I'm working on a project at the moment that has realllly thick concrete walls. Just some for example:

36" thick - 420 lb/ft2 weight

72" thick - 840 lb/ft2 weight

(architectural gave me the thickness and the density (140lb/ft3) and I just multiplied through by the thickness in feet for those weights)

For a reference, if you're using the CLTD method, the thickest wall-type group is ASHRAE Group A. These ASHRAE wall-types work fine for 99% of normal projects. As you know, if you are using CLTD, even if you're wall doesn't match up exactly, you can find something close enough. However, the thickest concrete this covers is 12" @ 156 lb/ft2 weight.

The U-Value portion of these walls is not a problem. Obviously I found the closest concrete construction type and used the resistance per inch to figure out the U-Value. The problem relates to the mass and weight of these walls. The ASHRAE wall types, as I'm sure you know, are primarily chosen because the mass of the building construction plays into the thermal transfer of heat through the wall into the building which affects what time of day the building will peak etc. Obviously a thin metal wall will see the heat inside sooner than a concrete wall, even if the U-Values are the same.

Have any of you dealt with construction this extreme? (ie this is for a nuclear power plant). If so how did you accommodate for this? With both CLTD and RTS, manually or using programs I have had to choose a wall type/mass value and it is significantly lower than what I am dealing with. As far as I know, HAP doesn't deal with walls this thick either and TRACE 700 seems to have problems as well (per someone I'm working with who messed around with the demo).

[Edited on January 13, 2010 at 3:07 PM. Reason : ]

1/13/2010 2:45:02 PM

http://hvac-talk.com/

1/13/2010 4:03:22 PM

forum looks like its mainly geared towards techs and craft type people. i'll post something up in the commercial section though and see.

1/13/2010 4:45:25 PM

Why won't using thickest thickness in whatever method you have to choose from work?

1/13/2010 5:06:03 PM

It might be a good enough approximation, I'm not really sure. I don't have any data to test it against. As I stated, the thickest wall in the ASHRAE groups gets to only 12" of concrete and 156 lb/ft2. I am not sure if this was done because after that the difference is negligible or what.

But as far as I understand it, the thicker the wall and the more mass it has, the slower the release of the heat will be into the inside space. With super thick walls I am assuming that there is a huge difference in 72" vs 12" on the peak load hour in the space. I could very well be wrong though. But seeing as that 72" is 6x as thick as 12", I can compare and say that using 2" of concrete in a load calculation is not a good approximation for using 12" of concrete.

I've considered using the thickest wall group and just overwriting the U-Value for the 12" with the U-Value for 72", I'm just concerned that the difference in the mass is significant enough to cause an incorrect answer.

1/13/2010 5:15:35 PM

so if you could find a U-value for concrete of that density, you'd be set?

[Edited on January 13, 2010 at 5:19 PM. Reason : cause you can just scale the thickness factor, right?]

1/13/2010 5:19:23 PM

What do you for a concrete slab on grade with infinite* length of soil below it?

Does the concrete get and value, then the subbase, and then the soil? Or do you just assume a number for some max amount?

1/13/2010 5:19:32 PM

^^ I have the U-Value. The density is the same as one of the ASHRAE standard list items and is 140 lb/ft3. the R-Value for that ~ 0.09 ft2*F*h/Btu*in. So I know the U-Value from the inverse of the resistance/inch*thickness. But separately, the mass of the wall also affects the calculation.

For example, 4" of 140lb/ft3 concrete is an ASHRAE wall type E, while 12" of 140lb/ft3 concrete is an ASHRAE wall type B (or with insulation is an ASHRAE wall type A).

^Depends, a lot of time I'm not that worried about temperature loss through the floor, depends on the calculated ground temperature. But what you just describe, yes thats pretty standard. You just assume the ground is isothermal and it doesnt matter how much ground is beneath it at all. Thats different than the thickness of a vertical wall exposed to outside air.

[Edited on January 13, 2010 at 5:39 PM. Reason : ]

1/13/2010 5:39:48 PM

oh, i see. i wish i had time to take the MAE HVAC elective.

1/13/2010 5:56:23 PM

yea I wished I had taken that too since I ended up in this industry. I'm not sure what all they cover in there.

1/13/2010 6:19:32 PM

sumfoo1, darkone, or anyone else that (I think) works in HVAC have any experience with this?

1/13/2010 9:45:42 PM

My suggest is to use your software do do a sensitivity test. Vary your wall thicknesses and the resulting U-values and see how much your heat loads change. You might be able to construct a reliable regression.

1/14/2010 12:40:04 PM

IIRC cltd is just a "shortcut" for some other full blown method where the ashrae tabulated some factors for common wall types/sizes. Seeing as how you have an uncommon wall, it sounds like you either need to calculate some custom "factors" (or whatever they're called), use the long drawn out math intensive method (a big bad ass energy balance maybe?), or use some software like ansys or trnsys.

1/14/2010 3:05:48 PM

^ Yea, I was actually thinking of doing that for a couple 100x100x10 buildings with no windows or something. Just to compare.


^ and yea, you're right. CLTD, RTS, TFM and all the little variations used by nearly all software out there are different variations/simplifications of the old cumbersome Heat Balance Method. If it really comes down to something like that, I'll probably make some approximations.

The space inside the building is only being kept at 104F in 90% of the building and the load of the building is driven by equipment anyways, so I don't want to spend too much time on the walls. I mainly wanted to make sure I was treating this right if the client questions anything. But, in reality, even in the summer, the heat gain to most of the spaces is going to be negative due to the indoor temperature

1/14/2010 3:30:01 PM

For anyone that cares, I think I did figure out a way to correctly account for this. I need to try and verify it but it makes logical sense.

If you look in ASHRAE Fundamentals 2005 Chapter 30 Table 17, that is where the standard RTS wall types are listed with their "Conduction Time Series" %s associated with each hour. As you probably know, one of the ways that RTS is useful as a method is taking into account the "carryover" of the conductance load through the walls from one hour to the next. The thicker the walls, the more future hours are affected by the load at the outside surface of the wall at the current hour.

I didn't really think about it before, but the way ASHRAE calculates this as a percentage, the total % conduction for the 24 hours of that surface have to add up to 100%. Obviously the less dense or thinner the walls are, the quicker it peaks, thus at your current hour of load, if you have a metal panel wall, something like 40-60% of the load after insulation affects the space in the same hour it was applied. So your conduction at that hour shows up as say 50%, the next hour 20%, 18%, 12% and then its done and the rest of the hours are 0. With concrete, it varies with thickness and type of course, it looks more like 1%, 2%, 5%, 6%, 8%, 10%..then starts to drop back down etc. all adding up to 100% over 24 hours.

After studying the thickest walls that ASHRAE lists by default in the Fundamentals book, you can see that, as the walls get thicker, the % of applied conduction for each hour has less of a difference. It appears that at some point, the wall gets so thick that, using this method, the only way to approximate it correctly is to divide the 100% evenly over the 4 hours, which leaves 4.16% for each hour. The closest example I could find to my situation was Brick + 12" of concrete + some insulation and gypboard which had 4% listed for 20 of the hours, with 5% listed for the other 4. With 72" concrete (even though I have no insulation, its just air + concrete + air) certainly having more mass than this, I am going to assume that each hour approaches 4.2% and just use 4.2 as the CTS number for hours 1-24.

2/11/2010 10:13:12 AM

Seems logical.

2/11/2010 10:18:35 AM

Yea, it made a lot more sense the more I looked at it today.

I guess you can kind of think of it like a damper in a vibration system where the thermal mass resistance is the damper. As some point it gets so large that the system's reaction to the temperature change outside over the course of the day is almost a straight line if you graphed it as opposed to the cosine wave of normal walls.

2/11/2010 9:23:02 PM