DIY Hydronic Floor Heating

[Ko_deZ]

Hi AC.

Thicker alu means longer distance between pex pipes. Higher alu-cost, lower pex cost. I suggest cc 200mm with about 17mm pex, and maximum 80m long circuits. Just divide in several circuits if the room is larger.

Regarding pops, there are two possibilities: Air or themal expansion. Both of them are error situations imo.

Air: This means that the water is "clucking" like in a stream. Trough the floor boards the sound might be an annoing clicking sound. This means that there is a significant reduction in waterflow in that circuit, reducing efficiency and increasing thermal difference across the floor. This can be avoided by flushing the circuits propperly and installing a micro-bubble filter (Direct translation, you can probably find one online somewhere. Mine is to the left: link). When flushing, make sure that both your collectors ( where all you circuits meet ) has a valve in each end. One is used for manual adjustment (or you can mount an on/off thermostat ) and the other to compensate for short circuits, making them all have the same waterflow resistance, thus making the manual adjustment consistent across circuits. When flushing, put tap water on one collector, open both valves of a circuit, let the water flow for 4-5 minutes, then close the furthest valve, and after that the closest valve seen from the tap. That will enclose some pressure in the tube, but make sure that no air gets into the circuits. Repeat for all, then somehow make sure that there is no air in the collectors. I have a valve system with each collector allowing me to open and close in all directions so that I can flush each part of my system separately. The system should have something like 2 bar preassure when running to avoid/reduce chance pump cavitation.

As for thermal expansion, that happes most frequently when you have a thermostat. When it opens for warm water, things will expand. Depending on how you mounted the tubes and the aluminium plates, something will make sound, or not if you have done a good job. Anyway, the effect is significantly reduced when using manual control and adjusting for higher heating needs by increasing water temperature slightly instead. Often people have 50C water running in the pipes, which not only reduces the efficiency of the HP, but also reduces lifetime of flooring, gives a high temperature difference across the floor, makes popping noises every time it opens and closes and increases your thermal loss in the to and from pipes. Also the water is not moving most of the time, making air bubbles more likely, which as said above, reduces waterflow and increases the problem by itself.

The best solution imo is a well adjusted system set up manually, with a central thermostat adjusting the accumulator temperature slightly, a couple of degrees only is enough, and a high waterflow to avoid bubble buildup and keeping an even floor temperature across the circuit. Once adjusted, you can forget about it for years, with no electronics or other stuff that can fail.

Hope this helps. As posted before, uponor.se has some great information. Use google translate, or find the link I posted here or in the manifesto. There is some great info there.

[pachai]

PS about the "stamped" plates....I bought 50' of aluminum soffit,
cut it into 6' pieces, and I could not stamp channels into it.

But a friend suggested making carefully calculated folds
to form a U channel 1.5" wide by 6' long by 8" high,
and then stapling it up (flattened) to be 17" wide, and the 1.5" would
become a neat U-channel. I did a few and it looked good.
When I run out of Radiantec plates, I plan to use my own.

I am hearing the snap-crackle-pop when the pump comes on.
Next year, I will try the steady state approach...Thanks

[AC_Hacker]

Originally Posted by Ko_deZ

Thicker alu means longer distance between pex pipes. Higher alu-cost, lower pex cost. I suggest cc 200mm with about 17mm pex, and maximum 80m long circuits. Just divide in several circuits if the room is larger.

Thanks for the suggestions.

So, you are suggesting about 200mm (8") spacing and 17mm (.5") PEX, right?

I think that you are assuming the use of thinner aluminum in such a case?

This is interesting because over here (US) PEX is assumed to be spaced 12" or more apart.

So you are thinking that with thicker aluminum you would get better heat transfer, so the PEX could be spaced further apart, right?

Well, I am thinking that it would be even better to use thicker aluminum AND space the PEX closer together.

With a spiral PEX layout, it would be possible to get spacings of about 6" and use the thicker plates.

What do you think would be the effect of doing that?

-AC_Hacker

[Student 07]

Hi,
I would like to add my two cents about radiant floor heat.

Like most things it is site specific. What works great in one area may not work well in another area. There are many variables that need to be taken into consideration. A properly designed system will outperform an improperly designed system. My recommendation is to have the design engineered and install it yourself.

We own a house in Germany with radiant floor heat. The tubes are embeded in the concrete floors. In Germany it gets cold and stays cold all winter. The tubes heat up the floors mass and radiates it into the room. The mass prevents a "quick warm up", it usually takes almost 3 days for the floor to reach our set temperature. But once it is on it works flawlessly and keeps a steady temp in the house all winter. When it is turned off in the spring it usually takes a couple days for the mass to cool down.

Here in the Portland OR area, we can have summer weather one day and winter weather the next. We needed a system that would respond faster and still have the benefits of radiant floor heat. We chose staple up. It has a very fast response time. We even use programable timers with ours, I would say it responds as fast as a forced air system. I do have tubes in concrete for the radiant floor in our basement which has a more stable temperature and is a better application for that kind of system. The rest of the house is staple up and is better able to compensate for our temperature fluctuations.

When I had our staple up system designed, I got a chance to question the engineer. I had many of the same questions that I have read here.

There are basically three different types of designs: in concrete, staple up, staple down.

In concrete works best in colder climates, gives a nice warm floor. With the proper spacing, correct length of runs, correct size of tubing and the proper pump/ flow speeds, it makes a great floor.

Staple up works best in mild climates, because it has a fast response time. The floor is not cold, but not as warm as a (larger mass) concrete floor. This is a great design for retrofits also or as a supplemental heat source. Floor coverings don't matter. This design has and is being used under carpeted floors (the system has to be designed for the flooring). The floor covering is like insulation, it slows the heat transfer down, but doesn't prevent it. It will still heat your home. I would recommend a radiant barrier and lots of insulaton under the tubing.

The engineer was strongly against staple down where the tubing is installed on top of the sub floor embedded in the underlayment. Over time this kind of design will cause what he called "telegraphing". Which basically means you will be able to see where the tubes are in the floor because of a discoloration of your flooring. I don't know how long it would take to have this occur, I guess it would depend on the flooring. He said there was no benefit to this design only drawbacks. Staple up disipates the heat through the subfloor then even more as it goes through the underlayment before it even gets to your flooring. Flooring directly on the heat source gets very warm over the tubing and causes telegraphing. Aluminum spreader plates may help spread some of that heat out over a larger area; however, I am still skeptical.

Some other things I have seen on here: Staple up requires a higher temperature. This is false. As long as there is a difference in temperature heat will move toward cold. A higher temperature in the tubes creates a larger difference (delta T) and the heat will move faster. You could heat your house with 85* water temps, but response time (change in room temps) would be very slow.

Aluminum spreaders: Some systems are designed with them. I was always told that they make lots of noise, because the tubing expands and contracts at a different rate than the spreader. My opinion is that if you have 3/4" subfloor and 1/4" underlayment, a spreader may be beneficial. If you have 1 1/8" subfloor and 1/2" underlayment, the heat will be disipated plenty before it gets to the floor. I am kind of neutral about spreaders, I have never used them.

I also read about somebody wanting to run their floor at 160* This would not be a good idea. Let me qualify this: it depends on your flooring and design. In Germany I saw many new home owners who had their tile floor "popping tile" because the floor was too hot. Through expansion and contraction and extreme heat to the thinset, it loses adhesion and the tiles come loose. It is better to be patient, lower the temp and the room will warm up.

One last thing, we did have a leak inside the concrete slab. At first it sounded like a nightmare. But, we found where the area was hottest, chipped out the concrete, patched the tubing and patched the floor. It was a hassle, but not a nightmare. We only had to chip out about 1 sf of floor.

Radiant floor heat IMHO is the best. I think a great DIY system is possible. Try to get information that will apply to your system. Research systems that others are using in our area. Call a local heating company who installs radiant floor heat, ask them for references in the area, call those people and discuss their design and how it works; hopefully, they know enough about it to give you enough information to copy it.

I want to point out that I am not an expert. I am not qualified to design a system. I am only passing on my experience with two different radiant floors, my experience of installing them, along with information that I learned.
Whew, I got long winded, and I still probably forgot something.

[AC_Hacker]

Well, radiant floors that will be warmed using heat pumps or solar should be constructed differently that radiant floors that will be warmed by fossil fuel. This is because heat pumps and solar do not produce the high temps that fossil fuel can.

Here is the content of an email I recently received from a friend who is very eco-oriented and is in the business of selling radiant heating systems:

The gist of it is: There is very little thermal difference between Ecowarm, Warmboard, Roth and RauPanel. The contiguous aluminum surface spreads the heat quite nicely and produces the highest surface temp for the lowest water temp. Others have also come up with contiguous aluminum surfaces and good contact with the tubing circumference. They're all good. They're all better than slab. And they're all better at thermal transfer than those without aluminum sheeting. And above floor is always better than underfloor.

Even spread and highest output with low water temps - that's what aluminum coverage will get you.

Then it just comes down to either price or application (or both).

Warmboard will be likely be the best thermal performer for use with a geothermal or air-to-water heat pump given their limited water temp. Warmboard is best applied as the primary structural subfloor. They also have an "R-Panel" for retrofit. Both us 1/2" PEX, which will be important to getting enough BTUs at low water temps.

Ecowarm is also branded as Radiant Complete Panel. Slightly lower manufacturing quality than Warmboard's R-Panel, but similar thermal performance for lower cost.

Thermalboard is significantly lower cost, but smaller tubing means lower BTU carriage through the water and shorter tubing runs (more manifold connections).

Quik-Trak or Climate Panel (exactly the same board) both use even smaller tubing and not enough aluminum to mean anything. Their lower profile, though, might fit the application if you can't get a 1/2" channeled panel under the floor covering without height issues. Very limited BTU output, so don't put this in a ski cabin unless you have supplemental heat. I wouldn't use it with a heat pump.

If you want more head-scratching, add the floor covering variable. Each one of these panels have floor covering caveats, but Warmboard and Warmboard-R stand out as the most friendly to floor covering installs. Acts just like structural plywood ('cause it IS structural plywood).

To be honest, my friend is not an engineer, although he does have computer tools to design installations, so he has a great deal of experience with thermal modeling and real world installations.

To be very honest, he formerly sold WarmBoard and currently sells EcoWarm.


Here is an interesting thermal diagram of two floors, one using aluminum spreader plates and one using staple up:

... here's a really great post that appeared in the Homemade Heat Pump Manifesto thread:

The methods we utilize for the space heating is extremly important. In our home in Canada burrr I had attempted to heat a room over a garage. The garage has in-floor concrete heating but being its not a living area we don't heat it very much (60-65deg.Slab temp75 deg.)The bedroom I had installed loops between the floor joists. Even for just a few more BTUs it was completly ineffective. The room is dam cool with a northern winter wind on the outside wall. Even though at the time we had an water temp of 180 deg.F there was no heating for that room. Now with the Geothermal the water temp is only 105deg.F that underfloor system has no hope. However the main floor concrete slab is the holy grail of low temp heating. There is nothing like a cup of coffee on a sunday morning with bear feet on a warm tile floor (88 degF) looking out the patio door at a -10deg F.wind wipping up the snow.
I don't want to discourage anyone but the benifits to the system are huge. The comfort, uniform steady warmth and the cost to operate. Oil fired heating was costing us more than $5000.00/yr. now with geo thermal $1100.00/yr

The method of tubing routed into a OSB strand board may be effective expecially with some aluminum sheet stock formed into the void and around the tube to dissipate the heat. Keeping the tubes closer helps 8" centres.

My project this summer is now the shop. We removed all the conventional oil fired forced air equipment, maybe a little too soon its still cold here. The real down side is we will have to remove 1200sqft of good concrete floor to install ridgid foam insulation and a new floor with a loop. We have been contacting excavators to dig in our loop 4 ton. For our area thats about 2400 lin ft. We have a very heavy clay soil here and we will be placing the 3/4" dia tube down 6ft. I've obtained 2 compressors 2 tons each. I'm thinking of a double compresor system like that was posted earlier.

Randen

I was in a house recently in Portland, OR that was hydronicly heated with a gas boiler and used staple up on the suspended floors and radiant in concrete in the basement. It was very comfortable.

-AC

[Student 07]

Hi,
Yeah, I have seen this information before. It is part of the reason there is so much incorrect information going around.

You have to remember where the information is coming from. Salesmen learn about what they are selling (usually it is the only product they know), and if you ask them what is the best product- it will be the product they are selling.

The pictures are also very nice; I think I first saw them on a website that sells aluminum spreaders. The picture compares apples to oranges to give their system the advantage. They compare a properly designed spreader system to an improperly designed staple up system.

I have nothing against spreaders, I am sure they work great in some applications. They don't produce more heat; they just spread it out over a larger area. My only reservation about spreaders is that they make noise from the expansion and contraction, from what I have been told.

I wholeheartedly agree with A.C. that the system has to be properly designed to the heat source, the heat load and numerous other variables. I don't believe Randen's system was designed by an engineer for his region. An improperly designed or installed system won't produce the results that are expected.

The beauty of a radiant floor is that the heat source doesn't have to be really hot. I run our staple up floor with 104*, I could turn it down lower; however, since I also use programmable thermostats, the higher temp gives a faster response time.

Of the three different radiant floor designs, I consider the concrete floors to be the only true radiant floor. The concrete has enough mass to absorb the heat and when the heat source is turned off the mass continues to release heat, like a rock placed near a fire.

Spreaders on top of the subfloor transfer heat to the room through conduction. The spreaders are a solid material and heat transfers very well. Technically all three systems use conduction. Spreaders store very little heat, so they would have a faster response time. But, as I mentioned before, having the heat source too close to your finish floor may cause telegraphing. Spreaders between the joists would prevent this, but may cause noise and not really provide any benefit over a properly designed staple up system.

With staple up the wood subfloor is an insulator. The heat has to go through insulation. Because wood is less dense than metal or concrete, heat will not transfer as quickly; but it transfers heat at an acceptable rate to heat a home. Staple up is technically conduction heat, but I would almost consider it as convection heat. It may help if I explain my design.

I use TJIs (truss joists) for my floor. On the bottom of the subfloor (plywood) I installed the PEX (staple up method). Across the bottom of the top chord of the trusses I installed a reflective radiant barrier (about 1 3/4" below the subfloor) creating an air gap between the bottom of the floor and the insulation. Below the radiant barrier I have fiberglass insulation and below that rigid insulation on top of the bottom chord of the trusses.

The air gap is where the PEX is radiating heat, there is little insulation above, with lots of insulation below, and so the heat goes up. The air gap is almost as warm as the PEX; this causes the heat to spread out as it heats up the air gap, sandwiched between two layers of insulation. A heated pipe in the free air would radiate heat like in the picture. Inside an enclosed space the wood floor above the heat source will be warm. The heat will spread across the surface, like a fire on the bottom of a frying pan. Another way to describe it is if you had a box and heavily insulated all but one side, then installed a heat source inside the box, the side without insulation would be warm. Heat easily makes its way through the small amount of insulation that a wood floor has.

Some people get leery of staple up because wood is an insulator. Wood is not as dense as metal, but heat goes through it just fine. Put your hand on the bottom of your wood roof sheathing in the summer if you don't believe me.

I think the key is to have a properly designed system, and a system that works well in your region. There are benefits and drawbacks to each of the three designs, but you can't beat a properly designed radiant floor system.

[AC_Hacker]

Originally Posted by Student 07

Hi,
Yeah, I have seen this information before. It is part of the reason there is so much incorrect information going around.

You have to remember where the information is coming from. Salesmen learn about what they are selling (usually it is the only product they know), and if you ask them what is the best product- it will be the product they are selling.

The pictures are also very nice; I think I first saw them on a website that sells aluminum spreaders. The picture compares apples to oranges to give their system the advantage. They compare a properly designed spreader system to an improperly designed staple up system.

I have heard experienced radiant installers say, "If you insulate well below the pipes, the heat has no place else to go but up."

But unfortunately, this isn't quite true.

If the space under your floor was the same temperature as the temperature in the room, and the space under the pipes was insulated to R-30, and the floor above was subfloor, then floor, then carpet padding, then carpet, the R-value above the pipes could easily be R-4. In this case the heat flow would be:

• Heat flow down = 13.3%
• Heat flow up = 86.67%

If here was just subfloor and floor (no carpet or covering), the R-value might be about R-1.5. In this case the heat flow would be:

• Heat flow down = 5%
• Heat flow up = 95%

If the pipes were put on top of the floor (R-value below the pipes = R-31.5), with aluminum spreaders and cement board (R-0.13) , and then linoleum were put on top of that (R-0.2), the R-value above the pipes would be R-0.33, so the heat flow would look like:

• Heat flow down = 1.05%
• Heat flow up = 98.95%

Here's a temperature chart for a particular product called Quick Track, that addresses a floor covering's effect on feed temperatures caused by different floor coverings R-values:

For simplicity, this chart has ignored the ratio of heat loss above & below, but it is clear that to achieve a given level of heat radiation, in this example, 20 BTU/square foot/hour... if the floor covering R-value is R-3, then the feed temp would have to be about 166F, and if the floor covering R-value was 0.25, the feed temp would have to be about 109F. I put the pink line in because most heat pumps have to struggle to get beyond 120F, so it would be wise to have a floor covering R-value that would be 0.75 or less, in the case of R-31.5, below a floor with Quick Trak as the radiant structure.

So, what I am trying to convey through all of this is that the closer to the radiant surface (R-value wise) the heat carrying pipes are placed, the lower the required feed temperature, and the lower the expended energy.

Also, I have come across a paper that was written by John Siegenthaler, who is highly regarded in the world of radiant heating design. The paper is called "Plateless in Radiantville".

I think you will find the paper interesting.

As far as I have been able to determine, the 'popping noise is caused by the oxygen barrier coating on the PEX. Non-barrier PEX does not seem to have that problem, and PEX-Al-PEX also does not seem to have that problem.

You mentioned that heat flows more easily through dense materials than less dense materials. This is often true, but lead is a much poorer conductor of heat than aluminum.

Best,

-AC

[Student 07]

Hi,
AC you're sounding like a salesman for spreader plates. I think John Siegenthaler says it best in your link Plateless in Radiantville, "A one-size-fits-all approach to radiant tubing installation can land you in hot water".
In the article John Siegenthaler pointed out three different installations, which were either designed or installed improperly. Of course they had problems, an improperly designed and installed spreader system or any system would have problems.
When we initially decided to do radiant floor heat in our new house, I contacted a company in Canada. The engineer there told me to do staple up and not use spreader plates. The company in Canada was backlogged and we would have had to wait too long, so we used a local company. The engineer there made the same recommendation, don't use spreader plates.

IMO they made that recommendation based on my system, needs, region etc., maybe for somebody else they would have recommended spreader plates.

I am not sure where you got your percentages. I was reading Radiantec's website. It stated that a reflective barrier reflects 95% of the heat. To me this means that only 5% even makes it to the insulation.
This is what radiantec has to say about reflective barriers. "This detail is good for crawl spaces and other situations where no heat loss downward at all is wanted. The aluminum reflective foil reflects radiant heat energy upwards. The fiberglass batt controls conducted energy downwards and the foam insulation board adds more R value, insulates the joist and keeps out the wind."

Did your examples take the reflective barrier into consideration? Here is some more information I found on the website.

"Will underfloor heat harm wooden floors?◦If the building is reasonably energy efficient, underfloor heat will not harm wooden floors. Much of the engineering data about underfloor heat was generated years ago when buildings were poorly insulated. As a result, floors needed to be maintained at temperatures of 90 degrees F or higher to keep the building comfortable, a temperature which caused wooden floors to split and crack. In an energy efficient building of today with R-19 insulation in the walls and quality thermopane windows, the floor need only be four degrees warmer than room temperature to heat the building to the mid-seventies when it is zero degrees F outside. The more stable humidity in an energy efficient building and the even temperatures that radiant heating maintains are actually good for wooden floors.
◦Before installing radiant heat in an energy-inefficient building with wooden floors, please consult our technical specialists."

160* feed temps are crazy. With a properly designed system, designed off of heat loss calcs, designed for your region, and your house meets current energy codes, you should never need 160* feed temps. As I quoted above the feed temps only need to be 4* higher than room temp. Of course this requires a properly designed system and that is why they (and I) say consult a technical specialist.

The chart from quicktrack is also misrepresenting information. I agree that floor coverings can add a very small amount of insulation; however, what is the chart showing? Are they referring to a single tube per sf? A properly designed system would take the floor coverings into account and possibly add two tubes per sf so the feed temps don't have to be so high.

Heat does move easier through a dense material than through a less dense material, I am not sure how dense lead is. Less dense material is considered an insulator. Most spreaders are made of aluminum, because it gives up the heat very fast. Car radiators were made of copper, now they are aluminum; it gives up heat fast and doesn't store it in its mass.

Inside solar thermal collectors, spreaders collect heat and transfer it to the tube, for a radiant floor they take the heat from the tube and spread it over a larger area. In some designs spreaders may be needed to do this. There is nothing wrong with spreaders when they are used as designed. I do recommend installing the tubes under the subfloor (if possible) to prevent telegraphing.

There is a lot of misinformation out there so Caveat Emptor. All properly designed and installed radiant floor methods work, each of them has their benefits and drawbacks. One size doesn't fit all.

[AC_Hacker]

Originally Posted by Student 07

Heat does move easier through a dense material than through a less dense material, I am not sure how dense lead is. Less dense material is considered an insulator.

Aluminum is more than 7 times more thermally conductive than lead.

Lead is more than 4 times more dense than aluminum.

Aluminum is not considered an insulator.

Lead is not prized for its thermally conductive properties.


-AC

[AC_Hacker]

below subfloor suspended PEX

Note weak heat transfer. Suspended PEX systems will require very high feed temperatures. Not applicable for heat pump systems.

below subfloor staple up PEX

Note stronger heat transfer, especially where PEX is not sagging. Staple up PEX systems will require high feed temperatures. Not applicable for heat pump systems.

below subfloor staple up with plates

Note very good heat transfer. Staple up PEX systems with plates will require lower feed temperatures. Marginally applicable for heat pump systems.

above subfloor aluminum covered grooved system

Note good heat transfer and even performance across floor. Very good candidate for heat pump systems.

So if there's still any doubt as to if heat transfer plates really work, check this out.

-AC_Hacker