Naked Science Forum

Non Life Sciences => Physics, Astronomy & Cosmology => Topic started by: meeotch on 18/02/2014 03:39:15

Title: strange claims about heat transfer
Post by: meeotch on 18/02/2014 03:39:15

I'm trying to evaluate floor heating systems for installation in my new studio, and I have to admit that my thermodynamics is a bit rusty...  But I came across the following, and I could use a sanity-check.

Basically, it's an in-slab heating element that claims to be 2.5x "more efficient" (and thus cheaper to run) than the resistive cable type elements that are widely used.  Link is here (http://www.warmfloor.com/en-us/floor-heating-advantages/radiant-floor-heating-with-a-flat-heater-is-more-efficient).  The page will spit a pdf at you, but don't freak out.  Direct link to pdf here. (http://www.warmfloor.com/images/stories/pdf/efficiency/Efficiency_(Imperial).pdf)

Q1:  I understand that more surface area would lead to a more efficient & even transfer of heat to the slab.  But iirc, resistive conversion of electricity to heat is nearly 100% efficent, and since the element is completely contained, there's nowhere else for the heat to go but into the slab.  So if your building is losing X heat per hour to the environment, you'll end up putting X heat / hour back in to maintain a given temp.  How could an element of one kind or another make any difference?

Q2:  More generally, afaik the point of radiant heating is that it doesn't "waste" energy heating the air, but instead radiates it directly to objects (e.g. people) in the room, allowing you to have lower interior temps.  But for a slab-on-grade building, I'd think the higher temp differential between the slab and the earth would mean greater energy loss through the slab, negating the savings - no?  (The pdf above weirdly doesn't even mention radiant transfer.)

Please don't let the clumsy advertising copy on that webpage bias you, I'm genuinely interested in the physics behind this.  (The company in question has been around for like 30 years, and has a 30,000sf factory entirely heated by their own system, so I don't doubt it's as good as any other in-slab solution.  Just possibly not better.)

Title: Re: strange claims about heat transfer
Post by: alancalverd on 18/02/2014 06:32:49

The trick, I suspect, is to use an element with a large area at low temperature compared with a small diameter cable. To get the same total heat output the cable must be at a higher temperature than the flat element, so the "downward" losses will be greater since the temperature gradient across the insulating layer will be larger (you can consider the underlying earth as an infinite sink at constant temperature).

I've chosen an air-to-water heat pump with 30 degC output for underfloor heating in a new building. It's a lot more comfortable than radiant heaters and has a coefficient of performance of around 3 even at -10 deg C outside. If it was a factory floor I'd probably drop the water temperature to 20 degrees for even better efficiency.

 

Title: Re: strange claims about heat transfer
Post by: CliffordK on 18/02/2014 06:55:12

Meeotch,

I've fixed your broken links.

It is hard to say.

I would be suspicious of their calculations as they seem to determine that their system is better with calculations, rather than with empirical testing.  Why not just build a couple of very basic sample one-room houses, and test out the systems being compared?

While resistive heating all has the same efficiency, there would be reason to believe that there would be differences between different systems.

 [ Invalid Attachment ]

Baseboard heating would get hot near the heater (and the wall that the heater is attached to), but may be cool on the opposite side of the room.  Thus the temptation to turn up the thermostat to compensate.  Thus it may be wasteful.

The floor heater would give better coverage of the whole room, and it is pleasant to walk on with bare feet.

The point with this system is that it covers the whole floor, rather than giving you local hot spots (my diagram of wire/tube heaters).  Also see their notes:

(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fwww.warmfloor.com%2Fimages%2Fstories%2Fimages%2Fdiv%2Floss-to-ground1.png&hash=2df409832c987d224c74c599cf3d1dc9)	(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fwww.warmfloor.com%2Fimages%2Fstories%2Fimages%2Fdiv%2Floss-to-ground2.png&hash=85c81c6df3b0e036adb08889f5c6eb0a)
Heat loss through the subfloor:  10sqft x 1/20 (84-50)°F = 5W	Heat loss through the subfloor:   10sqft x 1/20 (120-50)°F = 10W

They indicate that the hot water is warmer than their heating elements which may well be true.  However, they are calculating the entire 10 sq feet at the higher temperature which may not be the case, or perhaps the water system is only turned on intermittently.  Anyway, it seems as if their heat & area calculations are incorrect.

The calculations in the PDF seem to indicate that a 1 square foot flat element heats about 2 square feet of the floor, while a 1 square foot element (of the same wattage???) of a wire or water based system heats less than 1 square foot, and thus is half as efficient, but the calculations seem suspicious.

I agree with Alancalverd that you would likely find significantly better efficiency with either a forced air heat pump, or a water/water heat pump feeding floor tubing if you have a source for cheap/warm water (well water or closed loop underground heat exchange system).

I've seen descriptions of augmenting a water based floor heating system with water solar panels, or perhaps water coils in a wood stove which may help, at least part of the year.

Title: Re: strange claims about heat transfer
Post by: meeotch on 18/02/2014 19:45:54

Thanks to both of you for the replies.  And your points about the heat being more even & subjectively comfortable are well taken.

Also, I hadn't considered the "downward looking" temp differential...  Though if their argument is that their system is more efficient because it heats up a wider area of the slab in contact with the air, then presumably that argument applies at the area in contact with the subfloor as well.  I.e., cable must have higher temp due to smaller area, but then smaller area implies less heat transfer.  (And both deltaT and A are linear terms... Though I assume it's actually an integral across the area, with the cable having a hotter core and cooler edges, which is not a thing I'm going to try and figure out.)  It's also very weird that they assume all the heat is transferred via convection, when it's a radiant heating system).

I should say that there are other things I like about this system:  it's wired in parallel, unlike cable systems, and you can cut or drill through the elements without knocking out the whole system.  But it is more expensive, and requires several large 240-24VAC step-down transformers.

As for hydronic systems:  I looked into them, but it sounded like a huge hassle, requiring a new boiler, which take up floorspace & need to be vented, etc.  I live in New York City, so anything involving plumbing is a giant regulatory p.i.t.a.  The electric stuff could be installed with minimal or no permitting.  And since I already have a (sadly, inefficient) forced air system, I'm really just looking to "take the chill off" the floors to improve the subjective comfort of the space.

Though I can't afford to build test homes (this is NYC we're talking about), it would be cool to graphically simulate the heat loss / replacement of the whole structure.  Does anyone know of a software package capable of this?

Title: Re: strange claims about heat transfer
Post by: CliffordK on 19/02/2014 22:13:57

Quote from: meeotch on 18/02/2014 19:45:54

I should say that there are other things I like about this system:  it's wired in parallel, unlike cable systems, and you can cut or drill through the elements without knocking out the whole system.  But it is more expensive, and requires several large 240-24VAC step-down transformers.

Let's see,
The system is arranged in strips, 1 foot wide, 8 to 11 watts per lineal (square) foot.

If you could arrange it in groups of 5 identical strips wired in series, you should be able to run it on 120V, or 10 strips for 240V (or 5+5 for additional redundancy and control).

5 strips, 20 feet long would be about 1000 watts, and could be run on a 12AWG, 20A 120V circuit with a fair amount of redundant capacity.  Thus eliminating the transformers and simplifying the circuit.

Transformers aren't cheap.  Also, consider, say 1000W at 24V is 42A, and would need something like a 8 AWG wire if using a single main feeder wire.

I'm not sure how a series wired system would react if you had unequal sizes of pads (like might occur with an partial system failure), but that may be a concern.  The company suggests these will work with solar/wind.  Perhaps ask them what tolerance the pads have for over-voltage conditions.

Title: Re: strange claims about heat transfer
Post by: meeotch on 20/02/2014 22:49:33

Unfortunately, I'm pretty sure that would void the warranty.  (Which is significant, at something like 12 or 20 years - though I'm sure it doesn't cover the cost of smashing up my concrete floor and digging out the broken system to replace it.)  And yes, the transformers are something like $700 apiece, though that's only like 25% of the cost of the system.  Anyway, the amount of DIY on this project is probably limited. 

Hilariously, they (and most other low-voltage system purveyors, to be fair) tout low voltage as a safety feature.

I did notice that they suggest that alternate power sources are a possibility.  I don't know much about solar, but I'd think that would involve turning ?V DC into 24VAC, at an efficiency cost.  I'll ask them about it at some point - they put me in touch with their engineer who designed the transformer.

I'm also curious about power factor, as I'm on a commercial electricity account & will be penalized for presenting a highly inductive load to the supplier.

Title: Re: strange claims about heat transfer
Post by: CliffordK on 21/02/2014 01:09:03

Wiring several in series as mentioned above should create the same voltage drop as one would get with the transformer, assuming you don't get shorts.

However, I suppose the difference would be that since US electrical systems are grounded, then at the first element, you would have a 120V potential difference between the element and ground (even if you wire a 240V string, you would still get the 120V difference to ground). 

I don't know about safety.  I've been "buzzed" by 120V several times with no ill effects. 

If you ran a 120V system, it would likely blow the breaker every time you had a short to ground. 

Using the transformers and 24V system, it may not blow the breaker if it was grounded out.  At the same time, 24V isn't very dangerous unless you try very hard, pierce the skin, and use internal electrodes.

WATER????

The transformers, of course, are a source for energy loss (heating wherever you placed the transformers), as well as potential failure.

The prices of solar panels are coming down a lot if you have clear sky view.  However, the electric solar panels are expensive, and about 10 to 20% efficient. 

If you need heat (and have sunshine when you need the heat), then some kind of a direct heating solar panel is much more efficient, and probably cheaper to configure.

The thing about solar though.  Say you're using a 24V lead acid battery bank.  You might discharge it down to 22V, and charge it up to close to 30V.

If you already are attached to "the grid", then a grid tie system is good (less "wasted" energy) although you may have a very expensive system that is non-functional when the grid goes down.

===========

Thinking about this some more, I'd encourage you to seek out a competitor with a typical stranded wire system and see what they say, as well as observing how comfortable the system is on your bare feet.  I find it dubious that you'll see a significant efficiency gain with the flat panels system.

Title: Re: strange claims about heat transfer
Post by: meeotch on 21/02/2014 21:35:12

That's basically the conclusion I've come to...  I.e., not to count much on the advertised operating cost savings, and to make my decision based on install cost vs. pros/cons.  As mentioned, there's a lot to like about the system in question, especially w.r.t. robustness due to the way they're wired.  (They claim they've never had a failure in whatever 20-30 years of selling the product.)  A few $k savings on install doesn't mean much if your cable breaks down somewhere in the middle of a concrete floor...  Though some of the cable people claim do claim you can find the fault with a sensor, and minimize the amount of floor smashing required to repair it.