Adjacent Shells

7 posts / 0 new
Last post

Two questions in the same day...I may have more still :)

Has anyone taken the time (maybe Nick?) to determine if there is a major
impact due to leaving walls of adjoining shells as exterior instead of
changing them to interior, air-walls, or deleting them? I only ask because
I have a 4M+ SF building that is requiring 20+ shells stacked and adjacent
and I'm looking for every time saver I can find. I simplified the shells
as much as possible, but due to geometry this is the smallest number I can
squeak by with from an orientation/geometry/overlap.

I should note, this is an existing building and the model is being more or
less calibrated, so just having a rough idea of the impact of this nuance
will allow me to decide to take the time or to chalk it up in my model
error.

Thanks much!
Jeremy R. Poling, PE, LEED AP+BDC

Jeremy Poling3's picture
Offline
Joined: 2012-05-10
Reputation: 0

Jeremy,

Well, every one of those extraneous exterior walls (and roofs) is going to be simulated as envelope loads on those spaces. The model doesn't know that they aren't actually exposed to the outdoors. I would certainly delete them for any model, calibrated or otherwise, if it were me. You could determine the impact, relatively quickly, by modifying the U-factor of your wall construction based on the ratio of modeled exterior wall area to actual exterior wall area. Get your modeled wall area from the last page of the LV-D report, and divide by the actual wall area of the existing building. Divide the wall construction U-factor by the modeled-to-actual wall ratio, and modify the wall construction accordingly. Simulate and see how much the energy is reduced. Okay, never mind that sounds like too much work. Just delete the walls.

If you're in that much of a hurry, don't replace the exterior walls with interior walls. Just delete the exterior surfaces and be done with it. I usually add the interior walls, especially if the adjacent spaces have different thermostat setpoints or different internal loads. These edits can be pretty quick when you modify the 3-D view to more easily see where the extra surfaces are:

[cid:image002.jpg at 01CDF97A.3E198A80]

Regards,
Bill

[Senior Energy Engineer 28Jun2012]

Bill Bishop's picture
Offline
Joined: 2012-02-25
Reputation: 7

I agree with Bill - don't leave them as exterior walls - you will get non existent skin loads in those zones.

Another thing worth trying (if you are going through the wizard) is to make the building footprint slightly bigger than the zone footprint on that side of the building. Like this:

[cid:image004.jpg at 01CDF989.89343BF0]

You will end up with interior walls in those zones like so:

[cid:image005.png at 01CDF989.7EFDCB10]

Hope this helps

Vikram Sami, LEED AP BD+C

Sami, Vikram's picture
Offline
Joined: 2011-09-30
Reputation: -1

Thanks Bill and Vikram - with most buildings this would be an easy
solution, but with this particular building the problem involves hundreds
of these exterior surfaces. Unfortunately, the changes in building
geometry are not minor enough to simplify - bump outs exceeding 20 feet,
changes in direction and angle of the building, etc. I should have noted
that the 4M+ SF is spread across 4 floors...

I'll take a look at it once the architecture is finally complete in the
model. I'll have to do it as best as possible in the 3D view since without
a pan feature I can't get close enough to the tails of the building to
visualize the sections that are exterior.

If anyone does have an idea on what percentage of the load these phantom
skin loads can add up to in more typical cases, that would be helpful:
following ASHRAE 14 there's room for some small percentage error
adjustments due to things like this. To put it a little differently, we're
recommending the owner finish changing the last few incandescents to
alternate sources (depending on application, LED or CFL) but the threshold
for showing up as a percentage of energy use in the building is so high all
of the incandescent sources remaining wouldn't show up on the simplified
results in eQuest. :) Fun "little" project, yes?
Jeremy R. Poling, PE, LEED AP+BDC

Jeremy Poling3's picture
Offline
Joined: 2012-05-10
Reputation: 0

Jeremy:

I had a small project, just 12,000 Ft2. I needed to model a small shell that
was underground and fitted into the other building. It is that small area with
the 6 skylights on the west axis. See attached jpg.

Initially I modeled the main building with exterior walls where this section
fits in and modeled the section with interior walls. I then deleted the
exterior walls on the main building and left the interior walls on the section.

By deleting the exterior walls I reduced the unmet hours from 42 to 39. Energy
consumption reduced from $27,905 to $27,886 (.02%)

Paul Diglio, CEM, CBCP

Paul Diglio's picture
Offline
Joined: 2011-09-30
Reputation: 400

Hey Jeremy,

If I can correctly assume most of these are "sandwiched" walls as you've described them as adjacent shells (thus shading each other from solar gains)... Pretty simply, the impact or "degree of error" such phantom skin loads introduces will vary primarily with how heavily the annualized exterior conditions differ from the conditioned set points inside. More specifically: If it's a building in a very mild & temperate climate (San Diego comes to mind), the extraneous loads introduced will be much less severe than somewhere like Kansas City.

That said, "a pretty big deal" could describe the impact for your project just about anywhere. Particularly if you're making comparisons with/without self-shading, a la LEED.

I agree the fastest approximation to address this in detailed mode is probably to simply delete all 'false' exterior perimeter walls, as suggested. This might actually be MUCH easier using the 2D plan view option over the 3D view however. It'll show you one shell at a time (depending on what's highlighted in the tree), and you can very easily/accurately use the graphic to select the appropriate walls from the component tree (if you're familiar enough with the layout in 2D to recognize which walls go away).

I reckon correcting this may very well help bring your lighting consumption effects to the forefront. If not, the benefits of reducing installed watts should be easy enough to demonstrate with simple math however, outside of an energy model ;).

Regards,

~Nick

[cid:489575314 at 22072009-0ABB]

NICK CATON, P.E.

Nick-Caton's picture
Offline
Joined: 2011-09-30
Reputation: 805

Jeremy,

When two shells are located adjacent to each other the program automatically converts the walls in contact into interior adiabatic walls. Unfortunately the same does not happen to roofs and floors when shells are stacked on top of each other. Roofs have to be adjusted manually (although you can select the roof to be adiabatic in Screen 3). I forget how close the walls have to be for the program to assume they are touching, but I believe it is less than a foot. Click on the 3D view to see the list of surfaces that are behind where you clicked. In the view below the walls of both shells were automatically converted to interior walls.

[cid:image001.jpg at 01CDFA4E.5EA028C0]

In the wizard the larger wall in the example above is labeled as "partially adiabatic." Right click on the wall in Screen 2 to see.

[cid:image005.jpg at 01CDFA4E.5EA028C0]

If the entire wall will be adiabatic (like where a new addition connects to an existing building) you could click on "Entirely Adiabatic Wall" here in the wizard.

Hope this helps!

Keith Swartz, PE, LEED AP, BEMP

Keith Swartz's picture
Offline
Joined: 2011-09-30
Reputation: 0