Building Geometry

12 posts / 0 new
Last post

Hello,

Can someone please help me with this:

I have an AutoCAD file of an old home that was retrofitted with thick
(insulated walls). When I import this file and trace it to define the
building footprint on eQuest,should I be tracing the OUTSIDE of the
building (the brick face) or the INTERIOR of the building (the interior
walls)? or should I draw a line in the middle of them and use that as my
building footprint?

do you know the wall construction goes INWARD or OUTWARD from the building
footprint you set?

thanks,

maria

Maria Elisa Rumeo's picture
Joined: 2014-07-23
Reputation: 0

It should not make that much of difference unless your walls are rally thick. I suggest you trace the outside wall and then make sure of the wall thermal properties.
Farid

Farid Pour's picture
Offline
Joined: 2014-12-15
Reputation: 0

Hello,

It is a low-energy retrofit and the walls are really thick!
My fear is that by using the outside dimensions, then if the wall
contruction goes 'outward' of these defined lines, then I am modelling a
larger house ( the volume of the house in the model is greater than the
actual volume) .

please help :)!

thanks,

maria

Maria Elisa Rumeo's picture
Joined: 2014-07-23
Reputation: 0

As far as I know, it goes in.

Farid Pour's picture
Offline
Joined: 2014-12-15
Reputation: 0

Does anyone have experience with modelling thick-wall retrofits who can
also answer my question?
It is a small house, so if I trace the interior dimensions I will get
significantly different results than tracing exterior dimensions

Maria Elisa Rumeo's picture
Joined: 2014-07-23
Reputation: 0

Hi Maria,

We have done one university block recently which had 2.5 feet thick walls (retrofit) and we modeled using inner dimensions. Whether it?s thick or thin, its advised to use inner dimensions although project requirements vary sometimes. Thanks!

Harshul Singhal, LEED AP BD+C
Project Consultant
Thornton Tomasetti
386 Fore Street, Suite 401
Portland, ME 04101
T +1.207.245.6060 F +1.207.245.6061
D +1.207.245.6074
HSinghal at ThorntonTomasetti.com
www.ThorntonTomasetti.com

Harshul's picture
Offline
Joined: 2014-03-25
Reputation: 0

Inside edge is better in your case. Just make sure you model exterior
constructions using the layers input, so you account for thermal mass.

(Larger projects can be modeled with outside edge of walls so your modeled
area matches what's declared as total building area in the architectural
drawings, or in other LEED credits, as applicable).

CJ
On Mar 12, 2015 1:05 PM, "Singhal, Harshul"
wrote:

Charudatta Joshi's picture
Joined: 2015-03-07
Reputation: 0

The energy model is idealized ? there is no thickness to the walls.

Typically, modeling to the glazing line is preferred. Net leasable area is to the window line unless there perimeter wall HVAC units, induction units for example.

In the case of thick walls, modeling to the window line penalizes the baseline case as the lighting and equipment power densities are per unit area. For this reason, I would model this project to the inside of the exterior walls and to the centre line of interior walls.

Christopher Jones, P.Eng.
Tel: 416.644.4226 ? Toll Free: 1.888.425.7255 x 527

Chris Jones
crollinjones's picture
Offline
Joined: 2013-12-12
Reputation: 0

Hi everyone,

I have tread this path a few times before, and my position on the matter
has evolved over time. Here?s my current understanding of the issue:

1. Tracing the outside will result in:

a. Inaccurately large interior volumes and

b. Inaccurately large perimeter space areas (pertinent for
area-dependent inputs like LPD, plug loads).

2. Tracing the inside will result in:

a. Inaccurately small exterior surface areas for the purposes of
determining skin loads.

3. Tracing the middle of the construction results in ALL the
inaccuracies above, but to a lesser degree.

I consider the implications between these approaches on your baseline model
for Appendix G / LEED / etc. to be arbitrary in nature. Your choice should
reflect the degree of accuracy desired relative to the actual/Proposed case
and the time you have to invest in your results.

[TIER 1 EFFORT] You could make the case that #1 or #2 is the lesser of two
evils, considering the implications (and you might be right either way
depending on the circumstances of the project). #3 is a compromise between
#1 & #2, and is my first suggestion if you?d rather not get bogged down in
such hypotheticals AND need to perform a fast study, acknowledging the
above inaccuracies.

[TIER 2 EFFORT] However, a more correct/accurate approach is to put slot
yourself into either camp #1 or #2 and then make the necessary corrections
in detailed mode (post-wizards) for the known/stated inaccuracies with that
approach.

Of the two, I personally think #1: tracing the outside and making
corrections for space volume and lighting/rcpt inputs, is the more
intuitive approach and has the (subjective) advantage of a model that will
look correct in 3D view. Others will trend towards the second approach,
and I wouldn?t fault them for it.

[TIER 3 EFFORT ? HERE BE DRAGONS!]

This isn?t really part of the original ?black/white? question, but it seems
worthwhile to insert a heads up:

To first speak to what?s happening ?under the hood,? I believe it?s most
correct to say eQUEST/DOE2 doesn?t actually ?simulate? the thickness of a
construction (i.e. ?outward-in? vs. ?inward-out?), rather the construction
and heat transfer is 1-dimensional. (I swear, this is rather pertinent for
a thick-wall study ? bear with me!) While we DO enter thickness of
materials/layers for exterior constructions in eQUEST, those inputs are
used only to determine the construction?s thermal mass and conductivity.
This means each hour of the simulation, heat is determined to be moving
either in or out, perpendicular to the surface. Heat is not simulated to
move ?sideways? within the construction (as actually happens at building
corners and where materials change along/within the assembly, such as with
stud wall assemblies) ? that would be what people are referring to when
they speak of 2-dimensional heat transfer. Usually, we make efforts to
define our construction layers in uniform terms (i.e. for a stud/insulation
layer), and call it a day.

So what does all this mean for a thick wall study? (Consider how heat
would travel through the cross-section of a vertical building corner?) It
means the ?effective? 1-dimensional exterior surface area for a space with
skin loads actually lies somewhere between options #1 and #2, above. A
?most correct? eQUEST model might most closely resemble approach #3, above.

I will insert here: I have done some homework in the way of self-educating
to read up on the topic & attend related seminars, and I even passed a
thermodynamics course back in college, but I do not presently consider
myself an expert on multi-dimensional heat transfer in building assemblies,
and haven?t had a ton of experience putting this to practice. I do know
others on this list are more solidly situated in those camps and are
welcome/encouraged to fill in gaps & participate!

So take approach #3 as a framework: To arrive at a ?most-correct? study
using eQUEST (and it may be prudent at this stage to instead consider
software with 2D/3D heat transfer baked-in): Rather than arbitrarily
define walls along the centerlines of your envelope, I would propose
estimating at what mean construction thickness (between the interior and
exterior thicknesses) you would achieve the same result in 1-dimensional
transfer as you would with a more sophisticated 2D (or even 3D) analysis.
You would still need to later make corrections for interior volume and
space-dependent internal loads, as suggested under approach #1.

Procedurally, I would not be surprised to learn others have come up with
tools or a rule of thumb to estimate such an ?effective 1-dimensional
thickness,? accounting for the surface perimeter:area ratio in addition to
the usual conductive variables. You might also use a separate
study/software to externally determine what this effective thickness should
be. Again, I haven?t tread this path in practice, and am hopeful others
may chime in with suggestions along these lines.

I hope this was illuminating/helpful to some (even if only to help you fall
asleep!),

~Nick

*NICK CATON, P.E.*
*Owner*

*Caton Energy Consulting*
1150 N. 192nd St., #4-202

Shoreline, WA 98133
office: 785.410.3317

www.catonenergy.com

*From:* Equest-users [mailto:equest-users-bounces at lists.onebuilding.org] *On
Behalf Of *Charudatta Joshi
*Sent:* Thursday, March 12, 2015 1:21 PM
*To:* Singhal, Harshul
*Cc:* Equest-users
*Subject:* Re: [Equest-users] Building Geometry

Inside edge is better in your case. Just make sure you model exterior
constructions using the layers input, so you account for thermal mass.

(Larger projects can be modeled with outside edge of walls so your modeled
area matches what's declared as total building area in the architectural
drawings, or in other LEED credits, as applicable).

CJ

On Mar 12, 2015 1:05 PM, "Singhal, Harshul"
wrote:

Hi Maria,

We have done one university block recently which had 2.5 feet thick walls
(retrofit) and we modeled using inner dimensions. Whether it?s thick or
thin, its advised to use inner dimensions although project requirements
vary sometimes. Thanks!

*Harshul Singhal, LEED AP BD+C*
Project Consultant
Thornton Tomasetti
386 Fore Street, Suite 401
Portland, ME 04101
*T* +1.207.245.6060 *F* +1.207.245.6061
*D* +1.207.245.6074
HSinghal at ThorntonTomasetti.com
www.ThorntonTomasetti.com

*From:* Equest-users [mailto:equest-users-bounces at lists.onebuilding.org] *On
Behalf Of *Maria Elisa Rumeo
*Sent:* Thursday, March 12, 2015 2:51 PM
*To:* Farid Pour
*Cc:* equest-users at lists.onebuilding.org
*Subject:* Re: [Equest-users] Building Geometry

Does anyone have experience with modelling thick-wall retrofits who can
also answer my question?

It is a small house, so if I trace the interior dimensions I will get
significantly different results than tracing exterior dimensions

Nicholas Caton's picture
Offline
Joined: 2014-12-09
Reputation: 0

Maria,

The walls and other surfaces in the energy model are simple planes with no thickness. You can enter thicknesses of components like insulation, but the program just multiplies that thickness times the material R-value per inch as part of its calculation of the effective R-value (or U-value) of the assembly. The heat transfer calculation is in one direction perpendicular to the plane.

I often locate the exterior wall lines along the outside surface since gross square footage is usually based on the outside surfaces of the building.

For interior walls it would be ideal to go down the middle, but there is rarely anything to snap to and I don?t think it?s worth the time to create those points to snap to. I often just pick a side, whichever happens to be most convenient. I also try to keep straight lines as long as possible, like picking an average location for a long corridor with a small jog in it.

Sincerely,
Keith Swartz, PE, BEMP, LEED AP
Senior Energy Engineer | Energy Center of Wisconsin | Madison.Chicago.Minneapolis
608.210.7123 | www.ecw.org

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

Good morning, all,

The information provided regarding the modeling implications for this issue seem right, but the definitions in Section 3 are always a good place to find answers, although it tends to be a game of connect-the-dots.

In this case, Sections 9.5.1(b) and 9.6.1(b) both use the same terminology, gross floor area and Section 3 defines the gross floor area as being measured from the exterior faces of exterior walls. For interior walls, the area is measured to the center of the partition wall. Using the area from the interior face of the walls would actually underestimate the Baseline lighting power allowance in this instance. J

Have a great day!

Cam Fitzgerald

Energy Opportunities/a 7group company

Cam Fitzgerald's picture
Offline
Joined: 2011-09-30
Reputation: 0

Here are my two cents:

If you want to obtain valid results that also look good in the 3D drawing, specify outside surfaces to calculate skin loads and scale the inside loads (lights, equipment, infiltration, etc.) to the inner wall dimensions. Specify window setback if they are not at the outside planes of their walls. I do not think that LEED cares about the corner effects, but if you want to get the wall masses correct you can sacrifice the beauty of the 3D drawing by leaving gaps. Just subtract half the thickness of the adjacent corner wall from the actual wall width, leaving the wall to start at the original position, and you get one gap; all you have to change is wall Width. If your wall spans the full building width, you must subtract the two corner wall half thicknesses.

DOE-2 fails to subtract thickness (which is not always specified in the input) from the outer dimensions to calculate internal volumes. Remember that thermal boundaries (building shell) are assumed by DOE-2 to face outward, where the default tilt for walls is 90 degrees, horizontal roofs is zero and floors is 180. Material layers, if user specified, must be entered in the order of outside to inside.

Best regards
for KEMA Services Inc., USA

Glenn Haynes, PE
Senior Engineer, Policy Advisory and Research
DNV GL - Energy

E-mail glenn.haynes at dnvgl.com
DNV KEMA is now DNV GL
Mobile +1 615 856 3557 | Direct +1 615 577 2705
www.dnvgl.com | LinkedIn

[DNV GL]
DNV and GL have merged to form DNV GL
We are now the world's largest ship and offshore classification society, the leading technical advisor to the global oil and gas industry, and a leading expert for the energy value chain including renewables and energy efficiency. We've also taken a position as one of the top three certification bodies in the world. Read more here: www.dnvgl.com/merger.

Haynes, Glenn2's picture
Offline
Joined: 2015-03-13
Reputation: 0