I am evaluating a new 10-story hotel design that is proposing glass curtain walls and pre-tensioned concrete floors that project beyond the walls. On 3 sides the projections are minimal (6 inches). However, on one side, the projections extend about 3 ft.

For the small projections, I intent to use the following method recommended in the thread below. "In screen 4 of 25 of the DD shell wizard you can specify "Slab Penetrates Wall Plane" (check box)."

I am interested in any tips for modeling the thermal impacts of the larger projections. Thoughts? (I plan to use window shading for the light/shading impacts.)

I am hopeful that the analysis will help convince the design team to incorporate some insulation and thermal breaks!

Thanks in advance!

Kevin

Mon Dec 6 11:45:18 PST 2010

Previous message: [Equest-users] balcony

Next message: [Equest-users] Shading

Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

In screen 4 of 25 of the DD shell wizard you can specify "Slab

Penetrates Wall Plane" (check box). If you do, eQUEST created a new

wall type of height equal to your slab thickness and composed of 1' of

concrete plus any slab edge insulation you specify on the same wiz

screen. That provides a parallel path for the heat transfer from the

balcony or slab edge. Not exactly 2D heat transfer modelling -- but at

least you aren't ignoring the effect of the slab edge or balcony.

Kevin Coleman, CEM, LEED AP

Hello Kevin,

I would recommend looking at this report prepared for ASHRAE TC4.4 by Morrison Hershfield.? They have some pretty good data which may help you out.

http://www.morrisonhershfield.com/ashrae1365research/Documents/MH_1365RP_Final_%20small.pdf

Byron D. Burns, EIT, BEMP

Kevin,

I'm fairly sure eQUEST is not even going to closely approximate the heat loss/gain accurately unless you dial the loads in manually which means you have to calculate them by hand first. I suggest referring to ASHRAE-D-RP-1365 for guidance on estimating the loads. The section you are interested in starts on page 35.

Thanks,

David W. Griffin II

Here's a timely report on effective R-values of assemblies with slab/balcony thermal bridges:

The Importance of Slab Edge and Balcony Thermal Bridges

Regards,

Bill

William Bishop, PE, BEMP, BEAP, LEED AP

Great report ..... thanks.

Michael Hupel, B.Tech, CEM, LEEDRAP

Some more references

1. See attached " A Bridge too far' from the ASHRAE Journal, Oct 2007.

(This is freely available on the internet)

2. Also read the ASHRAE Journal July 2012 issue Thermal Bridge.

Redux

*Joseph W. Lstiburek.* P. 60. July. (Will require ASHRAE login info)

3. There was a presentation in the 2012 ASHRAE Energy Modeling Conference

Atlanta that detailed out the impact of extended slabs and the effective

R-values. Although I do not have details of that presentation, maybe

someone does.

Regards

Ramya Shivkumar

To me the most interesting part of the report isn't the conclusion that thermally broken balconies and slab edges give significantly better thermal performance (we all know that intuitively). Instead it gives a way of back checking what I would consider to be "standard practice" for calculating area-weighted U-factor for these same assemblies using default ASHRAE U-values to sophisticated thermal model results. Here are my thoughts, checking my calcs and comments appreciated.

Executive summary: It looks like our standard assembly performance tables and area-weighted U-factor calculations get us "close enough" to the results from this report that it doesn't seem to suggest we should switch methodologies.

Comparison calculation for steel-framed walls (stud insulation + exterior rigid):

For a 16" OC metal stud wall, R-12 Batt cavity insulation and R-5 continuous rigid insulation, the report states that the Effective R-value for the wall + slab will be R-7.4. This is based on 8'8" floor-to-floor, and 8" slab.

Using ASHRAE 90.1-2007 Table A3.3 for steel framed walls, we would get an effective U-factor for the wall portion of U = 0.0785 (interpolating between the R-13 + R-11 batt values). If we use Table A3.1A for the exposed slab edge (assuming 8" normal weight solid concrete walls), we get a U-factor of U = 0.740 for the slab edge.

Doing an area weighted U-factor calc: Ueff = U1*A1+U2*A2/(A1+A2) or in this case, factoring out the Length of the wall section we use: U1*H1+U2*H2/(H1+H2) = (0.0785*8 + 0.740*8/12)/(8+8/12) = 0.1294 or an effective R-value of 7.73, which is pretty close to the R-7.4 claimed, a difference of about 4% in U-value. Maybe I'm cynical, but to me that seems close enough.

I only spot checked this assembly, so maybe there is more impact with different construction types.

Another interesting tid-bit: Apparently the much vaunted fin-effect is actually pretty minimal. If you look at the results of the exposed slab-edge condition vs the full balcony fin-effect, the balcony scenario actually shows slightly BETTER thermal performance. Odd.

[cid:image003.png at 01CED972.79A92370]

[cid:image004.png at 01CED972.79A92370]

Nathan Miller - PE, LEED(r)AP BD+C, CEM

Nathan,

In the Balcony calculation, are you accounting for both the 'edges' and 'exposed area' of the balcony ?

In my opinion, a concrete balcony should be accounted for as a heat exchanger with 3 surfaces (perimeter edges, upper surface, lower surface); it looks like you are only accounting for the edges.

I had a 'heated' discussion about this with a vendor recently and the above is more or less a conclusion of that. Some of the structural thermal break products can be too expensive to justify based on energy savings, if area of balconies is not accounted for; the bigger your balcony, the better your payback from thermal breaks.

Best,

Kapil Upadhyaya, LEED AP

Kapil,

No, I am only accounting for the ?face? of the balcony, as if you sheared it off at the plane of the wall.

That is my point, the fancy thermal modeling in the report, which is supposed to account multi-dimensional fin-effects doesn?t come out with much different result than if you just take a simplified approach like a typical UA-trade-off calculation (one directional heat transfer, no fin effects accounted for). So by that line of reasoning, you would likely be OVER ESTIMATING heat loss if you throw in all three heat-exchanger surfaces that you mention in your methodology. I?d be interested if you compared your calculation results to the thermal modeling in the report if you feel like a little intellectual exercise.

Nathan Miller - PE, LEED?AP BD+C, CEM

It surprises me that the report indicates a better R-value for the fin

effect of a concrete balcony vs. the simplified exposed slab edge approach.

Seems to violate the basic teachings heat transfer 101. I can't think of

a logical explanation of how that could be the case and would chalk it up

to intricacies of the modeling tool. However I agree with Nathan that

since the report doesn't show nearly any difference between the fin effect

vs. the exposed slab edge that it's not necessarily worth the complication

in modeling unless the balcony slabs are the focus of the study.

My guess is that the 3-D thermal model uses a pretty simple surface

convection/radiation assumption, like an air film. To properly model

the fin effects, the thermal model would need to be coupled with CFD and

a radiation model to account for the more complex heat transfer effects

associated with "fins".

I would think the effect of the balconies would increase the thickness

of the exterior laminar layer on the building. The air flow is

perpendicular to the fins. In calm conditions the main driver is going

to be stack effect. The exposed edges are on a flat surface and would

have a normal boundary layer. Without a projection they would offer

little resistance to the airflow from the stack effect.

A 60% reduction in effective R-value for a building is quite a

significant effect depending on where you are starting from. A high

ratio glass building is a low starting point.

Bruce Easterbrook P.Eng.

Abode Engineering

I think the key factor in comparing the case of a balcony and a slab edge flush with the exterior wall face is the effect of the balcony on the temperature of the concrete at the plane of the exterior wall.

For the slab edge: the temperature of the slab edge depends on how quickly heat is transferred to the outdoor air, with a thin boundary layer of air slowing the transfer of heat between the concrete and the outdoor air. The temperature of the concrete will depend strongly on the rate heat transfer from the concrete to the outdoor air over a small surface area.

For the balcony: the temperature of the concrete in the plane along the exterior wall surface will depend on the rate of heat transfer to the protruding balcony. The temperature of the balcony will depend on how quickly heat is transferred to the outdoor air over a much larger surface area than in the slab edge case. Since the heat transfer to the outdoor air from the balcony happens over a much larger surface area, it is possible for the temperature of the balcony to be closer to the temperature of the outdoor air than in the slab edge case in which case the balcony would lose heat faster.

In determining which configuration leads to higher rates of heat loss, a lot depends on how the transfer between the concrete surface and the outdoor air is modelled. This strikes me as the item of greatest uncertainty in the Morrison Hershfield report, and the report does call this out on page 23: ?Selecting heat transfer coefficients for accurate calculations can be a challenge for components where a larger percentage of the overall thermal resistance is the surface resistance. However, the thermal performance of insulated opaque building envelope components is the primary output of this project."

All the best,

Dan

?

Daniel Knapp, PhD, P Phys, LEED? AP O+M

I wasn?t saying the effect of exposed slab edges and balconies is insignificant. What I was saying is that I don?t see much difference in using the thermal performance for those assemblies modeled in the report vs. using typical practice of just calling balconies and exposed slab edges ?uninsulated concrete walls? and assigning that default U-factor to them (hence you would have about a 1? high uninsulated concrete wall and an 8-9? high typical opaque wall).

Nathan Miller - PE, LEED?AP BD+C, CEM

True. I was thinking more on the comparison of insulation efficiency.

These buildings as modelled and designed have R-18.5 of insulation and

the net effect to the envelope is R-10 or less. Comparing this type of a

building and our current housing code in Canada where glass is limited

to 17% of the wall area, R-4+ glass, wall insulation of R-28 and roof

insulation of R-60 and intense scrutiny of thermal breaks, well lets

just say they are energy pigs. When you have an R-2 or 3 glass wall a

little bit of R-2 concrete in there doesn't make much difference. Even

if you stick a fin out 8' is still doesn't make much difference.

The research is interesting and quantifies the effect. It also shows

current modelling does a reasonable job of reflecting what is going on.

But for the building as a whole and the model it can have a small impact

if it isn't efficiently designed. The impact of how it is modelled has

an even smaller impact. 60% of R-20 is significant, 60% of R-2 is

meaningless and a 3% error in that is nothing.

The big picture is getting lost in all this work. We are using a

modelling technique in which a 10% error is considered very accurate, a

good model. Many of the early LEED buildings, as built, deviated from

the models by 100% brand new. Many others that did ok on day one were

way out by year 5. The models weren't the main problem, it was

application. Modelling is expensive. It seems to me that sometimes too

much time, effort and money is spent chasing less than a percentage

point of gain. There is a science to keeping the model efficient as

well. Also, the main way a model is used is as a "what if" test

platform. Holding most things constant and trying variations. This

reduces error as well. But if the sealing contractor does a poor job

and the building takes a 15% hit a little bit of R-2 concrete on the

envelope doesn't matter very much.

The science is great, this type of research is very illuminating and

will drive better design. The application of our technology, the tools

we use and how we use them is the biggest variable affecting the

efficiencies we are trying to achieve.

Bruce Easterbrook P.Eng.

Abode Engineering