Design load for supplemental electric radiant floor
GBA community:
My wife and I are considering the install of an electric radiant heat system in our kitchen renovation. The main heat will be provided by a hydronic kickspace heater, so the electric floor is merely for comfort. The kitchen floor will be tile.
I would like to understand the electrical energy that would be consumed by the system, so I can get a sense of operating costs. Unfortunately, I’ve been unable to find a plot on the Web giving the energy delivery per square foot of radiant heat based on temperature differential. I’m sure that this is out there — I just can’t seem to find it.
If I understand correctly, because the heat is supplemental, the temperature differential that I am considering should be the nominal room temperature (say 68 degrees) relative to the floor surface temperature (say 80 degrees). Can anyone confirm that for me?
Second question: if there a rule of thumb regarding how much heat would go to the desired place (kitchen tile) vs. going downwards into the subfloor? Obviously a simple temperature differential calculation would ignore this loss. As part of the renovation, I could install insulation under the subfloor without any problems — what R-value would be appropriate?
Lastly, I am not very familiar with how electric radiant systems would modulate load. Is it done by simply cycling on and off, or by modulating the level of current run through the wires?
Many thanks to everyone in advance. I’ve learned so much from this community, and am grateful.
Brian
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Electric heat is modulated by duty cycle.
Can you explain what you mean by "supplemental"? From a couple of your remarks, I'm guessing you're wanting the floor to feel warm on your feet. If that's the case, you've got a conflict. The floor temperature required for that will mean that the floor will be the primary heat. In some cases, if your house is very well insulated, then you can't even get "warm" floors without overheating the space.
You're right, I'm just looking for warm feet. As I understand things, a temperature differential of as little as 5 degrees would be sufficient for the perception of warmth. So my thinking is that the heating load is only what's needed to raise the floor temp 5 degrees higher than the indoor design temperature, rather than the outdoor design temperature. The hydronic heat would be responsible for heating relative to the outdoor design temperature. I have about 150 sq ft to heat, so according to another post's guidance of (2*dT*sqft), that would be 1500 BTU/hr. To be fair, I don't have a load calculation for the room, but it's a 17x15 kitchen in a corner of the house which is exposed to a fair amount of wind. I guess I would be surprised if 1500 BTU/hr covers the majority of heating load, but certainly I'm not an expert. Obviously to answer this definitively a load calculation is in order.
For floor heat, the amount of BTU delivered to the room is 2*deltaT*sqft. You can convert this to W by dividing the BTU/h by 3.412. For example a 300sqft kitchen with 75F floor into a 70F room your floor heat is putting out 2*5*300/3.412 870W.
Without insulation, you are delivering about 1/3 this above value to the space bellow, with electric heat that can be a lot of wasted power. Thus always insulate bellow heated floors. An R11-R14 batt is more than enough unless over an overhang. Make sure to seal around the rim joist in the floor heat area, air leaks there can cost a lot of wasted power.
Floor heat has very slow response, thus on/off control is usually what is used. Works great.
You probably want the floor heat less than 80F. That is very nice for warm toes but it could be a good portion of the heat going to the room. You want most of your heat to come from the hydronic heater. Generally I find the best is to run the thermostat with a minimum surface temperature, say ~75F.
Thank you, this makes sense. Based on the formula, the steady state load for a 5 degree differential would be 1500 BTU/hr = 440 W. If I understand correctly, a typical electric mat system can push about 12 W per square foot, which would be 1800 W for my kitchen. So it seems like at steady state the system would be cycling on/off quite a bit. All of this of course assumes perfect insulation below the floor.
You also don't need to deliver heat under the cabinets. So the actual area that you deliver heat for is quite low, and probably only useful in the mornings. For me, with 15x15 kitchen, I ended up with about 50 ft2 of wire + insulation - ran it at 400 watts for 2 hours a day. (So about 1 kw / day)
But I actually found that it wasn't nearly as effective in comfort as heating the basement evenly. If you are in the design, consider hydronic heat to the basement, or a basement mini split.
My cabinet maker would revolt if I told him I was installing heated floors under his work, so definitely not going there! Regarding the basement, it's currently unheated but I have a long-term goal to finish and heat it. Completely agree that conditioning that space will make a big difference in 1st floor comfort. Nevertheless, we live in MA and my wife is from TX, so the directive has been given to ensure warm toes ASAP.
At this point, I should probably mention that doing hydronic radiant for the kitchen is also an option, so I'm open to being convinced that it's the better strategy. Both installers I've consulted are concerned that a hydronic radiant system capacity isn't high enough to heat without adding a supplemental hydronic toe kick, and in light of that it seems to me that the upfront cost of a hydronic radiant floor isn't worth the investment. Hence my desire to calculate the operating cost of electric radiant.
This is a very squishy answer. Really depends on what your existing hydronics looks like. If you can tap off a high temp zone with minimal cost, you can do a decent staple up for similar material cost as a toe kick, but much more labor. You still need a thermostat with floor sensors and a zone valve for this.
With floor heat only setup, what I find is the floor is not warm enough in the shoulder season as the heat doesn't run enough to give you that warm toes feel. With electric you have the option to crank it up if you want.
bdharms,
FWIW, I experimented with a wire based in floor heat kit when I did over our kitchen in Illinois. The local market meant our house was doomed to demolition when we sold, but that was going to be years, so I took the opportunity to test how the heatwire would actually perform so I would be able to make a good design decision when we finally moved to build elsewhere.
The kitchen was only 11x13.5 feet so I bought a small area kit and distributed the wire pattern to be primarily in front of the main counter work space and under the small breakfast table. There was enough cable left to create two additional test areas in front of the stove where I set the cables at 2" spacing and 3" spacing to see if I could stretch the coverage.
The first thing that became apparent was the warm toes sensation set in at about 82-84F surface temperature on the tile. The second thing was even 2" spacing of the wires was detectable, the 3" spacing was very obvious. I can't remember now how the thermostat related the sensor in the floor to the heater wires. I do remember setting the thermostat very high in order to the wires on long enough to really enjoy the warm feeling. More on the sensor later.
As the house was old and not well insulated it kinda balanced out most of the time during the winter. Baking or slow stews with the gas stove made the air quite warm which added to the floor input made things excessively toasty. The kids did like lying on the floor like kittens, though the limits of my wire pattern were quite clear to them as well. Couldn't keep the floor toasty during early fall or spring though. It just made the kitchen too warm if we tried to keep toes happy.
The 12W per sq ft. output for in floor heat is typical and better distributed by the new film type heat materials. Expense and risks of damage to same are something to consider. I think the pre-placed wire mat materials might be more forgiving. Placement of and multiple backups of the floor sensors feeding the thermostat are an absolute must. I had been pre-warned and the first died less than two months in. The second one lasted over a year and the final one didn't ghost before we sold the house. By then it was on the second thermostat unit, so maybe the sensors were being damaged by the first unit. No idea for sure.
When we built, I nixed all ideas of in floor heat due to initial expense and the likely failure risk. Even with hydronic options the risk of major costs due to system failure left me "cold". We went with the PGH-to-better insulation pathway and to be honest the floors aren't that bothersome. Wear socks or have a few scatter rugs. Maybe your floor plan and current insulation levels will keep the kitchen tolerable most of the year. However, like others have noted before the warm floor goal will likely overdrive the kitchen heat needs.
If you need a test of temperature perceptions, run water over your hand until you perceive it as "just warm". Your feet would most likely agree with this assessment. I predict the water temperature will be between 80-84F. For fun, try a drink of the same water. I predict that you will find the "warm" water will be perceived as cool in your mouth.
If you set the faucet water to 74-75 and stick your hand into the stream it will now feel cool. Still, that is close to the temperature your floor should be emitting at to keep a balanced output in most homes. So pick warm toes that will potentially make the kitchen too hot to work in or run the floor temp at a practical but "cool" set point.
Hope this helps.