Subscribe or Renew Membership Subscribe Renew

What's the best type of radiant heat

byoung0454's picture

In a prior discussion I wanted some ideas on radiant floor heat. I was set on a gyp pour, but the more I look into the other systems I am beginning to wonder what type is better, a staple up, a radiant panel system, or the gyp pour. This is the last house I plan on building for my family and I want the best system. I would also like to offer it in my custom homes the more I get to know about it.

I like the idea of the thermal mass with the gyp pour but from what I have read about it, it takes longer for it to heat up and longer to cool down. I am not really wild about a system that harder to control. With the radiant panel system the heat source is closer to the finished floor and easier to control, same with the staple up method. If I decide with a dry system I would probably go with the panel system.

Now I know with a gyp pour if I were to have a power outage the thermal mass would contonue to radiate heat and the other would not have much of a thermal mass other than the finished floor. I plan on installing a standby generator so I am not to concerned with a power outage. So I guess what I am getting at is what really is the better system? I plan on doing most of the work myself, and I know the dry system would be a easier DIY project.

With the panel system it would be easier to install the hardwood flooring than with a gyp pour but then I have to use tile backer for my tile. But I will be using more hardwood than tile about 75%. So their are a lot of options to weigh but I am more interested in what system would have better performance.

(post #112730, reply #1 of 58)

What is interesting about your post is how you seem to be convinced that thermal mass is your enemy. True, a gypcrete floor may have a longer response time than a forced-air furnace, but as long as you're willing to plan around the limitations of the RFH system, I doubt you'll beat the comfort it offers.

IMO, whether the house is hard to control or not depends largely (again) on how you design it. Structures that have huge expanses of glass facing the south or west with minimal overhangs are a control nightmare, no matter what heating technology you use, for example. The longer response time of a RFH system simply exacerbates the hidden inssues, if any, in the structure.

That said, the thermal mass of a RFH system is an amazingly simply and comfortable way to heat a home. Due to historic reasons, we limited ourselves to 3/8" PEX installed above the sub-floor and love the system. If you plan to leave the home for extended periods of time in a deep setback, you can always install a whole-house controller that you can contact by telephone or internet.

As for the actual implementation, I agree that a panel system is easier to DIY. In the bathrooms and other areas of the house where there are tile, simply have a lower floor, put in the tubing and put some very dry cement mix on top. Just make sure that the rooms can be heated adequately with RFH, the tubing layout in each room, and that the loops are about the same length (and not too long!).

I have heard of at least one heating contractor who does radiat all the time who allowed the client to install the tubing to save on that part of the bill while focusing his attentions on the rest of the system. Perhaps you can find a similar resource in your area?

(post #112730, reply #4 of 58)

I don't have a problem with the thermal mass nor do I see it as a enemy I was concerned about the response time and comfort with the gyp, I don't want to have to open a window because it is to hot, or wait for 4 hours for it to warm up do to outside temps. I may be going at this wrong, I have never dealt with a system like this before. I don't want to regret wishing I have done one over the other.

(post #112730, reply #7 of 58)

If you engineer the envelope right, the slab will only add comfort by heating your home very evenly. A home that experiences wide ranges in insolation during the day can be a control problem, for example.

Some heating emitters work better in situations with wide temp swings, but the proper envelope (with overhangs to block out summer sun, for example) will reduce disruptions to a minimum.

Use the slabs mass to your advantage and you'll be very happy.

(post #112730, reply #8 of 58)

This is the exact issue I'm working through on my own custom house.  I want it to work and work correctly!

We also are planning about 75% hardwood, and 25% tile, and this might sway your end decision.   To install hardwood correctly over gypcrete, you install a deck, pour the gypcrete  with PEX embedded, then install another deck on top of the gypcrete, and then you can nail the hardwood down.  This seems to me to be very labor intensive, and expensive.

There are several alternatives, such as staple up from underneath.  My opinion is that this should only be used in very low budget applications or remodels.

Another that may prove useful is the Warmboard product that is used as the subfloor and the PEX holder.  They say this is product is pretty spendy but works well.  Along the same lines is something called Quik Trak, that operates similar to Warmboard, but goes on top of an existing subfloor.  With both of these the hardwood can be nailed directly to since the tubes are semi-exposed or open to view.  Hopefully the hardwood installer knows not to shoot a nail into the visible PEX.  Some sort of backerboard is needed for tile over these products as far as I understand.  Maybe it might be better to use a backerboard in place of the PEX holder in the tile areas and use electric radiant heat instead.  Then finished floor hights might be closer?

I'm convinced that the envelope (as mentioned above) is the key and that radiant is the best choice for home heat. 

Remember to plan in a mechanical ventilation system for air exchanges.  I would also have a well referenced qualified expert plan the system out if you plan a DIY install.

The last thing I'm working to understand is they say that you can't run high water temps under hardwood so the system must be sized correctly, proper installation, etc.  This is where having the expert design the system would pay off.

I'm leaning toward the Quiktrak system myself, since I'm doing it myself and already have a subfloor done, but have not made a final decision yet.  Construction is waiting until early spring to resume.  I'm interested to hear what you decide.


(post #112730, reply #9 of 58)

You're quite right, the envelope is where it's at. We have a QuickTrack-like system in our home and are very happy with it.

If our calculations are correct, the supply water temperature will never exceed 120 degrees F. So far, the floors are warm, comfortable, and do not exhibit any kind of splitting, warping, etc.

(post #112730, reply #12 of 58)

Any words of advice with this product?

Seems the water temps are quite high to use under HW? 

(post #112730, reply #14 of 58)

I don't think the water temperatures of well-installed QuickTrack/Climate Panel/etc. are that excessive, considering the much higher loop temperatures that staple-up, Ultra-Fin, or other under-floor systems achieve (180*F). The key is the envelope. Once that is tight, you'll lose little heat, which translates into colder supply temperatures on modern heating systems.

Trouble is, once the envelope is really tight, the floors don't get "warm" until you're in the midst of winter.

(post #112730, reply #10 of 58)

"The last thing I'm working to understand is they say that you can't run high water temps under hardwood ...."

One manufacturer (Wirsbo) recommends that the system be designed such that the surface temperature of a hardwood floor be limited to 80 degF and that the heating load of the radiant floor with 65 deg room temperature setpoint, be no more than 30 btuh/ft^2, at 70, no more than 20 btuh/sf.

The reasons are very straight forward. Higher heat densities and higher temperatures have been shown, through trial and error, to degrade the wood and cause it to dry and crack excessively. This is based on the "standard" 3/4" thick oak nail-down t&g type of flooring. Other types of wood floors may be more or less suceptable to the same effects.

"To install hardwood correctly over gypcrete, .... and then you can nail the hardwood down."  I would recommend against this option. If you can't see the tubes, don't use nails. Over gypcrete, you can glue the floor down or have a floating floor. Else, use the Quck Trak type of install, the you can see and avoid the tubing.

(post #112730, reply #11 of 58)

Thanks for the response.

I didn't intend to hijack this thread but its the very thing i'm sorting out.

The 80* temp range is what i understand, so i guess thats where its important to have the system designed appropriately.

PS I also would like to add that I would "hire" the expert to design the system, and not try to get a freebie out of a free bid even though I intend to install myself.

My hardwood guy actually recommended against gluing down the wood onto gypcrete, not sure of his reasons, and he says putting nailing sleepers in the gypcrete is a PITA.  Gypcrete does semm to be predominant in this area, but i am still leaning toward the Quik Trak for DIY.

(post #112730, reply #13 of 58)

Slightly different slant on this....

When you think about it, gypcrete for radiant floor heat is really only an EXTENSION of the old tried and true method of embedding tubes in a CONCRETE SLAB.

While I would hardly ever pour a concrete slab meant to be lived or worked over WITHOUT embedding tubes...even if the system is not immediately activated...., the sense of beefing up new wood floor construction and modifying your framing to aco[VERY BAD JOBSITE WORD] for the extra 2" just so you can add 2" of slurry that eventually gets sort of hard escapes me.  Then add the problems of attaching hardwood flooring.

I really think the gypcrete method is a hold-over from before the time when light weight, dry, efficient, fast responding, easy to work with and control systems like WarmBoard, Quik-track, Raupanel etc (not to mention the DIY sleeper sandwich with Al plates) became widely available.

If a HVAC contractor actually recomends gypcrete pour as the "best" or "only" way of effectively doing radiant over wood sub-floor, then I would wonder what ELSE he might not know.


(post #112730, reply #15 of 58)

Just a note of clarification: The 80 degF is the resultant surface temperature of the wood, not the water supply temperature. Water supply temperature to get that 80 degF surface temp will vary based on the tube size, installation method, flow rate and heat load, but is typically in the 120 to 140 range, for an "above the joists" installation. Add an extra layer of plywood/osb for a "between the joists" installation and add 20 to 40 degrees to that supply water temperature. This is what software is for.

Depends on where you buy your products. A good supplier can easily determine the correct parameters for the installation. I do this for any of my infloor system customers. Layout drawings, piping and control schematics I will do as well, but typically for a slight increase in the sell price.

(post #112730, reply #16 of 58)

Can on-demand propane water heaters be used effectively for closed radiant heating system?

Does anybody know if electric systems emit emf's through the flooring?

How about  RFH in a relatively mild climate ( Western Oregon ) where below freezing temps are not common?

(post #112730, reply #17 of 58)

Thanks everyone for your input and advice.  I feel that I am heading in the right direction.

I talked to another flooring guy today and he said the same about installing a redundant subfloor over the gypcrete to use as a nailer for the wood.  His reason are that the glue is very expensive and is probably more than the extra subfloor.

I asked specifically about the quiktrak, etc, and he said that the reason he doesn't use it is that he doesn't have much experience with it.  Also turns out that he trained most of the hardwood installers in this area (ex-employees).  Kind of like a lot of things, people are reluctant to change what they are doing unless put into a corner.  If it ain't broke don't fix it mentality.  No offense intended, my Dad is one of them, but I'm slowly getting to him :)

So do you think the Quiktrak products are better, equal or less than the gypcrete routes?   If it was your new house, what would you do?


(post #112730, reply #18 of 58)

First off, you're lumping a wide range of products together that shouldn't be.

Warmboard, Raupanel, and DIY sandwiches with Thermofin "U" have... by far... the best output at the lowest water temperatures. in roughly that order.

Lightweight DIY sandwiches are next (light plates/PAP/3/4" plywood infill).

QuikTrak brings up the rear. QuikTrak is actually beaten by underfloor installations using extruded plates, which are cheaper to boot, but sometimes harder to install.

Well, it's not really the rear. Thermalboard and Subray aren't so hot either.

There are other panels I have not rated. This is because I am not familiar with them :D

If your loads are really low, your floors may not be warm anyway. Radiant ceiling can be a very economical, efficient and comfortable option in that case.

If your house is vulnerable to rapid temp swings.. high solar gain, not so well insulated, compounded by your climate... then high mass systems are harder to control. It's generally doable, but it requires some doing on the controls end. This does include gypcrete. Also in superinsulated homes, you can have problems with overshooting if you use cheap thermostats or improper water temperatures. Note, I say "can", not "will". It's worth using good thermostats though.

However, high mass works fine. It's a cost/benefit analysis that differs region to region, house to house, owner to owner.

I will say if you want the best and cost is no object, Warmboard or, if the subfloor is down, Raupanel is the way to go. You get the best of everything: lowest water temps of any method out there, best response, highest output. For most people though, cost is an object, and so it's worth looking at other methods and seeing what matches up best for you.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #19 of 58)

Thanks NRTRob,  Sometimes it seems like you have to just keep trying until someone sees your post that can answer.  I wasn't aware of shortomings by quiktrak compared to other products.  They looked similar in design.

Do you have any idea of Raupanel costs?  I'm waiting to be contacted by a rep.

I've seen on other threads that you mentioned ceiling radiant before.  No one around here does any radiant ceiling.  How does it install?  Staple down? I'm clueless on this.  lol

My new home has ICF walls for the basement with insulated slab and PEX.  Main floor will be SIPs with good windows and R50 ceiling insulation.  HRV unit also.  So the envelope should be taken care of.

Here in Idaho, the weather can and does turn on a dime.  Spring or Fall will commonly see 65* beautiful day and the next 28* and snow, and back the next day.  I have heard of people cussing their slab heat on days like this.  Thats part of the reason I am inclined with the newer PEX panel approaches, better control.

Last question, We are planning hardwood for about 75% of the house.  Can hardwood be nailed directly over the Raupanel, quiktrak stuff (of course paying attention where the nails are going), or are you better off floating the floor?  It seems that the hardwood choices narrow considerably when you tell them over radiant heat.  Sounds like a engineered floor is best option?

Thanks again for your input!!


Edited 12/30/2005 6:59 pm by CRF

(post #112730, reply #20 of 58)

Raupanel is pricey.. I think it's actually the priciest option available right now.

Radiant ceiling can be installed several ways. We utilize the strapping most people are using for drywall these days anyway (the 1x3s). Add a 12" strip of 3/4" rigid foam between straps. staple up lightweight plates to the strapping, make some cuts, snap in the pipe. It's cheap and quite effective in many cases. But you do have to pay attention to surface temps and ceiling heights.

Floating floors are always best over radiant unless you like wood movement. If not that, quarter or rift sawn thin width wood flooring can be good. watch the moisture, acclimate, yada yada... but floating is best. the RPA has a good flooring guide that is about $5 or $10 and answers a lot of questions.

Sounds like you're doing a superinsulated type home in a high swing area. That would tell me precision would be important if you go high mass, so indoor feedback sensing would be best, probably. It is most certainly doable, but I wouldn't pour and throw in a dial T-stat and expect good results.

If you're DIY'ing, I'd say you probably don't need the highest output out there with a house such as you're describing, and prefab panels would probably be your best bet only if you are looking to save some time. It's quite a price jump to go from a DIY sandwich type (heavy or light) to a prefab panel product. It's true with Warmboard or Raupanel though, your water temps would be super low... probably something like 90 degrees, maybe less... but it would still be quite low with other methods.. maybe something like 100-105. If you want to do solar, maybe it makes sense to go the high end route. Otherwise, I'd save some money (if you like carpentry!) and put it into a higher end heat source. Or even better windows.

Or if cost is no object, do it all :D

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #21 of 58)

When you say DIY sandwich, what do you mean?

I've seen others discuss "making" their own warmboard by using plywood strips and laying PEX in between.  Is this what you're refering to?

The main floor is 2500 ft so unfortunately the budget can only absorb so much.   I do plan on prepiping for a couple solar panels down the road somehow...

(post #112730, reply #22 of 58)

yes, PEX-AL-PEX pipe (to limit expansion), 3/4" plywood, and heavy (extruded) plates like Thermofin U or light (flashing) plates to conduct heat laterally. Effective and cheaper, but admittedly there is more work involved.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #23 of 58)

I checked out your website, and would like to talk to you further on my project. I have learned a lot about radiant systems in the past few weeks,and I want to thank all who have posted their helpful information.

I had planed on going with a gyp pour for my particular system, but with more research I think I want to go with a radiant panel system. Looking at Tekmar's web site on control strategies I feel that a " Full Outdoor Reset with Indoor Temperature Feedback" would work best for me. I also like the way this system works. Going with this system allows me to go with my 24oc floor trusses also.

The house I plan on building will have a pretty good envelope,wall insulation above average and in the attic I plan on going with a R50 Plus sealing all points of air infiltration. I also plan on going with high efficient windows also would like to go with a outdoor wood boiler with a electric boiler for backup.

Before I keep going and going could I get your thoughts on what I have come up with so far?


(post #112730, reply #24 of 58)

Well, if you use a low-mass system, the need for indoor feedback is lessened greatly. It is nice, but unless you have a lot of zones needing a zone manager I would tend towards an outdoor reset control either on the boiler (for a modulating/condensing boiler) or on the mixing device (for other heat sources), and then just use a PWM thermostat (like tekmar's 507-512 series). Between the two, the end result is very close to indoor feedback and it can sometimes be cheaper. how much depends a bit on the zone count though.

However, then you don't get zone managing or central control, which certainly can be nice. and the upcoming internet gateway for Tn4 systems is making me drool... I can't wait until I can log on to a client's system and troubleshoot with all the info there in front of me!!!

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #25 of 58)

I would like to keep it down to 2 zones 1 on the main floor and 1 in the finished basment. For my heat source I want to go with a outdoor wood boiler with a electric boiler back up.

(post #112730, reply #26 of 58)

You'll have to mix down from the wood boiler. If this is a significant area (like an entire floor or more) then I'd go with variable speed injection mixing w/an outdoor reset controller and the PWM thermostats. I'd probably skip the indoor feedback.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #27 of 58)

"So do you think the Quiktrak products are better, equal or less than the gypcrete routes?"

I think the quicktrak/warmboard type of products are better for some applications than a lightweight overpour. Specifically, nailed would flooring. I know that you can put in sleepers or mark the tubes, or put a subfloor nailer over top, but IMO not the best way to go.

 "If it was your new house, what would you do?"

I prefer solid, nailed, hardwood flooring, and with that installation, I plan to use warmboard (or something similar, I haven't finished researching all the options at this time) in my next home. Under tile and/or stone I will use the gypcrete overpour. Personally, I do not like carpet and do not plan to have any of it in my house, carpet with a high quality pad over quiktrak is a suitable method for this type of flooring, as is carpet/pad over gypcrete.

There has been some discussion about the mass of floors, its benefits and potential drawbacks. I'm not going to add to that debate. All systems will have some level of compromise involved, that is just reality. Both lightweight concrete and the plywood/aluminum plate installation methods are relatively low mass. Doesn't make them better or worse than high mass (poured concrete slab) installations.

Edited 1/3/2006 10:05 am by Tim

(post #112730, reply #28 of 58)

Lightweight concrete is NOT "relatively low mass" by any measure that is important to the operation of the system. The lead/lag involved can easily result in temperature swings in the space. I've seen this many times.

Panel systems can be better than gypcrete... quik trak in particular, however, is not... in terms of thermal performance. It may be adequate to do the job, but water temps are higher with QT than with gypcrete by a significant margin. Others are similar, and Warmboard is actually better.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #29 of 58)

Would you settle for "medium mass"? :)

If a spectrum was created, with a thick concrete slab at one end, and a track system at the other, would a gyp overpour or a lightweight concrete 2" pour be in the middle?

(post #112730, reply #30 of 58)

I don't see the point in making the differential. It's a high enough mass to require dealing with the mass as a part of the control strategy in some fashion... nothing else matters unless you're planning a passive solar storage system or off-peak charging.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #33 of 58)

I see an advantage in setting expectations for the client. Background: I design houses--really tight, high mass--and have to talk to clients about what to expect from the systems that they choose for their home. I've found that the better a job I do of setting accurate expectations, the happier they are with the resulting house. Doesn't matter if it's high end or low end...matters only what they get relative to what they expect.

That said, a Warmboard floor will behave differently from a gypcrete floor and that will behave differently from a thick slab. If we humor me by creating a spectrum from low mass to high mass, creating a continuum between, then certain things vary from one end of the spectrum to the other. We could create a line chart for each variable.

For example, response time changes...time to increase the temp one foot above the floor by one degree (just to pick arbitrary numbers) will generally be longer the greater the mass. Other variables are installation cost, water temp, striping, etc.

Since these things vary, then there's an advantage to drawing distinctions. That way I can lay out all the nuances of each option for the client, and they can choose what best suits them. None are inherently superior to all others, but each has a place, and I gotta help those clients understand the distinctions.

That make sense? I'm not trying to promote anything...just trying to understand so that everyone gets the info they need to make decisions.

(post #112730, reply #34 of 58)

I understand what you're saying, but the strategy to control a thinslab will be basically identical to one controlling a slab... and should do it precisely in both cases. Neither should use setbacks, so the only question is... PWM thermostats or indoor feedback? That will be determined by building type, climate and heat gain severity/client preference.

That is, in a good system, time to ramp up and cool down should be a non-issue in most cases. The system should be set and left alone to govern itself.

Basically, the issues and solutions are the same in both a "high" and "medium" mass case. There may be a small window where a high mass may have issues a medium mass won't, but I wouldn't design that close to the edge to try and hit that window.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine

(post #112730, reply #31 of 58)

"Panel systems can be better than gypcrete... quik trak in particular, however, is not... and Warmboard is actually better."

On what do you base this? Test data? Pesonal use? What crietria do you use to define "better"?

(post #112730, reply #32 of 58)

Output at various water temperatures and floor coverings, using MFGs own data and common sense.

a thin, flat flashing of aluminum under 1/2" of plywood bonded with the pipe primarily through silicon will not perform like heavy gauge aluminum in tight fit with the pipe with the aluminum on top of the floor, for one.

Let's look at a situation with four types of install. Load is 20 BTUs/sq ft (about average) with a wood floor over (R1). This data is from Wirsbo's and Warmboard's output charts which my experience tells me is pretty accurate. DIY sandwich is ballparked from my experience since there are no seriously researched output numbers for it in existence, AFAIK.

A 9" o.c. thinslab will need about 115 degree water... not too bad.

Quik Trak, 125 degrees. Note, you only have about 15 more degrees to go here to cover heavier loads or carpeting before you start really getting hot...

Lightweight DIY sandwich, also about 115 degrees (having the aluminum on the correct side of the plywood helps!)

Warmboard, less than 95 degree water. That is not a typo.

now there is variance with floor coverings and such, but that should illustrate the point. For the price of quik trak, you're not getting the performance you could be getting for less, about the same, or a bit more money.

-=Northeast Radiant Technology=-
Radiant Design, Consultation, Parts Supply


Rockport Mechanical

HVAC Design and Installation in beautiful Rockport Maine