Why I chose to use radiators and not underfloor radiant

Gene Davis_3

For a house with its common area, that being foyer, kitchen, greatroom (dining and living), and that greatroom having a large ceiling vault and a lot of windows, I felt that there was no way to get enough heat (up to 23K btu/h) into the almost 400 sf space via its floor.

Thus the panel radiators.  And, if I had heated the floor with underfloor tubing and plates, I would have needed to supplement the room with emitters of some type, and they would want to be run with water at higher temp than the floor.  Two temps, mixing, controls, etc.

So I chose to try this with panel radiators throughout the house.  Small mod-con, 140 F. supply everywhere, TRVs.

Does this sound reasonable?

mor

Radiant walls

was that an option?

CMadatMe

Scratching

How does a 400 sqft space have a heat loss of 23K? Are all the windows and doors open?

Gene Davis_3

Images of the room

Its average ceiling height is about 18 feet.  Three of its outside walls are exterior walls.  Windows and glass doors penetrate all three of those walls.  Winter design temp is -15 F.  Pics attached.

Alan (California Radiant) Forbes

60 BTU/sq. ft.

You sure your numbers are correct?

CMadatMe

I just ran

a assumed heat loss based on the pics and your post. I came up with 13K using a -15 design temp for a radiant loss. I made many assumptions from the pics. Room is 17 x 24 or so, Ceil height 18, Wall height 12. I used no heat below with a temp below of -15.  I figured the doors and windows sizes based on experience and previous losses.  I also used R-19 for the walls and  R-30 Ceiling.  That gives me a 32.5 btu sqft and I'm sure my figures are a little fat.

Gene Davis_3

Forgive my lack of knowledge.

I used SlantFin's Hydronic Explorer 2 spreadsheet calculator to get my numbers, and you can see them here.

Since the foyer's exterior wall is an extension of the greatroom's, I thought it made sense to add those two rooms together, when I figured out my array of panel radiators for the space.

NRT_Rob

no way

the load is correct. I don't run into 60/sq ft even in old, uninsulated buildings.



also, adding some suspended tube to the radiators, at the radiator temperature, would give constant circ floor warming and reduce the size and/or number of radiators.



cheaper? no. but loads more comfortable. Radiators don't heat floors.



this could be done very simply and the cost of tubing the floor should be quite reasonable.

Alan (California Radiant) Forbes

OK

The numbers look good. 



If you're designing the radiators for 140°, they will be big and expensive.  But if fuel efficiency is your goal, then it's a good choice if you don't mind the added room ornaments.

Simply Rad

I agree its a HUGE load

The largest load I have come across was 50 btus/ft2 with a log home greatroom and a design temp of -20F. 

I would look at the calcs again.  Then I would look at upgrading the insulation package and windows to higher R vaules.  Reduce your load and use low temp radiation.



Jeffrey

Mark Eatherton

Me three...

I use to have that Explorer software, and as I remember, if you change something, it does not go through, and you might not notice it.



I would print your U values, redo the loss calc without making ANY changes mid stream, and see what it comes out to.



Also, one of the u factors for windows in the foyer is showing an R value of 33 for a window, and the only window I know of that has that high of an R value, would be my electric windows, and I don't think you are using them :-)



SOMETHING is not right. at [url=http://www.warmboard.com]www.warmboard.com



Also, look into a product called Warmboard. Excellent heat transfer and it is structural.



ME

Mark Eatherton

PS....

We were at "design" (zero degrees F) conditions here in Denver this morning, and I decided to do a quick check on the operation of my own home. I clocked the gas meter connected to my boiler to find that my house was using 14 btu's per square foot per hour. The heat loss calcs say it SHOULD be using 29 BTU/Sq Ft/Hr. And this is not the first time that I have seen these conditions and low loads.



These heat loss calcs are really nothing more than a S.W.A.G. What happens in real time is the proof of where the rubber meets the road.



I realize you are an engineer, and engineers are wont to overkill, but where do we draw the line of theoretical versus reality?



ME

Gordy

radiant walls

 If the load is higher then the floor can provide. Think radiant walls to supplement. Looking at your drawings there is space for this. Under that island counter where the stools are is a great spot also....even the counter itself.



I doubt the load is that high I used the slantfin program had the same troubles Mark calls out also.





Gordy

Gene Davis_3

Thanks for all the input.

I have been playing with the SlantFin spreadsheet calculator some more, and have made some observations.  First, the spreadsheet seems to be truly dynamic, in that the heat loss figure is jiggled by any changes made to the inputs in the fields in the column above.  Second, the factor that can whack the heat loss most effectively is the one that addresses air infiltration.



Another that affects the calculation, but not as much as the infiltration, is the "factor" for the windows and doors.  They call it a "factor," but the numbers they've got charted for various types of windows makes me pretty certain they are using U-value numbers for the fenestration assemblies.



Door and window companies have gotten better at testing and reporting their numbers, and modern windows can have real numbers down in the U = 0.020 range, if you go the whole nine yards with warm-edge spacers, triple-pane IG units, inert gases beyond even argon, and sputtered coatings like "low-e squared" and better.



I've said before in discussing this project in other threads here, that we will take great care in doing all the work that reduces air infiltration.  We will also use good windows, and half of all the openings in the greatroom are fixed units, both of these factors that can drive down the air infiltration.



Nevertheless, this great room is a big tall thing with three exterior walls and an attic/roof directly above, with plenty of fenestration.  The window and glass door area is equal to about six and a half full sheets of plywood



So, if I knock that factor down that correlates to the infiltration, taking it down to a maybe realistic figure of 0.012, and tweak the fenestration U-factor down a tick to 0.030, the calc is now showing the room loss at 13,300 btu/h.



This is all good.  If I apply these kinds of "better factors" to all rooms in the house, I may end up with a total for the structure at something like 25,000 btu/h.



And then that poses another problem for me and it is this.  With that low load, I can probably scrap the radiator thing and do this house with staple-up, plates, and PEX, but how does one heat the water with a mod-con, and not have it short cycling, at these kinds of loads?



Wouldn't it be nice if someone made a nice tiny wall-hung that ranged from about 6K up to 36K?

Simply Rad

Would that be nice

Glad to see that your load is not that big.  Like I said earlier my largest room load was 50 btus/ft in a log home great room with cathederal ceilings, dormers and lots of glass.  Lots of air infiltration too. 



I have spoken to manufactures about that size of mod/con.  The answer is....that below 15,000-18,000 btus its hard to stablelize the flame. 

I had -29F this morning, I normally use -20F for design.  I clocked my meter and I was around 25,000 btus.  Though my meter dials were slicky and jumping instead of just turning.  When its that cold nothing seems to work right, especially my fingers.  Anyways, I have a 18,000-80,000 mod/con.  I find that it short cycles until the OT drop below -5.  At that point it purrs like a kitchen.  I have not done anything yet but someday I would like to add a buffer tank.  Still it works great and keeps the house very comfortable and uses very little gas. 

I would stick with a mod/com. I have found anything else may give you troubles.

Jeffrey

NRT_Rob

I never recommend ACH quite that low

I recommend a minimum of 0.2. HRVs and such, dontcha know.

Paul Pollets

Calcs

The SlantFin software should not be used to calculate radiant floor heating loads. It does work well for radiators.  If you use either the Uponor or Viega software (not free), you would be able to get a better handle on the true radiant loads and be able to see the difference between applications (i.e. staple up with plates vs embedded methods). . The radiant software also flags the room if supplemental heat is required beyond what the radiant panels could provide. 

Alan (California Radiant) Forbes

Eleven years

ago, we did a house that had a room a bit like yours; it was a 2-story open dining room;  two of the walls were all glass.  It calc'd. out to 50 BTU's/sq. ft. and I was worried that I couldn't get enough tubing on the floor to heat the room properly on a design day.  But I trusted my numbers.



Radiant or radiators, I'd make that Greatroom of yours a separate and independent zone.  It has a very different heating profile than any other room in the house, save the stairwell.

Gene Davis_3

One comment with which I would like to disagree is . . .

. . . the one made earlier that HE2 "software" is OK for load calcs if one will use radiators, but for figuring out loads for radiant, one must use something else.

Why would that be?  Heat loss through a building envelope has to do with the makeup of that envelope, the temp differential, and other things, but a BTU getting through that envelope will do it regardless of the source.  Whether the emitters inside are sections of floor, walls, finned baseboard, or radiators, all is irrelevant when it comes to what is getting lost to the outside.

That is enough basic thermodynamics for now.  Let's talk radiant heating design.

Looking at my house design and its glassy greatroom, if we expand our view of that space and include all that adjoins it without walls or doors, that being the kitchen, foyer, and hall/staircase, and figure all the available floor area of those combined "rooms" into one, I get a total of 476 s.f. of floor that can be a radiator.

I did not count the front hall closet, or under cabs, island, or fridge in the kitchen.

I then went into my HE2 spreadsheet and de-tightened my infiltration factor per Rob of NRT's recommendation, setting it to 0.02.  Total BTU/h loss for all those rooms is now 20,840 BTU/h, and if we want the floor "radiator" to match it, it will do so at a rate of just under 44 Btu/h per s.f.

Howzzat sound?  Supplemental needed?  May we just presume, for those few hours each winter when the load is really that high, that either the floor cannot emit enough, and the space will be cooler, or better, that the wood stove will be used (if someone is actually there), and the place will be very toasty indeed?

Gordy

There is a difference

     Slantfins hydronic explorer program (original) is tailored to baseboards. HE2 is tailored to radiant. If you plug in the same data to both programs you will get different heat losses. Trust me I have done it with my own home. You will also find that HVAC cac tailored to FA will give you something else. Most of it has to do with infiltration FA tends to give more of a negative pressure on a structure then radiant. A dwelling is under negative pressure anyway.



Gordy

CMadatMe

Let's Put this To Bed

Supply the following information



Outdoor Temp

Desired InDoor Temp

Room Sqft

R-Value Walls

R-Value Ceil

R-Value Below Floor

What is the sub floor and finished floor?

Ceil Height

Average Wall Height

Sqft of Outside Wall

Sqft of Windows/Type of Windows

Sqft of Doors/Type of Doors

Heat Above?

Heat Below?

Sqft of Area without loops (cabinets, island, etc)

Gene Davis_3

Your required parameters require me to give you just the greatroom.

So here it goes.



Outdoor Temp = -15 F.

Desired InDoor Temp = 68 F.

Room Sqft = 14.07 x 24.42 = 338 sf

R-Value Walls = 21

R-Value Ceil = 50

R-Value Below Floor = fully conditioned  crawlspace below

What is the sub floor and finished floor = 3/4" OSB, 3/4" hardwood

Ceil Height = vaulted, min 13'6, max 20'6

Average Wall Height = 14.8'

Sqft of Outside Wall = (2x14.07x13.5) + (24.42x20.5) = 881 sf

Sqft of Windows/Type of Windows = 208 sf, doublepane, low-e squared, argon

Sqft of Doors/Type of Doors = both glass patio doors included in above

Heat Above? = no, above is an insulated attic with roof above

Heat Below? = the sealed insulated crawlspace below gets heat from system

Sqft of Area without loops (cabinets, island, etc) = 16 sf under woodstove pad



But if you wish to continue, the foyer that is open to this room has same floor type, same wall type, same floor type, is 5.25' wide with an 8.0' exterior wall, entire floor area available for radiant, 8' clg ht., heat above, one entry door, foamfilled fiberglass, 3'x6.7', in the exterior wall.



And to go the whole nine yards, the hall/stairway adjacent and open to the greatroom has a 9' width x 10.42' length, one exterior wall along the length with height of 13', R-28 in that wall, windows in that wall same type as elsewhere, 23 sf total, same floor makeup as elsewhere with same crawlspace below, same ceiling type and roof above as greatroom, and only 42 sf of the floor is available for radiant.



The kitchen in the middle of all this, with same floor as everywhere, has no exterior walls, 8' clg ht, is 9.04 x 10.42 in size, and has only 52 sf of its floor area available for radiant.



Floor plan of main floor attached here.  And thanks!

Mark Eatherton

Disagreements welcome...

Disagreements end up educating people to other peoples points of views.



A BTU is a BTU is a BTU, except where radiant heating is concerned....



The output of a vertical panel radiator if 60% radiant, 40% convective. It is the convective that causes stack effect, and stratification. Once the MRT starts maxing out, the ratio shifts more towards convective and less towards radiant.



With a radiant floor, 80 % of the output is radiant, balance is convective. If you could look at the ceiling, especially high ceilings, of the home under operation with the convective, you would see significantly higher temperature up high than you would down low. The exact opposite is true with a radiant floor. The ceiling is cold, and floors are warm. Little to no bouyant air induced stack action.



Additionally, there is a floor back loss factor that needs to be accounted for in the radiant floor system that is pretty much ignored on the convective loads unless the house sits over an unheated space.



So, heat loss is heat loss, but it is dependent upon the type of delivery system to determine how it leaves the home. As has been pointed out, forced error systems REALLY blow... Literally.



Additionally, convective systems do not significantly affect the MRT (Mean Radiant Temperature) of the buildings which they serve. Radiant systems DO significantly raise the MRT, and MRT drives the bus of human comfort.



In my mountain home, when I first get there, the house is ice cold. As soon as I fire up the radiant windows, and the windows get hotter than the surrounding ambient and MRT, I can feel the heat coming from the windows, and as the house approaches 65 degrees F, it "feels like" 70 degree F in a force air or convectively heated home.



Heat is but one component of comfort. Proper deliver makes all the difference in the world. Radiant heat rules, but not just floors. Ceilings, walls, window, countertops, radiators ALL deliver good comfort.



The other key to good comfort is being able to control the uncontrolled movement of air, known as infiltration, but I think you already have a good tight handle on that issue.



ME

CMadatMe

Heat Loss

I used the numbers you gave me but compared them to the picture on your post plus the picture on your previous post. I come up with a radiant heat loss of 11,583. Could it be done yes but you are going to have some very high water temps, By high I mean 170 degree water temp on your design degree day. That is not running with plates but with Quik Trak. You really don't need much as far as supplemental heat goes to get that water temp down. Some radiant wall here would work wonders for you but you are limited in wall space.



I personally would do radiant. For the little supplemental heat to get those water temps down I would add 2 corner panel rads on the wood stove side wall. You really will only need them at design but atleast you can fit them in the decor of the room and utlized them as corner shelves. Yes they make shelves that attach to them and yes you can ge them painted to fit the decor of the room. You can also run them at the same temp as the radiant so you won;t have to add another mixing of water temp.



Just my two cents but the call is yours.

Simply Rad

170F supply temp?

29 btus/ft2 = 170F  that's way too hot.  I agree with ME Warmboard is the ****.  The charts I am looking at show that at 29 btus/ft2 and R-1 for hardwood with a supply temp of 115F.  and that is using 70F room temp.  If you are using constant circulation control you can reduce that supply temperature by 10%.  On average I find that the Warmboard systems operate in the 90F range most of the time.  That mod/con boiler is just drooling thinking about it.  Quicktrak really 170F?  The biggest thing to overcome  I find with high efficiency radiant are floor coverings. 

Same load of 29btus/ft2

R .5  Tile floor=100F

R 1  Wood flors=115F

R1.5  Great carpet choice=128F

R2  Good carpet choice= 145F

R2.5  Poor carpet choice=157F

R3 BAD Carpet choice= 168F

Anything else I do not care about because the boiler we use is limited to a supply of 167F



So basically you start at 95%+ (tile and harwoods) efficiency and end at 87% +/-(bad carpet) for boiler efficiency.

I begin the flooring choice topic during the initial meeting with the homeowner.



Jeffrey

CMadatMe

I Actually Made a Mistake

The idiot that I am I forgot to change the R-Value of the floor when I switched from joist trak to quick trak...Thanks Rad for making me dbl check.....It's 139 degree water at design... Gene Sorry for the confusion......

EricAune

Radiant with constant circulation

Why not consider constant circulation radiant?  You can control this with an outdoor sensor and treat the supplemental as an additional zone, independent from the floors and only circulated when needed. 



I would design the supplemental as the same temp as the radiant, use panel rads for the supplemental fed off the same manifolds as the floors (if pumping works out) and control the rads with power heads and a stat. 



A basic heating control  will take input from an outdoor sensor, indoor sensor and control the boiler.  Depending on your boiler you might already be able to do this without adding another control. 



Control the power heads for the supplemental with a Taco ZVC and a heat only stat.  This would not be connected to the boiler, it would only enable the supplemental when needed.



Constant circulation will achieve a greater efficiency throughout the system with the maximum comfort.  

Simply Rad

No need for supplemental heat

With 29 btus/ft2 at design conditions you will not require supplemental heat. The floors will be adequate. Instead of Taco controls why not use true constant circulation with non- electric comfort dials. Oventrop has treated me well for years. Simple and effective. Customers even complement on the simplicity and comfort!

"the ultimate in comfort and efficiency"



Jeffrey

Gene Davis_3

This is great help you all are offering, but . . .

I'm not getting it all.  Forgive me, but as I said, I am not a heating pro.  Just a retired structural engineer that took physics and thermodynamics long ago.

Are we now saying that to get enough floor radiation to heat this common area consisting of greatroom, foyer, hall-adjacent-kitchen-and-staircase, and kitchen, that we do it with Uponor QuikTrak and 5/16 PEX tubing, and we'll have 139 F. supply at design temp?  And need no radiator supplement?

In my post made here on the evening of January 8, I reported my area for available floor radiation, for all that space described here, as 476 sf.  If we've got the ability to emit 29 Btu/h per square foot, that computes to 13,804 Bth/h total.

I took a look at my other rooms, the three bedrooms and two baths, and the loads I computed with my SlantFin HE2 spreadsheet.  Discounting all those loads down by the same factor as we now have for the greatroom space, it appears as if 29 Btu/h is adequate for doing all those spaces, too.

So are we now recommending doing the whole 1475 sf house with floor radiant using QuikTrak and 5/16 tubing?  And maybe something like the little Knight Lochinvar W50 boiler, 5:1 turndown, 10K to 50K input?

Can we keep discussing this?  I would like to understand more, and then maybe get into controls.

cattledog

Floor Temperatures

Gene--



Now that we are settling down on 30 btu/h/sq ft you need to address the floor temperature/flooring material issues. The coefficient of radiant floor thermal transfer is 2.0 btu/h/sq ft/degree F.



To achieve T=68 with you will require a surface temperature of 83F for 30 btu/h/sq ft or 80.5F for 25 btu/h/sq ft. Both exceed the maximum recommended surface temperatures for hardwood floors, which is 80F.



At 80F max, you can achieve a room setpoint of 65F at 30 btu/h/sq ft.



I don't know how frequently you will be at design day conditions, how you will use the wood stove, and how toasty you need to be, but per your original specifications, you are pushing the limits of a radiant floor.

Gene Davis_3

Thanks, Cattledog. Does this take us back to the beginning?

The beginning was the title, "Why I chose to use radiators and not underfloor radiant."



And in that opening post I went about feebly trying to explain that I had more heat required in my big tall glassy common space than could (I thought) be delivered with in-floor radiant tubing.



Which brought about a review of my heat loss analysis.  And then some very helpful and persuasive posts were made suggesting I lose the radiators idea and do floor radiation.  Which then turned the discussion into a how-to on load analysis, for infloor radiant, and how and why it needs to differ from that required for the mostly convective radiators type heating.  Very instructive.  Very helpful.



But the desired flooring finish, throughout, except for the two baths, is hardwood strip.  Furthermore, the budget is tight.  Oh, and one other thing.  The place will get most of its use in summer and the fall shoulder season, with only occasional use in winter.  Finally, that greatroom has a wood stove.



Here is my quick and dirty analysis of cost of radiant on a per square foot basis, and it is just for the top-of-deck QuikTrak, tubing, and the under-deck insulation.  $4.75 per square foot, and times the more or less 1500 sf, that comes to $7,125.  The labor to install all might be as much as another $1800.



A very rough look at radiators puts my cost, for the panels, tubing, and TRVs, at about $4000.  I am thinking the installation labor is the same as for the other, or less. 



Then there is the system, all the electics and pumps and valves and tanks, between the boiler and the supply and return ends of the tubing that goes to and from those emitters, whichever type they may be . . . floors or radiators.  My guess is that the control board side of a radiant floor package will cost more than that for one doing panel radiators.



Is my assessment wrong, and that I am better heating this place with in-floor tubing?



Or is this project best done with radiators?



And if best done with radiators, how can I best choose a pro to do the heating load analysis and system design for me, so this can be quoted and so we can know what our budget needs to be?  My resources, here in the thinly populated and poor northern Adirondacks or upstate NY, are slim to none.

Mark Eatherton

The funny thing about hardwood floor temperatures...

The national hardwood flooring council adopted the RPA's flooring guides a long time ago. The RPA's recommendation of limiting the floor surface to 85 degrees F is not a mechanical issue, but rather a human physiology issue. If your skin comes into contact with a surface that is greater in temperature than your skin, you WILL start sweating in an effort to limit your bodies core temperature.



So, back to hardwood maximum floor temperatures. If in fact, hardwood floor is limited to 80 degrees F surface temperature, then it should be against the code to allow the installation of hardwood flooring in any situation where it is going to see sunshine, because I have seen my own unheated hardwood floor at 140 degree F WITHOUT having any heat underneath it...



Just sayin' ...



ME

Gordy

Warmboard

Is the best bet its been said already. It takes the place of the subfloor one step your done with a good surface for nailing hardwood to, and superior layer of conducting aluminum.



  Any time you layer surfaces you are more prone to floor squeaks such as with a 3/4" subfloor then quick track then hardwood. Not to mention the extra labor of putting a floor down three times instead of twice to finish product.



 The 80* for hardwood all though recommended is a bunch of bull.  Shoot the floor with the sun shining in through those windows on a sunny day if you have a southern orientation, I'll bet your way over that. Humidity is the key to a stable hardwood floor, and aclimation of the flooring BEFORE installation.  After that it is about controling humidity in a tight range. Enough said



Speaking of orientation that has not been discussed. All though not a factor during cloudy days or at night. Southern facing solar influence will play a big part of demand during a sunny day on the system in this area.



Also design day is about 1% of the heating season. Maybe a weeks worth, and thats pushing it. Plus design temp is not usually going to be over a 24 hour period.  It takes alot of cold to make the mean daily temperature approach a design day temperature.



Just some more to mull over.



Gordy

Gordy

Sorry Mark

 We must have been replying at the same time.  Did not mean to be a parrot.



Gordy

Mark Eatherton

Contact NRT_Rob...

THAT is what he does for a living :-)



ME