DIY Hydronic Floor Heating

[AC_Hacker]

Originally Posted by pachai

I discussed insulating the slab with my builder, and he reported that he has another project where he is repairing a slab that collapsed on the styrofoam. So, right or wrong, I am going to let him pour a normal (up to code) slab.

It certainly is difficult to try to push for energy saving options once the construction process is underway.

The contractor's role is to get the job done and move on to the next job.

Your contractor may have had a failed insulated slab, but that does not mean that all insulated slabs fail. Insulated slabs are done routinely and successfully all over the world. It's a matter of proper design & construction technique.

As the above illustration of heat-loss gradient shows

• heat loss is greatest through underground walls and foundations,
• next greatest at the edges of uninsulated slab floors,
• next greatest at the perimeter of an uninsulated floor
• least at the central part of an uninsulated slab floor

So the pay-back will be most favorable in the highest loss areas. However, when a radiant floor heating is contemplated, the floor becomes the heat source and it is in contact with the world's biggest heat sink. So slab insulation is of the greatest importance if the radiant floor is to be efficient.

Foundation Insulation

Slab Insulation

From the second link:

Determining How Much Insulation

The RPA Guidelines provide a simple formula for calculating the R-value for heated slab insulation. Subtract the inside design temperature from the outside design temperature and multiply by 0.125. For example a 70° F inside temperature minus a 0° F outside design temperature multiplied by 0.125 equals R-8.75, approximately 2" of polystyrene board insulation. This insulation should at least be installed around the perimeter of the slab to below the frost line. (If this is not possible, it can be installed horizontally four feet under the slab.)

There are many good books that address the daunting task of building an energy efficient house in the country that is the [second] biggest energy consumer in the world (the book images are hyperlinked):

-AC_Hacker

* * *

[pachai]

Thanks again

[AC_Hacker]

I came across this dramatic photo of how various radiant floor strategies look to a Thermal Imaging camera (NOTE: I found a thermal image of warm board and have included it here as a composite insert. It was taken by a different photographer, using a different camera, but the indicated eveness of the heat distribution is quite interesting):

Staple-Up Construction



Suspended Construction (not touching sub-floor)



Channel Construction



Spreader Plate Construction (note fall-off at edges of plate in IR photo)



Continuous Aluminum Layer Construction

Regards,

-AC_Hacker

* * *

[Daox]

Very nice pics AC Hacker. I have in one section of my house a poured concrete floor with hydronic heat. I don't think it has any heat spreaders, so I imagine its closer to the "Chanel" picture. When walking on it you can definitely tell where the tubes are. When going forward with hydronic heat in my house I will definitely be using heat spreaders to even the heat out.

[AC_Hacker]

Originally Posted by Daox

Very nice pics AC Hacker. I have in one section of my house a poured concrete floor with hydronic heat. I don't think it has any heat spreaders, so I imagine its closer to the "Chanel" picture. When walking on it you can definitely tell where the tubes are. When going forward with hydronic heat in my house I will definitely be using heat spreaders to even the heat out.

Hmmmm.... The radiant slab floors are know for even heat. I'd be interested to know how far apart the 'warm stripes' are?

I'd also be interested to know if there is adequate insulation under the slab.

I'm assuming your slab is 3" thick...

-AC_Hacker

[Daox]

I think they are roughly a foot. Its hard to say. I'll measure them if I get a chance.

The previous owner said there is "a lot" of insulation under them. I imagine he filled the joist cavity. I'm not sure how thick the slab is though.

[Daox]

I finally got my 3rd edition copy of Modern Hydronic Heating: For Residential and Light Commercial Buildings. It is supposed to be the bible on hydronic heating and this is the latest edition that JUST came out. I just started reading it this morning. The preface states that interest in solar hot water, heat pumps, and renewable energy systems have grown a ton since the last edition and I'm hoping that means there is a good deal of writing devoted to those types of systems.

[mrd]

Originally Posted by AC_Hacker

However, when a radiant floor heating is contemplated, the floor becomes the heat source and it is in contact with the world's biggest heat sink. So slab insulation is of the greatest importance if the radiant floor is to be efficient.

I've heard this is heavily dependent on ground water. If there isn't ground water near the soil beneath the slab, then the earth acts as a huge insulation bank. Earth has a low R-value, something like 1 per foot, so you see the thermal gradient traverse deeply underground, unlike a short gradient through typical insulation in a wall. The large volume of heated earth has much thermal mass, so it takes a while to heat up.

But I've heard without ground water, the heat will just sit there in the soil. This is how a frost protected shallow foundation (FPSF) can resist frost heaving at low depths, the soil is heated. The heat loss occurs at the perimeter of the foundation, where it finds its way around the insulation up to the cold surface, where it radiates away.

[AC_Hacker]

Originally Posted by mrd

Earth has a low R-value, something like 1 per foot...

Dry earth with R-1 per foot would mean R-0.083 per inch. That's really not so much.

Pink XPS has an insulating factor of R-5.4 per inch.

This would imply that heat loss into the earth would be (5.4/.083) or about 65 times the rate of heat loss through XPS.

I think I would consider the money well spent, if I insulated under my slab.

In addition, the time lag to heat up or cool down an insulated slab is shorter than with an uninsulated slab.

-AC_Hacker

P.S.: I've heard tales of the radiant slabs that were in the homes that were built in Levittown, New York. They were put in without any insulation, as you describe. The residents were able to grow flowers all through the winter in the garden beds around the edges of their houses.

Pretty good thing if you like flowers in January.

* * *

[mrd]

Originally Posted by AC_Hacker

Dry earth with R-1 per foot would mean R-0.083 per inch. That's really not so much.

Pink XPS has an insulating factor of R-5.4 per inch.

This would imply that heat loss into the earth would be (5.4/.083) or about 65 times the rate of heat loss through XPS.

I suppose the comparison would be valid, if you were to replace all of the soil around and beneath the basement with XPS.. although I'm not sure at what point you would stop replacing the soil with XPS, as it just keeps on going.. Considering soil is free, and XPS costs something like $5/cuft.

The important point is the vast majority of heat loss is at surface grade. For example, let's say the basement is 12ft wide. A point in the center of the basement floor is 6ft from any slab edge, a point at the edge is 0 ft from the edge. For simplicity, let's say the average distance to closest edge for all points on the floor is 3ft. Additionally, the basement floor is 6ft below grade.

That is a total average distance of 9ft of soil between the floor and grade. If we assume R-1 per foot of soil, then we have R-9 insulation of the basement floor for free.

If you added 1-inch of XPS underneath the whole slab, you increase from R-9 to R-14. Using the cost saving calculator here, Cost Saving for Insulation Upgrades. If we leave the defaults of 1000 sqft, 5000 hdd, $1.50/therm, 9 to 14 saves us $90/year.

At 1000sqft, your basement is likely wider than 12ft, and thus your R-values are higher.. A better estimate may be more like 600sqft, which would be 12x50, which leads to $60/year. There are lots of variables.

If we pay $16 for a 4x8x1" XPS sheet, that's $0.50/sqft. We save $60/year/600sqft, that's $0.10/sqft savings. So 5-year payback. Not bad.

I wonder if a R-value of 1 per foot is accurate, I just pulled that off the top of my head. Some values listed here: soil layer models
Shows clay soil has a thermal conductivity of 0.25 W/(m^2*K) That's SI U-value, the reciprocal is SI R-value: 1/0.25 = 4. To convert SI R-value to US R-value, multiply by 5.682. 4 x 5.682 = 22.7

I'm going to shut up now, as I have severely underestimated my argument