Click here to Find a Contractor in your area.
Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit

thermofin vs rti - bbob

bill_8 Member Posts: 15
I'm close to ordering heat transfer plates for installation on the underside of the subfloor. Thermofin and RTI seem to be the only manufactures out there for the plates. Any input on one vs the other? thanx


  • Art Pieterman
    Art Pieterman Member Posts: 2
    Heat transfer plates

    HeatLink makes their own heat transfer plates for under floor systems. Our customers really seem to like them because of the cost. We make them out of galvanized sheet metal, instead of aluminum. They measure 5" wide and 20" long. As a manufacturer, we have done extensive testing and have determined that the output from using galvanized plates vs. aluminum are very similar. I am not sure what the American list prices are, but in Canadian, they list out at $ .84 each. Quite a significant difference in price vs. aluminum. We have compared them to whatever else is out there on the market, and performance wise, we measure right up there with them. But with the major price difference, we make doing staple-up systems much more attractive to both the contractors and the end-user.

    I hope that this has been helpful to you, and if you would like more information, please feel free to email me directly.

    Art Pieterman
  • Duncan_2
    Duncan_2 Member Posts: 174
    Thermofin vs. RTI

    Thermofin: Heavier cross section means more heat conduction. Larger surface area of u channel in direct contact with tubing means more heat conduction. Both cross section and tubing surface contact relate to the A factor in the heat conduction equation Q = U x A x dT. It comes in eight foot lengths. It is not pre-drilled.

    RTI: Good cross section, not quite as heavy as Thermofin. Raised ridges in the u section that grips the tubing means not as much direct contact with tubing surface. Pre-drilled holes are convenient.

    I don't know how superior the performance of Thermofin would be over RTI, but the laws of physics are on its side. I suspect the weak link of either is how much of the plate surface actually contacts the subfloor. Make sure the PEX you buy is a good fit for the extrusions.
  • Troy_3
    Troy_3 Member Posts: 479
    thermofin vs. RTI

    We have used both. about 30,000 feet installed. We only use thermofin now. Much better engineered product, with better performance. It will cost a little more but it is worth it. Don't cheap out on a radiant job, it always comes back to bite ya! Just my humble opinion.
  • Mark Eatherton1
    Mark Eatherton1 Member Posts: 2,542
    RRCC results

    For what it's worth, here are the results of some preliminary testing we did at Red Rocks Community College.

    Good to see you Duncan.


    To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"
  • larry
    larry Member Posts: 91

    As a homeowner looking to have radiant installed in the best possible way, I've gotten samples of Thermofin & RTI and compared. I agree that physics is on the side of Thermofin. Took a micrometer to both products and found Thermofin to be about 15% thicker. The big problem with Thermofin is its complete lack of distribution in this area whereas RTI can be picked up at many local supply houses. I plan on getting the Thermofin inspite of its higher cost plus freight cost. This week a contractor showed me the double groove Wirsbo plates he uses. They seem pretty light compared to either Thermofin or RTI.

    Mark - those test results seem to suggest that RTI outperforms Thermofin when using PEX-AL-PEX. Did they test the RTI plates with PEX?

    Duncan - I find your suggestion about the weak link being the actual contact area between plate and the subfloor interesting. By analogy, when cooling electronics you must use thermal interface materials/compounds between the semiconductor and the heat sink to get good performance. The key is to keep the thermal compound as thin as possible and only fill in the microscopic areas between the pieces. If you use too much, you can actually degrade thermal performance. Overall, it seems like the same principal should apply with heat transfer plates. However, haven't heard of any thermal interface products out there for heat transfer plates, nor is it known what difference it would really make.
  • Mark Eatherton1
    Mark Eatherton1 Member Posts: 2,542

    What you see is what you get. We had a limited number of test bays and the students dictated the types of testing done. Maybe one of these days HR will get some ASHRAE monies together and we can test all of them. Sorry.


    To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"
  • Earthfire
    Earthfire Member Posts: 543

    Guys we are dealing with pex, (soft plastic , "DUH") the ridges will make little, if any difference, on the heat transfer . If staple up worked with the loose fit of the aluminium foil plates think , will the extruded plates work better? As for personal preference I would rather work with 4 foot plates then try and maneuver 8 footers in a crawl spaceor joist bay
  • heatboy
    heatboy Member Posts: 1,468
    Having used both.......

    I think either will perform adequately for you, Larry. I have used a couple of miles of extruded plates both below and above the floor and am very happy with their performance. Almost all of the major tubing manufacturers offer these in some form. I think they are finally realizing they are the ultimate in heat transfer.

    I haven't used any PEX-AL-PEX with them. For me, it is much too cumbersome a product too use when working under the floor. Any advantage in output over PEX is quickly lost in the additional labor actually pulling it through the joists. The same would apply to any kind of thermal compound applied to the plate before attaching it to the sub-floor. Messy, possible smell when heated? My opinion, only.


    To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"

    The Radiant Whisperer

    "The laws of physics will outweigh the laws of ecomomics every time."
  • Duncan_2
    Duncan_2 Member Posts: 174
    point of contact

    EARTHFIRE, I think the ridges make a big difference where conduction is the name of the game.

    Plastic tubing is the one element that's in common with ALL the above methods of installation.

    Assume the surface temperatures and u values (conductivities) of the tube wall are the same for each tubing.

    The next (thermal) bridge to cross then becomes the area of the tubing in contact with the fins. In this respect, the ridges DO make a difference. Halfing the "A" factor of the Q = U x A x dT equation halves the Q output. Roughly, in the real world.

    The if in "if the staple up worked with aluminum foil plates" is a very big if. While Mark's temperature measurements of plate output are revealing, they measure one dimensional degrees of temperature, not necessarily an accurate measure of real thermal loading. I applaud Mark's proactive exploration of real world panel output.

    To see which installation can meet the real load demand, run some water across the surface of the different plate installs. Measure the before and after water temperatures over a given time period for a given quantity of water, see which panel you can suck the most heat out of. My best guess is a tight fitting Thermofin installation, for all the physics reasons I've outlined above.
  • Larry Ticknor
    Larry Ticknor Member Posts: 38
    Can someone explain

    to me why these results are logical.

    The supposed difference between placement 1 and placement 7 is that "1" has PEX and 7 has "PAP". Since PAP has as good conductivity as PEX (essentially the same stuff from the conduction standpoint) why is placement 1 so much better? Placement 1 is 30% better with only PAP instead of PEX?

    And why is placement 2 so much better than placement 6. Copper pipe conducts heat better, so why isn't it distributing more heat from the water to the plate?

    And if extruded plates are clearly superior, why is the lightweight Wirsbo plate with 1/2" pex showing up better than the extruded plates with copper or PAP?

    Is this data to be believed, or are there many more things going on that don't show up and make the results appear illogical? Are there more important parameters than the tubing and plate?

    Thanks for any help.

    Larry Ticknor
  • larry
    larry Member Posts: 91

    Good points... Does PAP dimensionally expand less than PEX when hot water is flowing through it? From an amateurs standpoint it seem like the aluminum layer should no only provide an oxgyen barrier, but also make the tubing more dimensionally stable when heated. Assuming this is true, the PAP wouldn't "push against" the channel as tightly as PEX, and would thereby provide lower conduction to the plate. If this is what's happening, it supports Radiant Engineering's statements about the importance of their smooth gripping channel vs. the grooved channel of RTI.

    Of course this is still at odds with placement #5 vs. #7.
  • Duncan_2
    Duncan_2 Member Posts: 174
    EARTHFIRE, you are right. - Duncan

    I re-read my post, and I'm fairly off base about the ridge thing.

    Heat transfer from system water to finished floor surface encounters a string of resistances working against heat transfer. Water to tubing wall, tubing wall to transfer plate, through the transfer plate cross section, plate to subfloor, subfloor to finished floor, finished floor to occupant.

    In the grand scheme of things, in the real world, the ridges in the plates probably make little difference.

    When looking at the big picture, there are at least three larger factors limiting overall heat transfer. They are all much larger factors than ridges or no ridges in plates: The boundry layers of air between plates and subfloor, the boundry layers of air between subfloor and finished floor, and the final floor coverings.

    The ridges may limit heat travelling by conduction from tube to plate (my point), but any large advantage of one over the other would be snuffed out by the other limiting factors mentioned.

    Heatboy and Larry's observations relate to this.

    Sorry about that, EARTHFIRE, I had my blinders on.

  • Earthfire
    Earthfire Member Posts: 543

    what I'm thinking is that the plastic conforms to the ridges . Example of that phenomenon is the failures to activate of the Orbit (and others) sprinkler heads that had the o-ring conform to the machining groves in the head passages and not release the stem when it was necessary to activate. Softer ductile materials will conform to what ever is encasing it, especially in the presence of heat and pressure. As to whether the ridges help hold the tube in place ( due to increased friction ) during installation is subjective . The best product for the job is the one that you are comfortable with, and has some support from the people selling it to you
  • Tim Doran
    Tim Doran Member Posts: 208
    Hey Duncan

    Nice to see your post. I had been thinking about you everytime I watched the news. I was out in Denver last week and ME said that you had gone missing for a few days due to the fire. Hope all is well.

    Tim Doran
  • Duncan_2
    Duncan_2 Member Posts: 174

    As of 7:00 PM last night, we were able to go home again after being evacuated for two weeks.

    Very scary stuff! And the closest it got was two miles away from my house. That thing was HUGE!

    The wildfire fighters amaze me... that is incredibly demanding and dangerous work. God bless 'em!

    All is well in my little bubble of the world. Thanks for asking, Tim. And Dave.

    I'm thinking of the people in Durango Colorado and ShowLow Arizona now... It sucks to live in a state of perpetual anxiety. It's going to be a long summer out here.
  • Larry Ticknor
    Larry Ticknor Member Posts: 38
    Too much we don't know


    The dimensions I have seen for Kitec make it slightly larger in diameter than the PEX dimensions I have seen. Though I'm sure these dimensions change and any length of tube the PEX could be larger diameter than a lenght of PAP. I agree with Earthfire that I would expect some conformation of the tube to the ridges under pressure and temperature.

    You are right that the PAP is more dimensionally stable than the PEX, I don't know how much that would change the heat conductivity.

    The main thing causing me to scratch my head with this data is that there seem to be too many inconsistencies with what I think are logical ideas: PEX and PAP are essentially the same so interchanging them should not have a large effect on heat output; thicker extruded plates should do better than thin pressed plates; increasing the conduction of heat from the water to the tube (by using copper tube instead of PEX) should increase heat output. The data seems to contradict all these.

    Below, Duncan talks about the boundary layers between the plates and the subfloor, and the boundary
    layers between the subfloor and the finished floor. Maybe there are boundary layers within the subfloor that
    cause some areas to transmit more heat upward? Maybe the largest variable for effectiveness of
    the plates may not be the plate, but the boundary layer between the plate and the tube and the boundary layer between the plate and the subfloor. This is just the idea you pose with your electronic thermal interface materials.

    Maybe thin plates that conform to the subfloor irregularities more closely work better than thicker plates that bridge some of the irrregularities? Maybe thicker plates with a thermal interface goo would do better than anything else or maybe thick and thin plates with thermal interface goo do the same. Maybe thicker plates without goo do as well as thin plates with goo and thicker plates without goo do better (on the average) than thin plates without goo.

    I guess it just seems to me there are many more questions than answers and the data gives us more questions but doesn't really answer any questions. There is a huge research potential here, but probably no funds to do the work. You would need lots of replicates and/or ways to minimize the variations between other factors not being studied.

    Until the study is done, what should the installer use? Are the thicker plates worth the extra cost? I would argue that the data is inconclusive, so how do you justify the extra cost in your house Larry?

    Just thinking out loud,

    Larry Ticknor
  • Earthfire
    Earthfire Member Posts: 543

    are the Wirsbo extruded plates out yet? so we can add their idiosicracies to the discussion.
  • larry
    larry Member Posts: 91
    Justify costs...

    Larry T,

    That is a very good question. With a lack of hard conclusive data, I'm following my gut (supported by a little physics) and the opinions of both local contractors and folks on this site. Since I'm going to work with a contractor who will allow me to prep the site and put up plates per their system design, I'm trading off some of my labor against the increased cost of Thermofin plates. The fact that RTI recommends caulk in the channel to keep the PEX from falling out sold me on Thermofin. Besides, caulking isn't a great conductor and over a number of years will loose its effectiveness.

    With regard to thin plates molding to the surface better than thick plates... I not sure I buy it. From looking at the Wirsbo double groove plates I saw, I'd guess the act of pushing a staple through the plate would create an area of very close contact at the staple, and the plate would deform away from the floor slightly until coming to the next staple point. Just trying to imagine how these go up - might be all wrong on this.

    Thermal goo is a interesting notion, but probably impractical from an installation standpoint. Maybe a light handed application of a belt sander where the tracks are to be installed would help. Of course all this working with tools above your head is tough to begin with without adding more steps.

    A while back someone mentioned that some aspects of this question are being researched by the RPA and ASHRAE.
  • Duncan
    Duncan Member Posts: 43
    area of tubing contact vs thin X section

    Larry, you said: "With regard to thin plates molding to the surface better than thick plates... I not sure I buy it"

    You are right on the money. It doesn't even matter how closely a foil plate conforms to either tubing or floor surface if the thin metal can't carry the heat. The molding-to-the-surface isn't the issue here, the area of the cross section is.

    There has to be a fairly substantial cross section, or you can't conduct any decent amount of heat.

    Consider wood stove door handles. Ever see the kind that has a spiral coil of wire around it? Yeah, it gets warm, but not too warm to put your hand on it. It's probably too hot to touch right near where it joins the iron part of the handle, but the heat is dissipated by the middle of the handle. That's because all the heat moving by conduction through the handle has to be transmitted through the small cross section of the wire coil. Something like an eight or ten gauge wire?

    Hold one end of a one foot long, one inch diameter aluminum bar in a flame, and see how long you can hold onto it. Now hold one end of a one foot long piece of aluminum flashing in a flame and see how long you can hold onto it.

    I'm at a loss to explain Mark's results myself. But there's a lot more to heat transfer than temperature alone.
  • Larry Ticknor
    Larry Ticknor Member Posts: 38
    Thanks Duncan

    Thanks for taking the time to educate.

    I keep thinking about your rod vs.flashing example. I was thinking about a rod and a piece of flashing, with
    identical amounts of material. The rod will become too hot to handle -- it's conducting more heat to the end you are holding. The flashing you will still be able to hold onto. My question comes from thinking about the surface area of each of these -- the flashing has a lot more surface area to transfer the heat to the subfloor, while the rod can't transfer as much heat to the subfloor since it has a smaller surface area. The rod conducts more heat but the heat stays in the rod -- (I don't know this -- I'm just guessing, so tell me where I go wrong). There seems to be a trade-off between surface area and cross-sectional area.

    If I increase the cross sectional area without increasing the surface area at some point I just have a hot plate and little heat transferred to the subfloor. The thin plates may not be thick enough, but when do plates become too thick so that your are just wasting heat by dumping heat into plates faster than it can be transferred to the subfloor?

    I don't know if any of this makes any sense. I'm just curious about what makes these things work and how to maximize the potential of a good job (from the heat aspect).

    Ducnan, glad to hear you house was untouched.

    Thanks again,

    Larry Ticknor
  • Mike Kraft
    Mike Kraft Member Posts: 406
    ridges help hold

    One thing that your missing.........water weighs and in a joist bay install the tube is facing down.And lets not forget Mr. Newton and his apples.........gravity.The tube will have certain rules of physics it has to obey.In a loose fitting plate I dont see the ridges and the pex working together to become more intimate.My minds eye sees the opposite ..........which would be flop out.I have heard of this happening to some contractors.That is why the manufacturers reccomend using silicone........or they did.

    With Rad.Eng.plates the fit is (to quote Forrest)......Peas and Carrots :) Tight and intimate.The tube has built in this will stop em from flopping power.The other day I was visiting another contractors shop.He had just recieved some cases of Wirsbos plates.I had samples of RTI"s and Rad Eng. C fin.The difference is quite evident in a not very upclose look.T-fin is tight!

    Nice thread.............cheese
  • Mark Eatherton1
    Mark Eatherton1 Member Posts: 2,542
    Wirsbo plates...

    are from the same source as RTI from my understanding.

    I can't tell anyone why there are differences that there are in the output data. The copper pipe did fit loosely in the plate. I'm sure that had something to do with its' poor performance..

    The plate from Montana is a superior plate. I will admit that. You get what you pay for. As they say in Austria, "Nothing is free..."

    In reality, the lowest output would probably work quite well as long as it was set up for continuous circ w/ outdoor reset. I think you also have to look at the fact that these readings were taken only one hour after the plates had been running. If you had gone back and checked them 3 hours later they probably all would have stabalized near each other. Once the thermal mass battery has made its major initial draw, the thermal conductivity of the surfaces as a whole start backing up, and the whole assembly gets hot. Carpets compound this effect.

    You'd be suprised at how low most heating loads actually are in comparison to their theoretical loss calculations.

    I'd still recomend that the manufacturers directions be followed, and use their outputs, not these.

    Here's some more difinitve information about the test bed we were working with.


    To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"
  • Larry Ticknor
    Larry Ticknor Member Posts: 38
    Stabilized ...


    Thanks for that reply and using the word "stabilized". I realize that's what's been bugging me. I think I've missed the simple idea again.

    If you put up a plate solid against a subfloor, then pack insulation up tight against it, it seems to me the temperature in the plate will eventually "stabilize" and the surface area may make more of a difference then the cross sectional area after the temperature has "stabilized"... Maybe plate thickness affects other response characteristics of the radiant floor.

    So it seems like you need to look at the heat ouput on top of the flooring through several cycles of different room heat losses and see how the surface reacts with the different plates underneath them. If the plate is not tight against the subfloor and/or the tube doesn't fit tight to the plate, maybe only response time is diminished or maybe response time and the maximum btu's available suffer ... Anyway, it seems like you need to look at the output of the whole system and not just look at the characteristics of the plate. (Which is what you are already doing ...).

    I think it's time to return to a non-plate life for awhile. There are lots of tools to get the job done.

    Mark, thanks again for the response.

    Larry Ticknor
  • jfox
    jfox Member Posts: 44

    Thanks from me too. I am going to start installing tubing next week.

    My thinking has been that if I am delivering X BTUs/ hour to an insulated space, it does not matter whether I am heating a volume of air first and then the subflooring and then the flooring or I am conductively heating the subflooring first and then the flooring. All of the heat/energy is under the subfloor and the whole mass will reach an equilibrium at about the same temperature. Plates will transfer a change in temperature faster but with constant circulation and outdoor reset that should not not be a problem.

    The assumption is that the change in temperature between the supply manifold and the return manifold will be the same with and without plates, i. e. that both systems will deliver X BTUs/ hour. Can anyone offer any insight into whether this is a good assumption?

    Thanks, again, to all.

  • Troy_3
    Troy_3 Member Posts: 479

    If that was the case why ever use commercial fin tube? Your reasoning says it all puts off the same btus eventually. There is our problem-(eventually). By the time you get enough btus out of plain tube in my geographic area- more btus have been lost than gained from the tube. That's why we add another perameter- time. Btu/hour. There is no way you can think that tube alone will transmit the same btu/hr. without an extruded plate that it will with a plate. You are adding surface area. The key is achieveing the design btu/hr. output at as low water temp. as possible.
  • hr
    hr Member Posts: 6,106
    It's an efficiency game, John

    When heating a home there are a number of efficiency trade offs. Start with the most important, in my opinion. Building efficiency. Obviously an uninsulated leaky old house would be less efficient than a new, tight, well insulated. Spend as much money as you can stand on the building, first. Windows, insulation weatherstrips, etc. An investment that will keep on giving.
    Next comes boiler efficiency. Once again the more efficient equipment will cost less to convert the chemical energy (gas, oil, etc) into useable heat. Less waste up the flue.
    Next is the transfer efficiency. Ah the beauty of using water as a heat transfer medium!
    This is the feature of transfer plates. On a given design day your house may require 25 btus / sq. ft. to maintain a given indoor temperature. If your radiant distribution can only supply 20 btu/ sq. ft. the building may not maintain a comfortable temperature. Yeah it will catch up, when it warms up outside!
    The plates merely increase the heat transfer efficiency. Which translates, generally, into lower supply temperatures.
    Perfect operating conditions for high efficiency condensing equipment, usually :)
    Tight building, low water temperatures distribution, and condensing heat source. Short of a total solarly heated home this would be my best choice.

    hot rod

    To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"
  • jfox
    jfox Member Posts: 44

    Hot rod and Troy. Both your replies make sense.

    The house is new construction, very tight and well insulated, and I am going to use a condensing/ modulating boiler with as much brains built in it as I can find, so I guess I am over half way there:)

    I need 17.5 BTU/ sq ft to meet my design day load so I am in a position where no plates, with reflective foil and insulation below, will meet the design load. I will have a hydronic coil in the Unico system for below design days.

    At this point, what does using plates give me. The higher boiler efficiency of sending lower temperature water back to the boiler is a given but the pay back seems dubious given the high installation costs and the high efficiency of the boiler already. Do plates allow me to deliver the design BTUs/hr with lower water temperature?

  • jfox
    jfox Member Posts: 44

    Hot rod and Troy. Your answers make a lot of sense.

    Do installations with plates have hot spots, or cold spots? If so, are there any fixes?

    My first floor is all tile, slate and hardwood, and socks and bare feet will be the norm.


  • Troy_3
    Troy_3 Member Posts: 479

    The plates really help the heat stripeing by spreading the heat across a 4" area. We have not had a problem with hot/cold floors. We install 8" on center. The only issue is you sometimes can't get under areas. also it is more difficult to seperate rooms when installing from the underside.
This discussion has been closed.


It looks like you're new here. If you want to get involved, click one of these buttons!