Kitec staple up - SteveMurphy

eleft_4

has a problem. 16in spaced joists are only 14 1/2" apart, when you need to feed more tube to loop it up the joist bay it needs to be twisted to start the loop. This very tough on the hands and time. That's why the loop hanging down in each bay as you feed it out and back to the manifold return side. Note it has the twist in it and is ready to pull down the bay, feed more or remove slak and staple up.

Al

Steve M

Staple up questions Steve M

I need to know if anyone is stapling up Kitec PEX-AL-PEX out there in Wallie land?? If so, I have the following questions:
What size do you use?
How do you attach it to the sub-floor?
What is the best way to install it through the joists?
How does it compare in expansion to regular PEX?
Do you use plates?
How's the transfer in comparison to Onix?
Do you prefer PEX-AL-PEX or Onix?
Thanks for any responses. If possible, I would like to keep the aluminum plate debates to a minimum, please. I'm more interested in the installation of the tubing itself.

Canuck

staple-up

We typically use 1/2" stapled to the floor joists as close to the sub-floor as possible. We then staple 'bubble/foil' insulation to the joists, at least 2" below the Kitec, to create a hot cavity. (This is actually Ipex's recommended procedure - at least here in Canada). The feed through the joists is always a chore. The best procedure we've found so far is to feed a large loop (I mean LOTS) through the first joist, then reduce that amount by feeding a second loop into the next joist. Keep going till you're done. If you need to provide more tube to carry on, repeat the whole procedure.Once all your loops are fed - start the stapling.
It may sound cluttery but we find that things move more quickly and are less likely to get kinked.

steve gates

much

better in the noise dept. w/Kitec because of the expansion. I have yet to have expansion noise with Kitec. Have had it with pex. I use 1/2" though am looking at the use of 3/8". Guys I've talked to say it is much easier to pull. Kitec is all I use unless allowed to charge more for pex. I find it harder to pull than Kitec(although esier to make the turns). I use 1"wide x 11/4"staples when not using plates(with an air gun/hand opertated is for the birds)

heatboy

If I were ever to...............

install joist bay heating without extruded plates (not that I would often consider it), I would opt for Wirsbo Multicor over Pex. My reasoning would be that with the high water temps needed without plates does not lend itself to regular Pex due to the expansion/contraction, possible noise, etc. I have started to do more 3/8" tube under the floor because of installation speed more than anything else, especially on 12" O.C. joists. I would approach actual installation the same way by getting all of the loops in their perspective bays, then pulling them out starting from the furthest loop. This method leads to much less of a chance of kinking the tube.

hb

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

Mark Hunt

I've used both 1/2" and 3/8". I would HIGHLY recommend the 3/8".

My wife had to take the children away from the house when I stapled up 1/2" in my own home. Don't want the little ones to know those words!!!

I was suprised by the lack of expansion noise. None at all!!

Good product and good people.

Hope this helps!!

Mark H

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

eleft_4

eleft_4

eleft_4

I used W/M alumipex. Not to bad if you start out with large loops between the joists and keep them large as you feed the tubing off the spool to the manifold. When you staple up and remove the slak keep the feed loops large also.

Duncan_2

Kitec joist installation.

Kitec may be a little harder than plain old PEX, but let's face it, pulling plain old PEX is a pain too. I dunno where the rumor started that PEX-AL-PEX is so difficult. The little extra attention Kitec takes is no big deal, in my experience.

This technique was recommended to me by Dennis Bellanti, who's now employed by Ferguson in Denver. It works. Maybe drawing a picture as it's explained will make it easier to understand.

From the boiler room, drill PAIRS of 1-3/8" holes to the joist space at the farthest end of the room you're doing.

Start with the roll of Kitec in the boiler room, and leave it there. You're stripping Kitec off this roll that stays in the boiler room.

Thread the free end of the roll through one set of holes to the farthest joist space. Let a long loop of Kitec hang down from that last joist space.

Make a U turn and thread the free end back to the boiler room through the other set of holes you drilled. Feed as much as you need from the roll in the boiler room as you go.

So now you've got both the roll of Kitec and the free end of the Kitec in the boiler room, with a big U turn in the last joist bay. Fasten the free end to something in the boiler room to keep it in place.

STARTING FROM THE FARTHEST JOIST BAY, pull loops of Kitec down the length of each bay, just letting each loop of tubing hang down freely. As you're doing this, you're feeding it from the roll in the boiler room.

Now you've got a mess of Kitec loops hanging all over the place. I think Kitec's shape-holding property is actually an advantage here. The way PEX snaps back up into coils, I think it would've been harder using PEX. Anyway, this pretty well takes up the whole room and other trades can't work in the area.

Fine tune the length to each bay, pulling or pushing to or from the roll in the boiler room as you fasten each loop into its final position in each joist space.

When you're done, cut the roll in the boiler room where you need to.

I did a three or four loop room with 250 foot loops of 1/2" Kitec this way. While it was a pain for one person involving ladders and planks and lots of running back and forth, it wasn't bad. 3/8" is probably easier, but then loops would have to be shorter.

Except for a couple tight spaces, paying attention was all that was necessary to avoid kinks. In the tight spaces, patience, extra attention and common sense was all that was needed. Stapleup just takes a lot more work than staple down. Bottom line is, if you kink it, cut out the kink and use a union and air test.

One advantage of Kitec is that it won't sag or droop when it heats up. I attach it to the subfloor using flat mount plastic pipe clamps (like the Mickey Mouse type, but flat bottomed) and zip-in screws with a cordless.

Nothing beats heavy aluminum plates for output. I'm absolutely in heatboy's corner on this one. I wouldn't recommend stapleup for anything but interior rooms that have almost no heat loss, or rooms that have supplemental heat. There is evidence that stapleup can cost more to operate than conventional heating. If you can meet the output needed to heat a room with only half the floor area's worth of plates, use plates near exterior walls, and suspended towards the interior of the room.

I wouldn't recommend Onix or any other rubber tubing product for heating.

Sorry about the length of this post, hope it's clear, and doesn't sound like a Kitec commercial. I like the stuff for residential. For commercial pours, I'll stick with plain PEX.

Duncan

steve gates

rolling platform

I've started using a 5'x2' adjustable rolling platform on my jobs. BIG help beats ladders when your alone. You can roll thru doors not climb up and down as much ect. Just keep the floor clean and off you go. Great for a work table too.

Steve M

Thanks for the responses,

I have some ammo now.

Duncan_2

Hanging loops

Eleft, that's the ticket, and there's a variation I like. Starting with the bay farthest from the roll... You can pull out loops completely to fill each bay as you go, rather than pulling a bunch of smaller loops out first.

That way, you're stripping pex-al-pex ONLY from the roll in the boiler room, and not through a series of small loops.

With a bunch of loops hanging down, you have to strip pex-al-pex from the preceding bay, then the next, continually feeding each small loop into the next bay, rather than feeding from the boiler room only, to the farthest bay, working back towards the boiler room.

The cross bracing actually makes it easier, it helps the installer support the loops.

It's been a while, I may be forgetting some details.

Duncan_2

Filling each bay as you go.

Filling each bay as you go, rather than pulling a bunch of smaller loops into each bay first, is an idea that I think saves a little work (but it's been a while since I've done a joist job). Here's why I think so...

If you pull the first loop COMPLETELY out to fill the FARTHEST bay, you're stripping directly from the roll in the boiler room only. When you get to your next bay (next closest to the boiler room), you're doing the same thing, stripping only from the roll.

If there's a bunch of smaller loops hanging down, you're stripping from a smaller loop hanging the next bay over. That means you have to feed (from the roll in the boiler room) through a series of small loops to get to the bay you're filling, rather than feeding from one place (the roll in the boiler room).

Like I say, it's been a while, and I think I started the same way, doing small loops at a time. Eventually, I just went for filling one bay, completely, at a time, if I remember correctly. Hope this makes sense.

heatboy

I give up.

Duncan,

I have been sitting here drawing so many circles my kids are asking me if I feel OK. I can't figure out how you can pull the loops from the uncoiler without looping. I can do that if I am dropping under the joists, but I just can't figure it when I have drilled joists. Help me, oh tinned one!

hb

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

Duncan_2

In the voice of the Tin Man

If I only had a scanner!

I knew a verbal description would be tough.

Put simply, start with a double tubing run to the boiler room, a supply and a return.

Pull one bay at a time, from the roll in the boiler room.

Just like eleft shows, but skip the step of pulling short loops down.

Pull 'em once, full length of the joist bay. Pull loops in the bay farthest from the boiler room.

Richard Miller_2

I use a lot of 3/8\" Kitec.

I use the clips from Peter Mangone and the nifty new air gun to put them in. Have been very happy so far. I install it exactly like Duncan mentioned.

eleft_4

Pull 'em once, that's a no no!

Duncan,
You need to pull the return back to the manifold first, thru the other hole. To do this you need to feed the loops from the uncoiler and last bay then straight thru the joists to the return manifold. After you hook up to the return manifold, go back to the last bay and pull the loop down to the end of the bay. To do this you need to work the tubing from the uncoiler thru all the hanging loops to get enough tube to make the loop. Now,in the last bay,start stapling plates on the return side to the end of the bay removing the slak. Make the bend at the end and staple up plates back on the feed side tube. Repeat this in each bay back to the feed manifold And connect. Tip: turn the loops like a steering wheel from the first bay to the last,pulling off the uncoiler. It goes smooth when you get going.
al

hr

The Duncan method

is what I use also. Fasten one end onto the manifold and write down the footage mark. Next pull one bay at a time and fasten at least the loop end so it stays there.
Then go to the next bay and do the same thing. I watch the footage marks on the tube and make sure I get back to the manifold within the loop length requirement. Generally 200 - 225 for 3/8".

One thing I do differently from Duncans method is I drill only one 1-3/4" hole in each joist. I can easily pull up to four 3/8" pex runs through a single hole. Cuts down on the drilling. Very important for the old "rock hard" oak joists found around here

I have never tried the multiple loop method, looks like a lot of adjusting until you get the hang of it I suppose.

hot rod

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

Duncan_2

Why is that ?

Absolutely, you have to run the return all the way back to the manifold, and fasten it to something, like the manifold itself - no question. I tried to make that clear, but it's difficult to explain with no drawings (don't have access to a scanner right now).

As for the series of hanging loops, I remember doing my first couple loops like that, eleft, and I'm trying to remember why I did it like that. You can tell I don't do a lot of joist installs, right? ;-) Maybe it was a question of slight bends and ripples in the tubing rubbing against a long run of drilled holes, making it too hard to pull through a lot of holes at once?

But then I figured, why not skip the series of loops and just do one loop at a time, always pulling a complete loop at the farthest bay? If the holes are drilled in line, the tubing is reasonably smooth and straight, and you can strip directly from the roll that's all the way back in the boiler room, why not? Why not do one loop at a time, instead of uncoiling through a whole series of hanging loops?

It's been a few years, remind me why not (seriously!). Is it a question of making it easier to pull a little more than you actually need for each loop being fastened as you staple up? Does having a little hanging loop in the next bay make it easier to feed back the little excess from the preceding bay? Is it a question of kinking it at the first hole out of the boiler room?

I figure the less you work the tube, the less wavy it gets. Not that it gets work-hardened like copper, but just that smoother is easier. Not criticizing your work (smooth loops you got going, obviously), just trying to save some steps.

eleft_4

The Duncan method

hr,

looking at the drawing, you need to twist the tube to start it back down the bay with 1/2" because the radius's (3)at once when you try to pull thru 14 1/2" space will kink.
I did this (Duncans method) on my first try, to much sweat and sware. It took all day, after elongating the holes, to pull one room.


My questions, how do you pull 225' off and where do you put this wild mess? Why not leave it on the spool? You are not feeding the loose end thru.

By the way, I work Alone so, no one hears me.

al

heatboy

I got it! I got it!

Well, I never claimed to be the brightest light on the tree. You pull your tube the same way, just not with the loops. Whew! I can rest now ;-) When using 3/8" tube, I do the same thing since there is little chance of kinking the tube when you twist it to get the loop. If I'm forced to do 1/2" I choose to loop the bays first. How's that?

hb

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

Bill Clinton

depends on the tube

I'm puzzled by all the different responses. I've been doing suspended tube (never called it "staple-up" for reasons I'll go into) for a dozen years or more--always very conservative on performance expectations and never unsatisfactory results.

I've found that pulling out individual loops as described by hot rod et al is the usually the easiest thing with 1/2" or 3/8" pex. HOWEVER, with stiff tube (radiation cross-linked) it becomes difficult to avoid kinking. I can't imagine how you could do this job with Kitec and not get kinks. Probably Duncan saw the same problem, hence the multiple loops. Of course, I've never tried it with Kitec, so maybe I'm wrong.

Also, I can't for the life of me understand why people are STAPLING up tube. I suppose the thought is to try to get as much contact with the floor as possible for direct conduction, but the obvious truth is that the contact patch approaches zero and the dominant heat transfer mode goes through the convection path. Given that, one would want to maximize convection by making it easy for air to circulate around the tube: Suspend that tube.

Also, I keep hearing concerns about the expansion noise
as tube rubs against staples; about tube migrating around, and so on. That seems to be caused by staples. Further, isn't there a long term risk involved in the high point stresses of tube against narrow staples?

Our standard practice is to suspend the tube about 1" below the floor by attaching to the side of the joist with full circle plastic nail clamps. We currently use 3/4" nail clamps for 1/2" tube, specifically to allow free movement without binding. We also put plastic isolators in every hole tube passes through--again using a 3/4" size for 1/2" tube. Result: Silence. No creaks, no pops, nada.

An interesting side-light that had to get pounded into my skull numerous times before I would learn it is that it is best to use smaller tube if flow rate and length allow. 3/8" tube will give off as much heat per lineal foot as 1/2" or even 3/4" because the lower thermal resistance of it's thinner wall offsets the incrase in surface area. Think small.

Bill

hr

You R right to clarify, Bill

To my way of thinking staple up, directly and tightly to the sub-floor, only works with rubber as it doesn't move in the staples. Anything with clips or brackets allowing an air space, in my mind, is suspended tube.

I agree suspended tube will work fine, IF you understand and are comfortable with the limitations. The plate manufactures will also agree with that opinion.

Suspended tube, obviously, is not for everyone Although that too may change.

hot rod

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

Duncan_2

All good points - the bigger picture, and a rant.

Discussing only technique narrowed discussion to one aspect of the bigger picture.

Suspended is the right term. Stapling PEX tightly to the subfloor is flat out wrong, both as a method of installing PEX and in naming a technique. I will never forget a contractor describing the noise a stapleup with PEX makes as it expands and contracts, and the nightmares he had to deal with.

All good points, Bill. Sleeving where tubing passes through ANY holes; terminology - suspended vs. stapleup; tubing size vs. output; that the contact area of true stapleup is nil.

Also, attention to expansion loops to avoid noise is important. I'm also thinking part of my reason for initially using a series of small loops before I went to the "farthest-joist-single-loop-method" may have been because, in my inexperience with suspended, I didn't drill large enough feed holes through the joists. Hole size isn't much of a problem with TJIs BCIs, etc.

In one of the couple ssuspended job I did, I used Wirsbo crosspieces to rest the tubing on. They look like the telescoping arms you see on lighting cans, the ones that you pull out to any length, with points on the ends to hammer into the joists. I'm not sure they're even made anymore. These crosspieces allow you to hang the tubing any distance you like from the deck OR from the joists. At least they do with PEX-AL-PEX, which holds its shape. Maybe even plain ol' PEX if you use wire ties, but that's prolly not practical. Incidentally, I used heavy plates along the perimeters of that "suspended" job to meet the heating load, which can be large in this climate.

And now onto the bigger picture...

Perfomance expectations of suspended's output are generally acknowledged. Though the exact number may not be agreed upon, I think the Kansas study pretty much tells the story about output.

System efficiency is another aspect of performance we need to look at with systems that rely on convection as a vehicle to transmit heat.

The potential savings of 20% or more often promoted as an advantage of radiant are probably not usually there with suspended. If the heat that's produced has nowhere to go, if it's not transferred efficiently, fuel savings are not likely to be realized.

Think about it... baseboard heat is designed aroung convection. Its emitters -the finned tubes- have a huge surface area, and are surrounded by 65 degree air. I'm talking about room air that is in direct contact with the heat emitter.

By contrast, suspended tubing is a small surface area, it's surrounded by hotter air in the joist space (which lessens the delta T of heat transfer). Not only that... this hot air heating medium (with a lousy btu content) is insulated from the room air by subfloor and finish floor covering.

How can it be more efficient than baseboard, or even forced air, fuel consumption-wise? It can't get rid of the heat produced by your heat source!

Not only that... we haven't even considered the higher parasitic losses associated with the higher temperatures of suspended.

If the heat produced cannot be efficiently transmitted, then where does it go? Back to the boiler! Voila! Short cycling and even lower efficiency!

The anecdotal evidence of higher operating cost is out there. At this point, it's new research waiting to happen.

I'm not arguing that I'm not comfortable in rooms heated by suspended, the comfort is definitely there. What I am uncomfortable with, are blanket claims that any form of radiant heating saves money.

This is why I'm in heatboy's corner when it comes to suspended and stapleup. I don't think I'm a heating snob, though. Maybe it's more of a perfectioist or a Zen thing: it's not the destination, it's the journey. Of the heat, that is.

jfox

Perfect thread for a couple more questions

In a suspended (not stapled) hepex installation, 3/8" tube on 8" centers (no plates), does anyone have an opinion on how far apart to space the clips and how far away from the rimboard to keep the turns? (the leaders are in the middle of the joists and I am pulling the loops to the outside walls).

Thanks

eleft_4

convectionradiation all in one ???

This topic could cause paralysis by over analysis.

Convection: the transmission of heat by the movement of heated particles, as in gas or liquid currents.

Radiation: the process of which in the form of rays of light, heat etc. is sent out from atoms and molecules as they undergo internal changes.

I think Apples vs. Oranges?

al

Bill Clinton

I'll concede one point re efficiency:

The suspended tube method will have greater downward heat losses unless extra insulation measures are taken--which they usually aren't. This is not a problem if there is a conditioned space below as the downward losses become useful heat. With a ventilated crawl space below, a considerable amount of heat is wasted and you have to factor that into the decision process. How much loss, I don't know.

That consideration aside, the suspended tube method should be about as efficient as other methods. True, you have to run higher temperatures at the boiler, but most boilers have to be run at those temperatures anyway to prevent condensation. Short cycling is not a problem of heat delivery method; its a problem of boiler sizing.

Suspended tube has definite drawbacks, but in the proper application it's a good heating solution. I'd just like to see it understood and done right.

Thanks for your response, Duncan, and you are right on that "system" efficiency is the big picture to be attended to.

Bill

eleft_4

Post quoted from cheese


Duncan said,
> "The suspended tube method will have greater

> downward heat losses unless extra insulation

> measures are taken--which they usually aren't.

> This is not a problem if there is a conditioned

> space below as the downward losses become useful

> heat." Date: August 09, 2002 10:25 PM


"Author: Mike Kraft (thebigcheese@pikeonline.net)
Subject: Wirsbo sez.............



(from the CDAM)Suspended Floors: Downward loss exists in areas with a heated space below.If the heated area below uses the same heat plant as the area above,the loss does not increase the total load/heat plant load because the heat is not lost to the structure.If downward loss to the heated space below exceeds either the upward load or 10BTUH/sqaure foot,insulate the suspended floor.Without insulation the room temerature below is impossible to control.

From Dans description of the kitchen he's warming the floor.Not heating the room.The mechanicals below sound pretty warm with a 400K boiler not to mention DHW,piping etc.Ultimately your concern will be noted.I'm sure the numbers dont lie:)

Keep us posted Dan.

very curious cheese "






al

Duncan_2

efficiencies

Bill, you said: "[apart from downward losses]...suspended tube method should be about as efficient as other methods"

I think not. Maybe someone has some information to offer on this point. But the stories I've been hearing indicate large fuel bills, sometimes larger than forced air, and at a minimum, scant savings, if any.

Admittedly, there are many factors involved. But when I think about a heat source that has the efficiency of its heat emitters stifled as much as possible, images come to mind. Imagine putting a radiator in a box constructed like a floor system, and letting that box heat the room rather than the radiator itself. This is essentially what one is doing when one heats a joist space rather than a slab. The inefficient mode of the convection step gets in the way of efficient heat transmisson.

Informal observations of outdoor reset (with other types of emitters) have been done that indicate parasitic piping losses may account for more for system inefficiency than they're given credit for. So higher running temperatures (not boiler temperatures) may be more of a factor than we think.

I'm not saying higher boiler temperatures are inefficient though. I'm saying produced heat that is not delivered is inefficient, (heat that is run right back to the boiler).

If you get a large delta T across a (relatively) high temperature boiler and bypass, you're still extracting the heat produced. This is fairly easy to do with lower system temperatures. And temperatures of these boilers don't always need to run in the 180 degree range. Definitely not as efficient as a condensing heat source, but very likely a boiler/bypass setup in a slab system could cost less to operate than a condensing source in a suspended system. Heat emitters are the delivery point of the system.

There is no question in my mind that suspended tube makes comfortable floors, I installed floor tempering this way in one room of my home out of curiosity. I have no illusions: suspended will continue to be popular as an installation method. Unfortunately, this is based on installation cost considerations and not on system efficiency.

I guess sometimes the glowing claims of efficiency and lower fuel bills with radiant hack me. Been down this road before.

Thanks for your considerate response, Bill. I re-read my previous post and thought that I may have gotten carried away a bit. I do realize it's not a perfect world where every system has a condensing heat source and purely conductive/radiative emitters. Hey, as long as we keep asking questions and keep exploring the options.

Duncan

Duncan_2

Hey, wait a minute! I didn't say that!

> Duncan said, > "The suspended tube method

> will have greater _BR_ > downward heat losses

> unless extra insulation _BR_ > measures are

> taken--which they usually aren't. _BR_ > This

> is not a problem if there is a conditioned

> _BR_ > space below as the downward losses

> become useful _BR_ > heat." Date: August 09,

> 2002 10:25 PM

>

> "Author: Mike Kraft

> (thebigcheese@pikeonline.net) Subject: Wirsbo

> sez.............

>

>

>

> (from the CDAM)Suspended

> Floors: Downward loss exists in areas with a

> heated space below.If the heated area below uses

> the same heat plant as the area above,the loss

> does not increase the total load/heat plant load

> because the heat is not lost to the structure.If

> downward loss to the heated space below exceeds

> either the upward load or 10BTUH/sqaure

> foot,insulate the suspended floor.Without

> insulation the room temerature below is

> impossible to control.

>

> From Dans description

> of the kitchen he's warming the floor.Not heating

> the room.The mechanicals below sound pretty warm

> with a 400K boiler not to mention DHW,piping

> etc.Ultimately your concern will be noted.I'm

> sure the numbers dont lie:)

>

> Keep us posted

> Dan.

>

> very curious cheese "

>

>

>

>

>

> al

Duncan_2

Actually, I didn't say that...

But that's OK. The quote should probably read:

"If the heated area below uses the same heat plant as the area above,the loss does not increase the total load/heat plant load because the heat is not lost FROM the structure"

In other words, they're saying that any downward losses are not really losses, because they're still heating the building envelope. That's my take, anyway.

Kinda like the DOE efficiency calculations.

eleft_4

oops, sorry Duncan

It was Bill Clintons post I quoted from.

The line in cheese's post that is important here is: Without insulation the room temperature below is impossible to control.

Been there done that. The basement was very uncomfortable before insulating the bays, even at 95* water temp. Kind of broiler effect

al

Mike T., Swampeast MO

The morass

In this thing called "radiant" heating it's easy to forget that the space is ultimately heated by an extremely high portion of radiation. As long as the panel is sufficiently large and at a sufficiently low temp, convective currents do not reinforce themselves. Convection is still occurring, but it's essentially unnoticeable. The lack of temperature stratification with very high ceilings is testament to this fact. The actual proportion of radiant to convective heat is highly debated and likely indeterminate in any but a perfectly controlled lab situation.

Since the ultimate heat source to the structure is constant (the panel) the key is getting heat to the panel as efficiently as possible.

The earliest method, building a fire directly underneath a stone floor, was also extremely efficient. Reconstruction of a hypocaust system yields efficiency in the 90% range. Since there was no intermediate heat exchanger, the vast majority of the heat from the fire was available to heat the floor via all three method of heat transfer. Since the panel remained relatively cool in comparison with the fire, nearly all of the available heat was absorbed by the structure. The structure itself was the flue, so flue loss in the current sense was non-existent.

The next evolution of heat invariably involved less direct means--massive masonry "fireplaces" or free-standing iron stoves. Fireplaces suffered from terribly inefficient operation as the majority of the available heat went up the dedicated flue. The initial convection of the heat source was a near total loss. The space was heated by a high percentage of high-intensity radiation. This was much less efficient than the old hypocaust. It was also uncomfortable unless you were on a rotating spit close to the fire.

Free-standing iron stoves were a definite improvement as convection could now occur freely around the stove. Flue losses were reduced by controlling the rate of combustion and modifying the shape of the combustion chamber to expose a maximum of stove area to the heat of the fire. Unfortunately the intense heat of the iron stove resulted in intense radiation and high convective currents. The heat zoomed (and tended to stay at) the ceiling. Those near the stove melted from the intense radiation while those far away froze from the lack of radiation.

Note that the first heating systems were MUCH more comfortable and efficient due to the reliance on LOW INTENSITY radiation. The later suffered because of their reliance on HIGH INTENSITY radiation.

To combat the comfort problems caused by the intense radiation and rapid convection, the next revolution in heating involved various methods to "split" the heat and reduce its intensity. Heating the air alone, splitting it and delivering it evenly to the structure was the most simple method. Radiation was essentially eliminated for sake of nearly pure convection.

(continued)

Mike T., Swampeast MO

This nearly complete lack of radiation had a previously unheard of consequence. You could heat the air quite warm, yet still feel uncomfortably cool. Without a source of radiation, the body itself became the radiator, liberating its heat to the surroundings instead of RECEIVING radiation from the surroundings.

A new revolution evolved. Again big hunks of cast iron were used to supply the heat to the space. Unlike the old stoves however, these operated at a MUCH lower temperature. Both convection and radiation were slowed considerably, but since the heat source was split into numerous low temperature sources comfort increased significantly. The body was again RECEIVING RADIATION and it was possible to be comfortable without directly heating the air to a high temperature.

It was found that the larger the radiating surface (and the lower its temperature) the more comfortable the occupants became at lower space temperatures.

This continued until "full-circle" was achieved--again heating the structure through a very large, low temperature panel. The BIG difference is the removal of the initial heat source (the fire) from the final heat source (the panel). It's very important to note that the efficiency of the hypocaust has only recently been achieved AT THE INITIAL HEAT EXCHANGE! Only with near perfect heat transmission to the panel can this most ancient method be bested.

The obvious way to heat a panel (say the floor) was to embed one good conductor (metal) into another (concrete). This resulted in near perfect transmission of heat. Problems occurred though--the transfer medium could not be too hot as it would overheat the panel. The panel was also liberating its heat in ALL directions--remember that with a FIRE directly underneath the floor the under side is the "receiver" and the upper surface is the "liberator." Amazingly, it was discovered that mother earth did a decent job of insulating the slab--I believe this is because the deep earth itself is a low-intensity heat source. As long as ground water does not contact the slab a stasis is achieved over time and most of the heat makes it where we want. Later it was found that insulating the edges of the slab significantly reduced its "wasted" heat.

People again became impressed by the wonderful comfort and practicality of nearly pure, LOW LEVEL radiant heat.

Now, the debate rages on just how to achieve even, low heat in the panel.

Despite the losses through the earth, the tube-in-slab method continues to gain efficiency through the use of low-temperature boilers. As boilers that are able to safely produce water at the actual temperature required by the slab become more prevalent, transmission losses lessen throughout the entire multi-stage transfer process. This is a key point--by keeping the transfer medium relatively cool and extracting the majority of the available heat, the intermediate heating becomes more and more efficient.

Adapting this heat to wood or existing construction can be difficult. Despite the lack of heat loss into the ground, it's not easy to force LOW INTENSITY heat through an insulator like wood. The simplest way is to turn joist spaces into little heat chambers by attaching or suspending heated tubing. There is very little direct conduction of heat here, so convection AND radiation are the only means of transfer. At first look, this seems a good idea--it's quite similar to the ideal hypocaust system.

The problem is that the heat source IN THE CAVITY is extremely limited in temperature and often the tubing itself is an insulator. Convection AND radiation in the cavity both occur in a constantly-changing interplay. Significantly less heat is extracted from the system and inefficiency in the initial heat transfer (at the boiler) mounts. Also, heat tends to travel down through the joist cavity as well and being surrounded by free air, it is readily liberated. Insulating and air-sealing the cavity helps, but this slows convection, increases radiation and the system still looses heat downward. It works, but with inherent inefficiencies. The real conundrum is that the best efficiency AT THE BOILER is achieved with low temperatures, but the best efficiency IN THE CAVITY is achieved with high. Think again of the hypocaust with the fire directly under the floor.

To make wood construction behave more like masonry, a number of methods are in use to enhance the CONDUCTION of heat into the panel. Invariably, these involve surrounding the tube as much as practical with a good conductor with "wings" to spread the conduction as much as possible. Downward losses are minimized by nearly eliminating convection either by putting the wings in a "sandwich" between sub- and finish-floors or directly insulating down-facing wings. The temperature of the transfer medium can be kept to a minimum and a high percentage of its available heat can be extracted.

Are any of these methods perfect? NO! Do they all work? YES! Is one better than another? Yes AND no--it's mainly a function of economics and diminishing return.

The REAL point in all this? We STILL haven't bested the efficiency and comfort of 2,000+ year-old technology!!!!!

Earthfire

Me Thunks

That puts it quite plainly and even astutely