked around but didn't doShoot, you're talking about building your own Unico air handler/AC/heater module! That's what mine look like!!! We disconnected the supplemental heater though (the other 3/4 lead I mentioned) be/c the blown air was useless in our sized spaces.>I'll turn it on for 4 hours or soI used a PLC called Magic Module to control when the RFH pumps operate so that I can stick to convoluted elec time-of-use schedule as much as possible. You could use it to save you the manual off/on every day. Uses astronomical clock, so program it once and done. $300 and a little wiring.>Worried about neighbors shooting at you?Nah. Prospective suitors of my daughter! Pleasure as always,Jim
*David ,You mention gypcrete for RFH but would regular concrete work just as well? Radiantec claims it will but most others say use gypcrete. I plan on regular concrete for my floors , easier to get here than gypcrete.
*Don, some of the decision depends on:* thickness of the pour. Gypcrete in our house is 1 1/2". I wouldn't pour concrete that thin.* ease of placement. Our top floor was wood framed and the walls were done. Gypcrete is self-levelling and easier to place. Can't imagine the difficulty of placing concrete up there.* gypcrete is lighter, me thinks, which has effect on structural decisions.* gypcrete helps with reducing floor-to-floor sound-transmission.Now, I had another floor where that didn't matter, and did an elevated concrete slab I love. Point is that other building conditions can drive you one way or the other. You saw the blatant advertisement above for warmboards. That's another option of course. as is staple-up from beneath. Me? I'll take thermal mass any day, which is not an invitation to get into a debate on this with Barry. My stand on this is as fixed as his.Jim
*Sorry for the "blatant advertising" as I do not want to start a spam session. It's just that many of the issues being addressed in this thread, Warmboard speaks to very well. I will entertain any discussions offline regarding why I think that, but I am interested in Jim's opinions about why thermal mass is better for RFH than low mass. Seriously. I promise to be good.Barry
*In the SouthWest USA they frequently put brick floors over sand in adobe houses. More than 1 1/2 inches of sand though. Some using the Pex in their radient heated floors. Depends on where you are in the USA.
*Thanks for the info Dave , I was told that the gypcrete would shrink much more than concrete . I do know that gyp is lighter than regular crete. I would not be afraid of 1 1/2 concrete on my floors ,12"oc 16ga. 9 1/4" steel joist 3/4 ply deck,..But ilike the sound deadning part of the gyp.
*don... i wouldn't use 1.5 " concrete on that floor system.. i'd want to see some history and mix specs before i'd even consider it...do you have any history with it ? or know anyone who used it successfully ?
*What Mike said.Gypcrete shrinking? I don't think that's accurate. Not my experience. We have 1600 ft of it and it's tight to the walls and barriers and has no shrinkage cracks. My concrete floor's got em though. The only thing I don't like about gypcrete is the surface. Takes more prep before laying tile, for example (you need an anti-fracture membrane). Easier to scuff when finishing construction and installing flooring, for another example.
*Barry,I didn't say that thermal mass is better than low mass. I said I'd take it any day. That's different. I'd also take my Dodge truck over a Ford and a Mac over a Wintel, and I can do so without having to say one is better. Sometimes there are several right answers (well, never Ford ).That said, here's what I like about mass. First is the way is moderates temperature swings. I've compared with a neighbor on a day when neither of us used heat or cooling. Identical sites and sun exposure. Her low-mass house fluctuated 15 degrees and mine fluctuated only 3 or 4. That's because of thermal mass. On a typical day, that fluctuation would force her to heat or cool for comfort, while I'd leave my HVAC off. $$$ Not suggesting that's scientific, but it's still meaningful to us.Second, the ramp up time is a non-issue for me, and the ramp-down time is part of our strategy. Ramp-up is no problem because our indoor temps change so slowly even w/o any HVAC, that having a lead time for a slab to heat just doesn't matter and isn't noticed. Our 8" slab can give off heat after about 8 hours of circulating water. It will continue to stay warm (maintain room temp) for from one to several days, depending on the part of the house. This allows me to stage the zones (all automated) so that only one is heating at a time. But the others aren't cold--they're radiating heat until it's their turn to get more hot water. That means I never have to run the whole house at one time, and so can use a smaller and cheaper heat pump, and it's more efficient to run that, too. Our gypcrete floor, by contrast, starts to heat within one or two hours, it takes about 4 hours to raise all bedroom temps from 67 to 71 (obviously lots of approximated numbers), and the heat lasts all night. During the night, when our elec rate are lowest, the heat loads into the big slab.Without mass, a room would start to chill as soon as the heat source disappeared. Without any thermal retention, I probably couldn't stage the zones because the temp drop might be felt before I'd cycle around to that zone. The bottom line for me is that mass leads to much more balanced temps and a smaller heating system.Other people in other houses may have other results. I have nothing against your low mass solution. I saw it (or something like it) used on a TV home show remodel, and that seemed particularly appropriate. But for me, mass properly placed and insulated is leading to heating/AC costs (install and operating) of about 1/2 to 1/3 of comparably sized low-mass house, according to my HVAC contractor.Jim
*I think I will put in gypjust for the sound deadning with the steel joist the floor can get very noisy. All I have is hearsay on the floor systems and what info I got from Radiantec.
*Jim:Thank you for your thoughtful reply. I also understand and respect your assertion that you don't have to base your preference on what is better technology. As a society, we have frequently taken the lesser technology and promoted it to prominence (ie: Beta/VHS, Mac/IBM). Regarding the issue of cost: I don't know how the HVAC contractor you refered to came up with his ratio, but it is a fact that low mass systems like Warmboard operate at lower water temperatures (sometime as much as 20 degrees F) to achieve the same BTU/H output as the high mass (concrete, lightweight concrete, gypcrete).Wirsbo's software, among others, comes up with either a higher water temp or closer tube spacing in a slab to achieve comparable output to low mass systems. I will grant, however, that in certain parts of the country concrete is very cheap and easy to come by, and effects the cost of installation marginally in favor of high mass. But concrete adversely effects the structural and controls cost which sometimes offsets the material cost advantage.For operational costs: It takes "x" therms of energy to heat one gallon of water one degree, and energy costs "y" dollars per therm to purchase. If you have to heat it up to 20 degrees more heat in the high mass to achieve the same output, it would appear that you would be spending more dollars to do it.Also, the heat you get while the heating system is idle isn't free. You paid for it while you were waiting for the mass to come up to temp. Feel it now or feel it later for 15-20% more cost.I don't remember what wise old sage told me this, but it has stuck with me for a few years: "Thermal mass is only of benefit for residential heating when two conditions exist simultaneously:1) When the heat is free (as in solar heat), and 2) When the heat is released at the time of need.The first condition is rarely met in standard residential heating system design as we all have to pay for gas, electricity, propane, or some energy source of some kind. The second condition is also rarely met by natural means, but can be met with intentional, artificial effort, such as reset controls and computer brains in the controls that try to anticipate the heating need 4-6 hours from now.With low mass systems like Warmboard, you can program a simple thermostat to call for heat as soon as the need arises. You only get the heat you need when you need it. For total comfort control, you can still employ all the fancy computer stuff and have the system respond for NO fluctuation in heating comfort--mainly because the low mass obeys the system calls so quickly.The biggest complaint I have heard about any radiant system is about "overshoot," where the mass kept cooking while the sun contributed extra heat through the picture windows in mid-afternoon. Mother Nature likes to keep us guessing, especially in the mild climates near the coasts or in the Pacific NW. In Oregon, we can see temperature swings of 20 degrees in a couple of hours sometimes in the spring or fall seasons. High mass systems can't respond to that no matter how many outdoor resets are sold into the project.Another issue related to high mass is one of striping (variations of temperature across the floor--uneven heat from one spot to the other). Concrete slabs usually stripe to the tune of 7-9 degrees F whereas Warmboard tests out at 1-3 degrees. That may not mean as much to the cost issue, but I have heard complaints about slab systems from homeowners who notice the difference. Admittedly, the striping diminishes as the mass loads up, ultimately resulting in even temperatures, but it takes a while.HOWEVER, in favor of high mass I must say that some folks in Fairbanks, AK had a good argument for it for when the power goes out for hours at a time, and the high mass buys them time to get alternate heating up to speed. For them, it is more than a matter of a small inconvenience of having to use candles and flashlights. A couple of hours at -50 deg. can be life threatening.What is really encouraging is that whether low or high mass, radiant floor heating is taking off in popularity, which is what the industry needs to get volume up enough to drive down costs. It is being talked about at the Dept. of Energy in Wash. DC as well as at other energy organization's offices because radiant heat is so much more energy efficient than scorched air. Just as in the vinyl window industry, it may take government regulation to drive consumer interest in this area so cost of implementing radiant heat can compete with forced air.Barry
*Cant speak to radiant tubing but...thermal mass rules! I heat with a masonary heater and have an inch of sand in the upstairs floor (between sleepers), mostly for sound absorbtion but the thermal mass is a bonus. Heat is even, comfortable and cheap all winter, house stays cool through spikes of heat in the summer with no a/c (we're in VT so we dont get a ton of hot days.) Thermal mass acts the same way a flywheel does: you cant start it up quickly or slow it down quickly, but once it is going you can maintain its "speed" with minimal inputs.
*Logic tends to agree with David T's suggestions. I have never done anything with radiant floor heat, but it is all any of my friends talk about. I would assume that the concrete and aggregate turns into a retention radiator and the more solid (molecular) the concrete is in contact with the tubing, the more efficient it would be. I would also be afraid that loose sand would be terribly messy when you do future construction and it would continually filter through the cracks to the room below as people walk on it and as the building contracts and expands.I dream (someday) of tearing out the floor in my garage and repouring it with a network of piping that I could pump heated antifreeze through. Then, I would add some valves and connect some other lines to my new steps (which I haven't installed yet, either) in front of my home, and perhaps even the sidewalk up to the porch. Aw, heck, I could even repour the driveway and install a piped liquid heating grid in it too!!! MY DREAM, however, is to build a woodburning boiler that I could use to burn all the wood that my trees generate all summer and recirculate the heated antifreeze solution to parts of the system or through the entire network. I figure that heated concrete and brick steps would last longer because the heat would not only melt ice and snow, but would also dry it out, preventing re-freezing in cracks and crevaces. Also, a heated garage floor would be so soothing to my wife's car in the winter. Also, if I ever had to work out there, a warm floor would be so comfortable to skid across on a creeper. The garbage man would also appreciate not having to haul away my tree/shrub trimmings also, and I could spread the ashes on the lawn with fertilizer each spring. The most beautiful aspect, though, is that I wouldn't have to pay my utility company anything for this heat except for the minimal electricity to power the recirculating pump. This could put the snow blower out of commission!! kidding of course. Scotty/sioux city.
*Hi David, greetings from Ely MN. You never know who you'll find at FHB. I've got PEX in 4" concrete slab, it's nice. 1 1/2" doesn't sound like much of a thermal mass. Call me at home, I'll tell you all about my system. Wayne.
*Scott: I kicked around a radiant driveway for a long time, trying to come up with a way to operate it cheaply. But looking at the cost of natural gas and heat content of firewood (8,000 BTU/pound), it worked out to $140 of natural gas/1 foot of snowfall (I have a long driveway) or a few thousand pounds of wood. Seemed easier and cheaper to have someone plow it off.I did do a radiant sidwalk because having that ice-free is a very nice convenience and safety issue. That gylcol is heated off of the same HWH that heats the slab in my garage. Costs a few bucks for each use.I'd encourage you to do RFH in the garage. The snow melts off the car much quicker when heated from below. And the warm slab helps evaporate the melt water. Make sure to slope the slab towards the center so the meltwater doesn't run under your tool benches. -David
*Anyone ever tried to finish gypcrete as the finish floor...like done with concrete?near the stream,ajAnd what about the additive Disney World said they put in their concrete sidewalks to supposedly make a long day of walking less tiresome...Rubber granules maybe like we put in cushioned tennis court coatings?
*We walked on ours for about a year until I got some boards on it. Can't imagine it as a finished floor. One reason...the guys used nails to mark the pour depth throughout the field of the floor. Quite a few of the nailheads are flush, but still visible. Reason two...wherever there was a tiny bit of ponding, the surface would flake and crumble. The gyp beneath it would wear away from traffic, and was a real bitch to get glue to stick to. Can't imagine that a finish would give it significantly better abrasion resistance, not that it would look good.
*I've just starting prepping for a 3/4" layer of sand with 1/2" PEX tubing, between strapping 12" OC. The theory (as told to me by the radiant company) is that will get more BTU's per foot output with the sand than with aluminum fins, and the materials cost is a lot lower. So get a 'medium' mass system so a compromise between high and low mass virtues.However, maybe I should use a fine sand to reduce air gaps between grains of sand? And yes there will be a layer of plastic underneath to keep the sand from falling out over time.Any holes in this theory?
*Get sand that you can pack. If it can be pacted, then it is going to be denser. The sand would have many sizes to it... not just smaller.
*AJ is right about the sand size. A distibution of sizes packs more densely and is slightly more thermally conductive. What engineers called "well graded" and geologists called "poorly sorted".Absolute size doesn't matter. Imagine perfect, uniform spheres. You can calculate their density, porosity (40%), and thermal conductivity without even knowing their dimensions.So get a mix designed for road bases or concrete pour bases. And then, as they say, go pound sand. -David
*cool... makes perfect sense now that it's been said. Off to visit the local beach, er, sand & gravel yard.
*Why not stop by the local masonary shop and grt some portland to mix in with the sand , that should help with the packing.
*Regarding the Driveway snow melting. Anyone used a ground source heatpump to provide the required BTU. Maybe circuit that routes available heat to driveway only when residence is not demanding heat?
*We've staged zones like this with a programmable logic controller. No reason it wouldn't work, though when it was cold enough to create a driveway "need", it'd probably also be cold enough to have the house needing the heat.