Attention Physics Geeks: How do we improve the functioning of our in-floor radiant heat?

November 30, 2007 7:11 AM

Physics Geeks Please Help: Is it better to turn our radiant floor heating system on and off or to let it run? My partner and I have been debating this, for days and days now. Please help us stop the hate.

We have converted our 120 year old garage/mini-carriage house into a studio. It has an upstairs which is temporarily used as an attic, but will eventually be made into a loft area. (It is well insulated, like with piles of insulation past my knees. There are other issues with "leaks" but we are working on fixing those.) We also have a carpet on the slab, an office type with some backing, but not thick.

The in-floor radiant system was installed by a professional with our help. The slab is on the low end of acceptable thickness, and we're using an electric hot water heater to heat the water. (Someday to be converted to solar, and we can't afford a boiler.) We have a fifty-gallon water heater, and ten gallons in the floor. Closed system.

The problem is that here in Northern Minnesota, we're having trouble getting the heat to "catch up with itself." Since it's a closed system, I have insisted that if we just let it run, it will eventually, but slowly, heat up. My partner says that we should turn the heater on and off in half-hour intervals to let the psi rise, for a few days, to allow the water to heat up enough. I have been resisting probably because I'm an idiot, but also because I'm not sure my partner's philosophy degree means his physics is up to snuff. I believe that the slab, in 10 below weather, will cool too fast, and the cold 10 gallons and cold slab will not be offset by a pause in which the rest of the water is heated in the water heater.

Online research has come to nil, mostly because the information out there is spiked so heavily with people trying to sell the systems.

So, physics geeks: should we turn the system on and off in half hour intervals, or should we leave it run? And does it matter if we do it when it's cold outside or marginally warmer? Night or day?

posted by RedEmma to home & garden (30 comments total)

Did I say "physics geeks" enough times? How silly of me. Apologies for sleepless editing.
posted by RedEmma at 7:13 AM on November 30, 2007

Let it run. These systems are designed to be turned on and left alone. Would you tolerate a furnace that you had to stand over and turn on and off every half hour? Also "letting the PSI rise" is code for "I have no clue what the fuck I'm talking about." There is not a noticeable difference in pressure when a water heater is hot or cold. Turn it on and leave it alone for a day. Run a space heater in the mean time.
posted by cosmicbandito at 7:22 AM on November 30, 2007

Isn't there a thermostat? Usually these kinds of things have a mechanical device that turns the heater on and off.
posted by GuyZero at 7:23 AM on November 30, 2007

cosmicbandito: there is a psi meter on the pipes. it does rise as the water sits in the water heater. don't know what that means, but it does. we've been running the system since September, and we do have a low-impact space heater to supplement. the problem is that the water in the pipes--to the touch--does fluctuate, and doesn't stay hot. (last winter, we had a smaller water heater, which didn't work as well. so we replaced it with the 50-gallon.)

the thermostat is just a regular one we've connected ourselves to turn it on and off. the problem is that the room never gets warm enough, without the space heater, to ever turn the system off. this is a very simple system, so there isn't an in-floor sensor, or zones or anything.
posted by RedEmma at 7:31 AM on November 30, 2007

I agree with all of cosmicbandits comment. Word for word with the PSI statement.

The only way to get the system warm is to warm everything up - stands to reason. You aren't going to get anything hot by turning the heater off.

You have a large amount of water. It will take a relatively large amount of time to heat, as will the floor it is in contact with. Half an hour is probably not enough to get any appreciable heat in the thing that will not immediately dissipate into the cold floor. Once everything gets hot, it will use less energy to stay that way, it will just use more to get it there. I am assuming that the system has a thermostat (if not, fit one).

Is there something to circulate the water? Is this controlled by the thermostat or just the heater?

For maximum efficiency, you may allow the system to turn off for a few hours overnight (if you don't use the space then) but it may take too long to warm up again in the morning and so make the saving moot. When I used to live in a house with an open fire (back in the day) it was an expensive lesson to learn that a very hot fire doesn't work - a lesser, constant, fire was the only way to get the house warm. Albeit only after a few days. Let the thing run.
posted by Brockles at 7:36 AM on November 30, 2007

Ah. More info:

If the system doesn't get the place warm enough, it hasn't got enough capacity to overcome the heat loss. If there is sufficient insulation (it sounds like there is), then there isn't enough heat coming out of the system.

Have you tried leaving it on 24/7 at a sensible thermostat level and seeing what happens? It may need to be on for a few days to get everything warm. The air in the room (where the thermostat is) will be affected by the heat of the structure. Let the whole thing warm up.
posted by Brockles at 7:40 AM on November 30, 2007

i should say, in clarification, that the water in the pipes according to Partner T, is, when left alone to run all the time, getting incrementally cooler. he says that since the slab will hold the heat longer than it will take to heat the water in the heater, and so will gradually improve the heat of the overall system, so we can stop using the space heater so much.

we are exploring the idea of getting a timer.

what i really want to know is whether the slab and the water in the floor will cool too much when it's off (and 5 below zero outside, ground freezing) and negate the effectiveness of letting the water get much hotter in the water heater and then circulating that. will the stopped 10 gallons cooling in the slab (which is also cooling) offset the water which gets hot in the 50 gallon tank?
posted by RedEmma at 7:43 AM on November 30, 2007

we have tried leaving it on 24/7. i thought it was working pretty well, but we were still having to use the space heater. (it was staying at about 60, left alone.) Partner T is basically saying that we could help it along by this process.
posted by RedEmma at 7:46 AM on November 30, 2007

Partner T is basically saying that we could help it along by this process.

What he is suggesting will have the exact opposite effect. His assertion is wrong.

when left alone to run all the time, getting incrementally cooler....

This is because the slab/building is taking more heat out of the system than the heater can output. Turning the heater off will in no way help this, it will just make the pipes you can feel hotter. It is a simple heat in/heat out situation. Once the temperature of the slab rises (very slowly) this effect will lessen.

If the heat loss from the system (from heating the slab and room) is higher than the heater can produce, it will just not get the place warm. Period. The most efficient way of getting it warm is to put as much energy into the system as possible - that is the heater in 'full hot' mode 24 hrs. If it STILL can't get the place warm, nothing you can do will make it better save putting in a bigger heater.
posted by Brockles at 7:55 AM on November 30, 2007

Sounds to me like you still don't have enough capacity. How big a slab? Think about it this way. You're pouring hot water into the slab 10 gallons at a time. The slab has lots of mass. The 10 gallons of water doesn't. So all of the heat energy in the water is quickly dumped into the slab. From there, the water recirculates back to the heater, and the next 10 gallons of hot water comes out. Repeat. Repeat 5 times and you've very quickly used up your 50 gallons of hot water. It's very unlikely that your hot water heater can keep up with the initial demand and quickly just starts recirculating lukewarm or even unheated water. You could try turning off the flow to the floor and giving the hot water heater time to catch up, but that's going to mean standing around and feeling the pipes every 20 minutes or so. Not really a viable option either.
posted by cosmicbandito at 7:58 AM on November 30, 2007

Let it run.

The PSI meter is just there because it's a closed system, and if it was filled incorrectly, or if the water heater was filled incorrectly, the pressures could get dangerous. Also, if there's a leak in the tubing, or too much air in the system, the pressure will drop, and the PSI meter will let you know that there's a problem.

As for the heating, I urge you to consider two things:
1) Electric water heaters are slow. They work, but they take a long time to actually heat water. Turning it on and off might make it so it can *never* recover the water to the expected temperature.

2) I'm not sure how big your installation is, but its likely that there is 10 or 20 gallons of water in the tubes of your radiant system at all times, that go up, go into a cold slab, and nearly instantly turn cold. A problem that will continue to happen until they're able to actually warm the surrounding slab a bit. This will not happen if you're turning it off and on.

FWIW, I'm not a physics major or heating expert, but I have lived in houses that used radiant heat to heat rooms that were built on slabs, and to melt snow on walkways.
posted by Tacos Are Pretty Great at 7:58 AM on November 30, 2007

I think you can't find a satisfactory answer because you haven't asked a clear question, so you can't pin down enough variables to make sense of the situation.

The radiant heat system is putting heat into the concrete slab. Heat is constantly leaving the slab, virtually all of it into either the ground or the room. Having a carpet (aka insulation) on the slab inevitably means that more of the heat is going into the ground, and less into the room, than would if there were no carpet. That much is clear. The thing is, for the radiant system to maintain the room temperature where you want it, more heat has to be coming into the room from the slab than leaves through the walls, ceiling, windows, leaks, etc. Obviously, that isn't happening. You could improve the situation by running the heater more of the time, by increasing the water temperature (if feasible), by removing the carpet, by reducing air leaks and insulation problems...

I have electric radiant heat in my bathroom. Its thermostat does not respond to air temperature; it uses a small sensor that's buried in the mortar under the ceramic tile. Because your slab has so much mass and heats and cools so slowly while air temps can be much more volatile, I don't think it makes any sense for you to use a conventional air temperature-type thermostat to control your system.
posted by jon1270 at 7:59 AM on November 30, 2007

Let it run continuously for an extended period. These systems are intended to reach a stable point and then maintain it, not to act as on-demand heat.

I'm not personally aware of any radiant system where the line PSI is meaningful, except as a means of identifying a broken impeller or something.
posted by aramaic at 8:07 AM on November 30, 2007

The size of the hot water tank is not important since the water it heats is in a closed system. The power consumption of the heater is all that matters, as that tells you how fast it heats up the water and whether it can keep up with the heat transmitted to the room.
posted by cardboard at 8:18 AM on November 30, 2007

Not coming from any physics or home design background, but I do have a question that comes from just thinking about this logically: Does the system not have some kind of automatic control? If so, why would your partner assume that y'all could do a better job of heating your home through manual control than the engineers who spent months (at least) designing the automatic control system?
posted by dondiego87 at 8:21 AM on November 30, 2007

"the water in the pipes according to Partner T, is, when left alone to run all the time, getting incrementally cooler."

Yes! Exactly! This is exactly the principal upon which radiant heating works. You heat the water up and circulate it. As it circulates, it surrenders its heat to its surroundings. By the time it makes it back to the heater, it's cooled down.

Concrete has a lot of thermal mass. It takes a long time to heat up. I don't know the numbers, but I'm thinking many hours. 30 minutes is nowhere near enough time to bring the concrete's temperature into equilibrium with the water coming out of the heater.
posted by adamrice at 8:52 AM on November 30, 2007

What circulates the water? Is there a pump?
posted by exphysicist345 at 8:54 AM on November 30, 2007

A friend of mine had a one floor condo that was over his garage and was only heated by an infloor radiant system. His front door was accidentally left open overnight by someone who had left - in the middle of a -30celcius cold spell.

He told me it took almost 3 days for his house to get back to normal temperature again.
posted by jeffmik at 9:16 AM on November 30, 2007

I think when RedEmma says "leave it off" they mean turning off the pump that circulates the water, not turning off the heater. I don't know if that changes the answer(s) or not, but I think that's what they mean.
posted by entropic at 9:31 AM on November 30, 2007

Yeah, these systems really work at an equilibrium -- it takes a while to warm up, but then they stay that way pretty easily. Let it run.
posted by barnone at 9:36 AM on November 30, 2007

I work in the HVAC-R field. I agree with the above posters.

What your parter is proposing will not work.

In all probability the reason it does not work is because you will not stop screwing with it. Leave it alone for a week, see what happens.

Failing that call a plumber who does service work. Get him out for an hour to look over the system. Tell him your concerns, ask him to ensure everything is set up properly. Yes, it will cost you money. No, you can not get people on the internet to tell you how to fix it. You do not understand the problem, how could you explain it to us so we can tell you the solution. All we can tell you is your partner is out to lunch. Anything further is a guess.

I will make no further recommendations or guesses because I have not seen your system.

You do not need a physics geek, you need a plumber. Skip the online research - call up a plumber who does service work - the information he gives will not be baised towards selling because he gets payed by the hour. Do not go back to the installer, because if the install is undersized or improper, he will never admit it (I see this all the time).

Do not try to jury rig some silly hack appliance timer from kmart. I also see this all the time - and although usually nothing bad happens, it never helps.
posted by vonliebig at 9:44 AM on November 30, 2007

Is there insulation under the slab? If not, you're fucked either way.
posted by electroboy at 9:46 AM on November 30, 2007

Let it run. I fail to see how putting less energy into the system will result in more energy (higher temperature) output, unless you managed to install some newfangled homeopathic radiant heating I'm not aware of.
posted by true at 9:55 AM on November 30, 2007

Your carpet, by the way, is drastically interfering with the transfer of heat into the room from the slab. It's as if you had an electric blanket and a down comforter on your bed, but then put the blanket outside the comforter; most of the heat would go into the cold room rather than the bed. Similarly, much of your heat is going into whatever is under the slab rather than your room because of the carpet.
posted by jamjam at 10:01 AM on November 30, 2007

If entropic is right and that's your partner's belief - that getting the water hotter before you turn the impeller back on is the goal here - then no, there's no advantage to the approach being proposed.

The heater is taking water contained within it and applying energy to make it warmer than it is. Temperature N becomes N+1. The garage loses heat to the ourside, so if it's Temperature M it becomes M-1.

The floor gets heated because when it is at Temperature M and comes in contact with water at Temperature N they will both eventually come to be Temperature (M+N)/2.

Your partner wants to fiddle about and thereby change what number N is. But it doesn't matter - the end result is going to be the same either way, it's just a matter of timing.

If you want a visual metaphor for this, get two jars, three people, and a shitload of marbles. Jar one you name WATER TANK, jar two you name GARAGE. Friend one is HEATER, friend two is PUMP, friend three is FUCKITSCOLDOUTSIDE.

Once a minute, HEATER puts a marble in WATER TANK to represent increasing the temperature of the water. Also once a minute, FUCKITSCOLDOUTSIDE takes a marble out of GARAGE to represent the fact that inevitably the room will lose energy, to radiation or the ground or magic elves - let's just accept that the room will get colder if you don't do anything about it.

PUMP, on the other hand, we'll let do his or her thing whenever they like. Every minute? Once an hour? Doesn't matter. What does matter is they're allowed to do only one thing: move marbles from one jar to another, and only from the jar that's more full to the one that's less full, since when PUMP is working the two jars will average out.

Looked at this way it's obvious that running the pump constantly or intermittently has no impact on making that room any warmer than it was going to get anyway. PSI claims or not, that water can only add energy to the garage that it has in the first place.

There are two possible results to your partner's plan, though. One, the water heater has an upper limit to how hot it will make the water. Once it reaches that point it turns off and you're LOSING heating time till you start pumping again.

Two, the garage has a lower bound to how much the temperature can drop: 32F, or 0C. Maybe other things prevent it from getting that low, I dunno. But if they don't then I have no idea what the repercussions, if any, would be to letting it happen.
posted by phearlez at 10:12 AM on November 30, 2007

If your system is hooked up to a simple on/off thermostat (which it is), then any time which the system is on, the heating element in the water heater will be on full power. It becomes an issue of energy; that is in order for you to receive that maximum amount of heat energy from the system, you must put as much electrical (or gas) energy into the system as possible (a basic thermodynamic principle). Since the system is either on or off with no intermediate setting, you must keep it on at all times in order to get maximum heating benefit.
So set the t-stat, keep it on until the room reaches an equilibrium temperature, which may take a while, but let the thermostat do its job.
posted by goHermGO at 11:35 AM on November 30, 2007

What your partner proposes doesn't make any sense. Why would you turn the heater off? If it's not warm enough, you need to keep the heater going at full output. If it's still not warm enough, you need to make some changes to the system.
posted by mr_roboto at 1:18 PM on November 30, 2007

+1 let it run. All systems using the thermal density of water are intended to run without being shut off.

You might consider glycol or something instead of water. If you're having issues getting it warm enough, you've got an issue somewhere along the system.
posted by TomMelee at 1:58 PM on November 30, 2007

okay. thanks for all your responses.

yes, it's true that we're talking about turning the pump off, not the water heater, for 20-30 minutes so the water heater can heat the water, then running it for over an hour before doing it again.

yes, there is supposedly adequate insulation under the slab, though the slab itself is as thick as we could make it, but not the optimum.

i understand that the carpeting, thin though it is, does have some effect, but it is not an option for this particular room to operate without carpet. (coziness is important, and we both agree that rugs just won't cut it for comfort in this case.)

thanks, vonliebig, for your suggestions. i agree that the warning about not calling the installer is a good idea. however, this particular installer has been very good at not upselling us, helping us out at no cost, and teaches classes at the local community college on in-floor systems. therefore, T trusts him. (i am more skeptical... but see his phone advice at the end of this post.)

phearlez, thanks. that was the kind of explanation i was attempting, muddling through my strictly intuitive understanding of physics, in arguing with T. i just didn't see how mucking with it was going to make things better. he was driving me craaaazy with it.

but he *finally* called his guy, the installer-dude, who's been helping us DIY for a year now. the installer-dude, and y'all, have convinced him to stop this on-off-on craziness. (oh, thankyewthankyewthankyew.) installer-dude says the lack of adequate heat is probably because we have not bled all the air out of the system, which means that the water doesn't move as fast through the system as it should, and explained how we can go about doing that properly. T thought he had done this, but it was obviously not adequate.

so, tomorrow is Get-the-Air-Out Day, and then T is going to let the system do its work. the worst-case scenario is that the pump is not adequate, which will cost about $200. this is unlikely, thank goodness.
posted by RedEmma at 5:34 PM on November 30, 2007

The air answer sounds interesting. The pump, however, doesn't sound likely to me. If the water is getting cooler (while still circulating) then the heat going in is the issue, not the speed of the water. An uprated pump will just mean it just cycles through the system (and thus the heater) faster - I'm not sure how (being as the water is not reaching maximum temp in the time it is in there already) this would help at all. It will get back to the heater faster, but will conversely spend less time in there as well.
posted by Brockles at 5:41 PM on November 30, 2007