I have been researching a new home build like crazy, one of many things to decide is what type of system to run radiant floor heating. I am doing a full basement ranch, radiant in the basement slab and then for the first floor too. I priced out warmboard and its waaayyyy too much money for me. Is a staple up system the best bang for the buck? Any other systems out there I should consider? I am not afraid of doing extra labor to save money and get a better performing product. All my reading and research just leads to more questions, if any "experts" out there can just tell me what to do and why, that would be great lol. Thanks again!!

Not an "expert". More of a DIYer. Take a look at this way of doing it.

http://www.builditsolar.com/Projects/SpaceHeating/SolarShed/House.htm#Press

Kind of making your own warm type board.

Staple up systems are very inefficient and very expensive to operate unless you have access to low cost fuel (e.g., onsite timber). You can use aluminum plates or products like Warm Board to create more efficient above-floor and below-floor hydronic emitters than staple up, but doing so makes the acquisition cost of these approaches unattractive to very many people and you will still end up with sub-optimal performance. Slab-on-grade and thin slab (i.e., Gypsum) are the lowest cost and the most efficient hydronic radiant emitters and what one should being considering for new construction. Above-floor and below-floor systems are more for remodels and where cost isn’t a significant factor to the buyer. We have a fair amount of information about hydronic radiant floor heating and free DIY design tools on our website.

Slab on grade is a great choice for the basement ,
Platform can be done with a surface mount system.
We offer RHT Floor panel system, this is an easy install runs about 2.75 square foot for materials (pipe, manifold, plates turns and site purchased plywood).
Surface mount systems can run low temps so if using a combi condensing boiler it is an easy efficient fit in that both slab and platform can be served with water in the 95 degree range (condensing on the boiler unit).
Topping pours are a great choice as well but you will need to check your structural, adds about 17 lbs per square foot and you will need to double plate the platform bottom plate when building.
Staple ups typically will require 135 degrees or so, and your pump panel will now be a 2 temp panel, default temp on the boiler will be 135 or what ever staple up requires thus eliminating some efficiency should you be using a condensing boiler.
Dan

It is only in contrast that suspended--other than in-slab--radiant floor heating is expensive. Since "very" is not a technical term, let's say that the cost of operation has more to do with the heat load and the heat source than the emitter (radiation).

How you deliver the heat generated is matter of budget and performance expectations.

Aluminum plates attached below the sub-floor are said to increase the efficiency of heat transfer from the PEX to the floor above by 175%. My own sub-floor hyrdronic heating system has never operated above 140F, well within the range of condensing boiler efficiency curves. Remember these design conditions only occur during a few days of the year in most cold climates.

Mine is a 100 year old house with 3/4" sub-floor and 3/4" oak, but the 2# foam and upgraded windows makes my system workable.

Each case is different and you must have someone experienced in hydronic radiant floor design do the math to give you realistic expectations.

Warmboard works because it allows you to eliminate one element of new construction combining sub-floor and aluminum heat transfer plates. We use them regularly depending on the floor plan.

Don't be discouraged by the ill-informed.

Before you can begin to answer the question with any precision you'd have to specify:

What is the heat source/fuel options?

What is the design heat load of the first floor? (I'm not too concerned about the basement...)

Until you know the load, you can't really say what it takes to serve that load, and if cost is a factor it pays to avoid building overkill "just in case".

Since it's a new house you have the option of lowering the loads by the envelope design. Improving the building envelope is often more cost effective than resorting to high-output radiant floors when a simpler/cheaper lower output floor system can get you there.

> ...cost of operation has more to do with the heat load and the heat source than the emiiter

Agreed. Use electric resistance heat with exposed concrete or concrete+tile and you will find it less responsive, no more efficient ($/BTU) and that you often have to overheat the room to make bare feet comfortable (surface temp should be >= 79F).

Cover 25% more surface area and a radiant wall with plates will be more efficient than bare concrete. Even more so for gypsum or concrete with wood over it.

Thank you for your input. I don't think a thin Gypsum slab is what I would like to do because A- added weight of the floor and B I dont want to mess with plumbing/electrical penetrations, framing, or trying to put a wood floor over the concrete. I am building myself as a DIYer and that I think is over my head. I am all ears if you can convince me otherwise though.
Dana - I believe you responded to some of my other threads. I am building a custom log home. I really have no idea how to figure out my necessary heat loads. Footprint is a 40x40 square with 32,000 cubic feet of air space and lots of thermal mass and a very, very, rough estimate of R16-18 for walls. Roof is r50
Blue Ridge - I would be totally fine going with a system such as the RHT. Does you happen to offer it able to accept 5/8s pex piping?

I am planing to use a combi boiler running on propane for heat. Also, I should have mentioned, with the staple up, I would have used a near full coverage of aluminum plates. Does that make a difference?

Log home,
We will be glad to work with you, best way is to go to our site and click on the design submittal page,
Follow the steps and we can contact from there,
We have 5/8" pipe but that is more typical of shop's, warehouses or pole building type structure slabs. rare to use it on residential. 1/2" Barrier pex is the standard.
Dan

The only people who advocate plate systems are those who sell these systems... The best plate system doesn’t come close to a slab-on-grade or thin slab hydronic radiant emitter and will cost 4-5 times more. For a few cents on the dollar, your building engineer can ensure you can handle the load of a thin slab too. Dana is correct that you should first get a heat loss analysis (simplified ACCA Manual J methodology or more accurate ASHRAE methodology) and do your best to reduce this heat load. You might want to read John Siegenthaler’s “Modern Hydronic Radiant Heating” to get better informed and avoid being scammed.

Interesting,
So with regard to a topping pour with pex pipe including concrete or gype, manifolds, staples and engineering, can you provide a set of real numbers as to cost to support 4-5 times less?
US dollars square foot perhaps Pipe, manifold, concrete placed ...
Can you supply real numbers of several of the surface mount platform systems to support your 4-5 times more?
US dollars square foot perhaps Pipe, manifold, sandwich or platform system placed ...
I am just curious how you arrived at these values.
I am relatively neutral as we provide for both systems all the time but am more just curious how you support the values.
Dan

From a cost per BTU-delivered point of view propane is usually substantially more expensive than air source heat pumps, but the avaliable equipment (and design expertise out there) for hydronic air source heat pumps is pretty limited, and to get decent efficiency & capacity out of them at 0F & lower requires heat emitters with very low water temp requirements. The basement slab would not be a problem, but you'd need to design the main floor radiation carefully to be able to go that route. Designing for condensing propane is a lot easier, but in most locations it'll cost more than 2x more to run. The thin gypsum slab is about as low temp as you'd be able to get though, and not outrageously expensive.

Low mass combi boilers generally SUCK for most ~1600' houses, since the peak domestic hot water heating loads are nearly an order of magnitude higher than the design temperature load of even a code-min house. A tank-type combi can work without short-cycling, or a boiler right-sized for the space heating load with an indirect tank for the hot water operated as the priority zone.

The volume of the house isn't always relevant for the heat load unless you're looking at code-max 3 ACH/50 (your leakage will likely be much higher). The exterior surface area is where the conducted heat loss occurs. Here's the lipstick-on-0mirror rough calculation:

The U-factor of an R50 foam roof is about 0.02 BTU per square foot hour per degree temperature difference, the U-factor of an R17 whole-wall assembly is about U0.06, and a code-min window is U0.3, a 2" thick solid exterior door is U0.5.

Measure and up all the surface areas by U-factor, and use 65F as the temperature difference. (The code min interior temp is 68F, the 99th percentile outdoor temperature bins in your area runs +3F to +5F, and the thermal mass of the logs would make even a 65F analyisis a bit conservative.)

So, a 40' x 40' house is 1600 square feet, and assuming a typical 15% window/ floor ratio you'd then have 240' of U0.3 window for window losses of:

U0.3 x 65F x 240' = 4680 BTU/hr

A couple of exterior doors runs about 40 square feet, so door losses would be:

U0.5 x 65F x 40' = 1300 BTU/hr

With a 160' perimeter and 10' walls you have 1600' of gross wall area, less 240' of window and 40' of doors comes to 1320' of U0.06 wall, for wall losses of:

U0.06 x 65F x 1320= 5148 BTU/hr

Assuming a simple gable cathedralized ceiling and an 8:12 pitch the roof area over the 1600' conditioned space will run about 1925 square feet of U0.02 roof, for losses of:

U0.02 x 65F x 1925'= 2503 BTU/hr

Add it all up and you're at 13,631 BTU/hr

Even if you try to air seal it a log home could easily see leakage as high as 300 cfm during winter storm winds. The thermal mass of air is about 0.018 BTU per cubic feet per degree-F. 200 cfm x 60 minutes= 18,000 cubic feet per hour, so the infiltration/ventilation losses could be:

0.018 x 65F x 18,000= 21,060 BTU/hr

That's well over all of the conducted losses combined, which is why air-leaky log homes generally suck on thermal performance, no matter how high the R-values. At a code-max 3ACH/50 and 32,000' of volume, a plywood sheathed house would be looking at only about 5500 cubic feet per hour, and ~6435 BTU/hr of infiltration loss, for about 20K total.

So, with 300cfm infiltration you're at a total of ~35,000 BTU/hr for the losses on the first floor, and 1600' of total floor area probably only 1400' of floor suitable for the radiant, which is a ratio of 35,000/1400= 25 BTU per square foot of radiant floor. To do that at condensing water temps would require either a gypcrete or WarmBoard type of approach, not metal heat spreaders under the subfloor. With sufficiently low temp floors you can probably do that with a 4-ton hydronic air source chiller too.

You're building an ICF basement with an insulated slab. With out running the napkin math on it, let's call it 5000 BTU/hr. (It could be less, but probably not a lot more unless you leave some windows open or don't even try to air seal it.) With on the order of 1400' of fully radiating floor space not covered by boxes furniture appliances etc, that comes out to about 3.6 BTU/ft, which can be delivered at sub-100F water temps.

Run the real loss numbers on the real U-factors and areas, but unless you blower-door test to prove the air infiltration will be much less, keep the 300 cfm infiltration assumption. If you miraculously make code-max (about 1600 cfm/50) you'll need active ventilation, but a Heat Recovery Ventilator (HRV) running 75-150cfm adds only a tiny amount of heat load.

> Any other systems out there I should consider?

I'd seriously consider just using mini-splits on the main floor.

I am an advocate and user of aluminum plates but haven't sold any in years.

We use them regularly in our full-service design work for retrofit situations where nothing else will do. They also make sense in smaller new construction projects where DIY is the only affordable installation option.

Slab-on-ground, such as a basement or ranch is a no-brainer, thus the attraction I suppose. But once you get off the ground so to speak, it takes experience and an open mind to get what you really want.

Whether you use aluminum plates below the floor or sandwiched in between sub-floor and finished depends on the heat load for that particular room and the floor coverings. I like the above-floor or "sandwich" systems topped off with solid wood as they have similar performance gradients to gypcrete and can be installed for less if you provide the labor.

Gypcrete adds thermal mass, of questionable value in most applications, sound attenuation and fire protection. Nice but pricey as you suggest. Concrete will work, but you will need a structural engineer to sign off before pouring. I have done both with good results.

You will find all of these methods mentioned and illustrated in John Siegenthaler's "Modern Hydronic Heating" 2nd addition chapter 10. See section 12 page 348 for the sub-floor illustration.

Having designed, installed and serviced all sorts of hydronic radiant heating systems for the last three decades I prefer a radiant-specific heat load program. We use the Uponor module with our enhanced Manual 'J' 8 Wrightsoft design program.
As for the cost, you may want to take a hint from that self-serving capitalist Dan, and shop for the plates on his full-service website. They are not all the same.

http://www.blueridgecompany.com/radiant/hydronic/527/rht-floor-panel-system

prej·u·dice
ˈprejədəs

noun

A preconceived opinion that is not based on reason or actual experience.

Posted By jonr on 11 Aug 2016 04:30 PM
> Any other systems out there I should consider?

I'd seriously consider just using mini-splits on the main floor.

+1

It would surely be a lot cheaper to install, with very low design risk, and much lower operating costs than anything propane.

Running the basement slab off a propane water heater (or even an electric water heater) would make for at least some cushy-warm floor and reduce the cool basement feel without a huge operating expense, but a 3/4 ton Mitsubishi FH09 would probably be comparable to or cheaper than a radiant slab running off the water heater.

Because I am still at the planning stage, should I move to a gyp-crete system? What is the cost for a 1600 sf house? My current joist spans are 13 feet, and I was planning on using 2x10s at 16 OC. Going to 2x12s wouldn't be too much extra cost, would that be enough for the weight?

I also want to finish my floors with 3/4" hard wood. Could I install 1" wood sleepers in the concrete 12" OC as a nailing surface for the wood? Is there a dry material that a DIYer could use in place of concrete that has good thermal characteristics yet would be easy to install with sleepers?


I also would not be completely against a mini split unit however, I would need to more info on them before going that route.

1- other than for providing AC, they are not popular in my area due to our cold winters. I have never heard of someone being pleased with the performance from them. Are there any newer units that perform better in single digit temps? The year before last we didn't get above freezing for the entire month of February and most of that was single digits or low teens!!

2- With such high ceilings I am worried about the heated air just rising to the top and not staying at the floor, where its needed. I liked the idea of heating the floor and not the air.

3- Would you need to put an air handler in each room? How do you heat say, two bedrooms, two bathrooms, and a great room with one unit?

4- The that Mitsubishi the best unit? It looks like it is for small spaces, and I cant find one with enough BTUs...

Feel free to sell me on which way is best to go. We preferred Radiant however if mini splits are better or if gypcrete is better than staple up please share your expertise it is very helpful.

This is a valid method for installing radiant floor in new construction or renovation...if the ACCA Manual 'J' heat load gives you the information you need to validate such a decision.

Dan can do it with or without gypcrete.

The second hydronic radiant floor heating system I designed and installed (circa 1987) worked perfectly. Just dumb luck, since I did not have the tools to perform a room-by-room heat load, nor the experience to interpret the results if I had.

It is never a good idea to design a building based on internet forum advice. You will likely end up more confused than when you started... I suggest that you consider spending a couple hundred dollars and hire a structural engineer to review and refine your plans based on what you want to do...he or she will likely save you way more then they charge you.

If you want a 3/4” hardwood floor, you should consider something else other than hydronic radiant floor heating. Effectively putting wood insulation between the PEX and the living space will further significantly reduce your operational performance. Yes, there are folks who will advocate doing this too...but they are often the same folks who advocate plate systems... Again, you really need to do some homework and get informed.

1- A properly setup mini split will out perform most everything. Personally, I don’t like the way they look, the way they sound or the way they blow air all around. However, they offer the best operational performance and often offer the best economics too. Mini splits are really a no brainer if you need both heat and AC.

2- Hydronic radiant doesn’t heat the air. Hydronic radiant heats the objects in direct line of sight of the PEX (i.e., not blocked by insulation). Hydronic radiant floor heating excels in high ceiling buildings...although high ceilings ultimately detract from operational performance because they create more exposed surface area and larger air volume.

3- No, one or two heads per floor is typically all that is required. A proper heat loss analysis and proper heat source selection/installation is required to get good results.

4- Mitsubishi makes a great mini split, but Dana likely has the best current vibe given that mini splits have been evolving rapidly.

With regard to operational performance, there is no question that gypcrete is better than staple up. Don’t ever do staple up, period! Gypcrete is better than everything except slab-on-grade. The only possible exception would be a poorly insulated and/or poorly sealed building that is sensitive to outdoor temperature swings. In this case a more responsive heating system might be preferable. However, there is no excuse to build a poorly insulated/sealed building these days. Again, locate a copy of John Siegenthaler’s “Modern Hydronic Heating”...and consider hiring contractors and engineers licensed to operate in your State.

More gibberish.

This is outside the skill set the average structural engineer.

3/4" hardwood is perfectly fine over a properly design radiant floor heating system. I am standing on one now.

Wood is not a good insulator. If it was we would not bother with fiberglass, cellulose, foam, et.al.

Mini-splits are the cheapest to install and operate depending on your available fuel source.

I encourage to use your brain, regardless of any "internet form advice" you may receive.

I have a single-head Mitusbishi in my home and a 2-head Fujitsu for the shop/studio.

"Stapleup" is technical term meaning literally "tubing stapled up to the sub-floor". Some get this long-obsolete technique confused with a suspended tube system, also of marginal utility except in the warmed of climate or in the best construction where "floor warning" is the goal.
In contrast plate diffusion plates allow the competent designer to overcome less-than-ideal floor covering choices to satisfy the homeowner's stated goals.

Radiant panels, be they floors, walls, ceilings or wall-hung heat other objects in the space and indirectly heat the air since convection is a consequence of any space heating appliance in any residence.

I have both a mini-split and radiant floors in my 1921 renovation. The extruded aluminum plates went below the floors as it was the only practical way to install them. Not a problem since I use a Uponor-one of the world leaders in hydronic radiant heating--program that calculates the design temperatures and required outlet for the room I am designing for.

In my slab-on-ground shop/design studio I used mini-splits, radiant slab, radiant ceiling, radiant walls and two designer wall radiators with varied control systems to showcase the infinite possibilities.

Far from the stogie, don't buy it there and never to that mentality.

Stick to your guns and get what works for you from someone who has actually done it.

Gypcrete is very difficult to DIY. It will cost $3-4 per sq.ft. in most locals. If you have a heat load and fact in your floor covering and gypcrete still makes sense, you will have the added benefits of acoustics and fire protection. "Performance" is a relative term. If the gypcrete somehow lowers the design supply water temperature and your heat source operates with greater efficiency in this marginal range, by all means.

If you want 3/4" T&G, as I did, the sub-floor system will give you uncompromising comfort at more than reasonable cost of operation. In fact, the my wood floors in the house often run warmer than the studio as a consequence of the wood floors and their added resistance to the radiant energy from below.

Comfort is my thing, followed by efficiency.

Until you have a proper Manual 'J' heat load, you don't know anything.

I have worked as a licensed contractor in 5 states and designed systems in 15 US states, 4 provinces and 3 countries. The physics of radiant heating don't change with local.

would a mini split be better than even a geothermal system? Those could heat hot water couldnt they? But do they really work in the cold north?