Interior courtyards

6 posts / 0 new
Last post

Hello all,

I have a particular building geometry that I am unsure of how to simulate. I know that you cannot simulate an exterior space that is completely surrounded by interior zones. In the past when I have come across an interior courtyard type of geometry I was able to conveniently split my shell in two parts with minimal effort with the split occurring along the edges of the interior courtyard.

This time, however, my interior courtyard (which is actually more like a large light well than an interior courtyard) is relatively small compared to my building, but not small enough to ignore, and spans across three floors, including the interior parking garage in the basement. Before taking the time to split my shells like I did in the past, I was wondering if anyone has a more clever way of approaching this particular geometry?

One option I thought of was to separate the spaces surrounding the courtyard on each floor into their own zones and simply adding appropriately sized external walls and fenestration to these zones in detailed mode, while simulating the courtyard itself as non-conditioned. Could it really be so simple?

The attached image is of the second floor of the building. The courtyard is in the red circle.

[cid:image001.jpg at 01D0C307.D13BF7E0]

Jason Ng_ing. jr. M.Sc.A., LEED green associate
jng at bpa.ca | www.bpa.ca | t: 5143833747x2813

Les informations contenues dans le courriel que vous venez de recevoir, y compris les pi?ces jointes, sont destin?es ? l'usage exclusif de la (ou des) personne(s) identifi?e(s) comme destinataires et sont confidentielles. Si vous n'en ?tes pas le destinataire, soyez avis? que tout usage en est interdit. Si vous avez re?u ce courriel par erreur, veuillez le retourner ? l'exp?diteur et le supprimer compl?tement de votre syst?me informatique.

jng's picture
jng
Offline
Joined: 2014-06-20
Reputation: 0

Of course, I forgot to attach my image......

[cid:image001.jpg at 01D0C30D.95886700]

Jason Ng_ing. jr. M.Sc.A., LEED green associate
jng at bpa.ca | www.bpa.ca | t: 5143833747x2813

Les informations contenues dans le courriel que vous venez de recevoir, y compris les pi?ces jointes, sont destin?es ? l'usage exclusif de la (ou des) personne(s) identifi?e(s) comme destinataires et sont confidentielles. Si vous n'en ?tes pas le destinataire, soyez avis? que tout usage en est interdit. Si vous avez re?u ce courriel par erreur, veuillez le retourner ? l'exp?diteur et le supprimer compl?tement de votre syst?me informatique.

jng's picture
jng
Offline
Joined: 2014-06-20
Reputation: 0

Hello Jason,

Great building to try to simulate.

However, real courtyard buildings are very difficult to simulate accurately because, depending on climate, they have the potential to expose the building to (2) or more "weather files", one for the exterior surfaces of the building and one for the "courtyard". In addition, if there are any openings from the exterior through the building into the courtyard, then there can be significant cross ventilation, which further complicates things.

Finally, if the building is in an extreme climate (for example hot and dry), and if there is any vegetation in the courtyard or fountain, then you may also have significant evaporative cooling in the courtyard (i.e., wetting of the porous floor and walls) and in such cases you'll need an industrial strength CFD program coupled with the building thermal simulation, that is capable of tracking air flow and moisture in and around the building and its courtyard...not for the faint of heart.

If there is no interaction between the building and the courtyard, if you don't have evaporative cooling and if you have some measured data, then one could develop an empirical or "inverse simulation" of the courtyard, split the building like a donut and "simulate" the weather condition inside the courtyard for the inner portion of the building and the normal weather file outside the building with the outside portions and then knit the two simulations together with adiabatic, mass-type thermal walls after iterating to converge the solutions.

One additional difficult issue is that many of the traditional courtyard buildings have amounts of thermal mass that go way beyond the capabilities of today's whole-building thermal simulation programs, which can require a rather long period to converge (i.e., weeks or months).

Finally, although there has been some previous work in this area, most creditable efforts attempt to resolve one of more of the above issues.

Jeff

8=! 8=) :=) 8=) ;=) 8=) 8=( 8=) 8=() 8=) 8=| 8=) :=') 8=) 8=?
Jeff S. Haberl, Ph.D.,P.E.inactive,FASHRAE,FIBPSA,......jhaberl at tamu.edu
Professor........................................................................Office Ph: 979-845-6507
Department of Architecture............................................Lab Ph:979-845-6065
Energy Systems Laboratory...........................................FAX: 979-862-2457
Texas A&M University...................................................77843-3581
College Station, Texas, USA, 77843..............................URL:www.esl.tamu.edu
8=/ 8=) :=) 8=) ;=) 8=) 8=() 8=) :=) 8=) 8=! 8=) 8=? 8=) 8=0

Jeff Haberl2's picture
Offline
Joined: 2011-10-02
Reputation: 0

One of the standard shapes in eQUEST DD wizard is a rectangular building
with a courtyard. If you select this, eQUEST creates a ?C? shape building
with the ends of the C very close together and (ideally) set as adiabatic.

You can do something similar with your shell(s). Trace the outside of the
footprint, then cross over at some logical point, trace the interior
courtyard, then go back to the outside and finish the footprint. Be careful
to put the adjacent vertices close together but not one atop the other or
such that the footprint crosses itself. In the wizard screen 2, you can
right click on a wall segment and set it to be adiabatic. You will want to
do this for the walls that are within the footprint.

I hope this helps.

bfountain's picture
Offline
Joined: 2011-09-30
Reputation: 201

Hello Jason [and all];

Maybe I am overlooking an important detail here, but I think Jason?s ?is it so simple? question can be answered ?yes? with some minor clarifications.

Jeff detailed out some of the complications of this situation very well, but I am not sure we need to go that far with it. If this for a code compliance model or LEED performance model, the simplified approach is best. If we are truly attempting to understand this courtyard and the building interaction, or cross ventilation, and such, the Jeff is spot on.

The way I would/have approach a situation such is two part. In Wizard Mode, draw the shell[s] all the way around the building. For zoning leave this area empty/don?t make a zone for it. If this is truly surrounded by a single zone, split this zone into two zones through this hole, some of each zone bordering this ?hole?. This is necessary to actually create the ?hole?. Then, in Detailed edit mode, you add/create your exterior walls/fa?ade/windows/etc. in the ?hole? area. If you had to split a zone into two, take the interior wall that splits these two zone and make it/them an ?air wall?.
I have to use this approach on both smaller and larger ?holes? in buildings with no major issues. Just the time it takes to create the other exterior walls, which adds up on big models.

I?d be interested to hear other?s thoughts or approaches.

Thanks,
DARIC R. ADAIR PE, C.E.M.
Henderson Engineers, Inc.?|?Mechanical Engineer, Energy Analyst
dir 913 742 5530? tel 913 742 5000? fax 913 742 5001? tx id #F-001236? email daric.adair at hei-eng.com? www.hei-eng.com
Licensed in KS.?

daric.adair's picture
Offline
Joined: 2014-06-18
Reputation: 0

I?d add over the past years (in the onebuilding archives), you can find
more discussions and illustrations by searching for the term ?donut.? For
a time it was popular to call these situations donut hole zones! This
illustration is from one such thread for a visual:

Further to Brian?s suggestion, my tendency is to deliberately ?slice? my
building footprint along the lines of some interior zones whose thermal
interactions shouldn?t matter much (i.e. similar temperatures/exposures).
In CAD processing, I?ll draw parallel lines to snap to in eQuest
approximately 1-2 inches apart, for both my zone and perimeter references.

If your zoning is fairly orthogonal, and/or you just forgot during CAD
cleanup (been there!), you can save also ?manually? define the necessary
offset vertices by editing the X/Y coordinates to the left as you?re
otherwise defining your footprint/zone vertices by snapping.

The time-saving advantage to this approach overall is you can automate
creation of the interior ?donut? exterior surfaces (exposed to ambient
conditions) and of course any windows involved. When done with wizards, a
first step in detailed mode is to delete the resulting ?sandwiched? walls
along the donut ?slice.?

Regarding ?microclimates? as can occur with large atriums/courtyards:
Jeff?s post does a good job of summarizing some of the limits to be aware
of. Conditions leading to ?microclimates? are a topic that sits squarely
on/beyond the edge of today?s building simulation development, and
eQuest/DOE-2 notably does not perform much of the thermal massing & CFD
analysis that seems necessary to accurately simulate the hourly thermal
characteristics for some such cases.

There are many situations where an hourly simulation such as with
eQuest/DOE-2 could really use a second weather file. These include dense
cityscapes where direct solar loads are eliminated/magnified based on your
neighbor?s glazing, semi-exposed exterior bazaars/shopping malls with lots
of thermal mass to keep corridors cool, and unconditioned gymnasiums in
Florida where I lost many pounds of water weight playing dodgeball!

I don?t think there is any single software package that is today claiming
to squarely and directly address all the variables of concern within the
simulation (though I believe EnergyPlus and IES-VE are getting there, among
other problem-specific tools), but it?s very healthy to recognize where the
science of simulation is still actively growing so that you can best
evaluate/temper your results when there are ?known unknowns.?

~Nick

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

*Caton Energy Consulting*
603 N Ferrel

Olathe, KS 66061

office: 785.410.3317

www.catonenergy.com

*From:* Equest-users [mailto:equest-users-bounces at lists.onebuilding.org] *On
Behalf Of *Daric Adair
*Sent:* Tuesday, July 21, 2015 3:37 PM
*To:* Jeff Haberl; Ng, Jason; equest-users at lists.onebuilding.org
*Subject:* Re: [Equest-users] Interior courtyards

Hello Jason [and all];

Maybe I am overlooking an important detail here, but I think Jason?s ?is it
so simple? question can be answered ?yes? with some minor clarifications.

Jeff detailed out some of the complications of this situation very well,
but I am not sure we need to go that far with it. If this for a code
compliance model or LEED performance model, the simplified approach is
best. If we are truly attempting to understand this courtyard and the
building interaction, or cross ventilation, and such, the Jeff is spot on.

The way I would/have approach a situation such is two part. In Wizard Mode,
draw the shell[s] all the way around the building. For zoning leave this
area empty/don?t make a zone for it. If this is truly surrounded by a
single zone, split this zone into two zones through this hole, some of
each zone bordering this ?hole?. This is necessary to actually create the
?hole?. Then, in Detailed edit mode, you add/create your exterior
walls/fa?ade/windows/etc. in the ?hole? area. If you had to split a zone
into two, take the interior wall that splits these two zone and make
it/them an ?air wall?.

I have to use this approach on both smaller and larger ?holes? in buildings
with no major issues. Just the time it takes to create the other exterior
walls, which adds up on big models.

I?d be interested to hear other?s thoughts or approaches.

Thanks,

DARIC R. ADAIR PE, C.E.M.
*Henderson Engineers, Inc. | *Mechanical Engineer, Energy Analyst
*dir* 913 742 5530 *tel* 913 742 5000 *fax* 913 742 5001 *tx id* #F-001236
*email* daric.adair at hei-eng.com *www.hei-eng.com*
Licensed in KS.

*From:* Equest-users [mailto:equest-users-bounces at lists.onebuilding.org] *On
Behalf Of *bfountain at greensim.com
*Sent:* Tuesday, July 21, 2015 12:31 PM
*To:* 'Ng, Jason'
*Cc:* equest-users at lists.onebuilding.org
*Subject:* Re: [Equest-users] Interior courtyards

One of the standard shapes in eQUEST DD wizard is a rectangular building
with a courtyard. If you select this, eQUEST creates a ?C? shape building
with the ends of the C very close together and (ideally) set as adiabatic.

You can do something similar with your shell(s). Trace the outside of the
footprint, then cross over at some logical point, trace the interior
courtyard, then go back to the outside and finish the footprint. Be careful
to put the adjacent vertices close together but not one atop the other or
such that the footprint crosses itself. In the wizard screen 2, you can
right click on a wall segment and set it to be adiabatic. You will want to
do this for the walls that are within the footprint.

I hope this helps.

*From:* Jeff Haberl [mailto:jhaberl at tamu.edu ]
*Sent:* Monday, July 20, 2015 5:20 PM
*To:* Ng, Jason; equest-users at lists.onebuilding.org
*Subject:* Re: [Equest-users] Interior courtyards

Hello Jason,

Great building to try to simulate.

However, real courtyard buildings are very difficult to simulate accurately
because, depending on climate, they have the potential to expose the
building to (2) or more ?weather files?, one for the exterior surfaces of
the building and one for the ?courtyard?. In addition, if there are any
openings from the exterior through the building into the courtyard, then
there can be significant cross ventilation, which further complicates
things.

Finally, if the building is in an extreme climate (for example hot and
dry), and if there is any vegetation in the courtyard or fountain, then you
may also have significant evaporative cooling in the courtyard (i.e.,
wetting of the porous floor and walls) and in such cases you?ll need an
industrial strength CFD program coupled with the building thermal
simulation, that is capable of tracking air flow and moisture in and around
the building and its courtyard?not for the faint of heart.

If there is no interaction between the building and the courtyard, if you
don?t have evaporative cooling and if you have some measured data, then one
could develop an empirical or ?inverse simulation? of the courtyard, split
the building like a donut and ?simulate? the weather condition inside the
courtyard for the inner portion of the building and the normal weather file
outside the building with the outside portions and then knit the two
simulations together with adiabatic, mass-type thermal walls after
iterating to converge the solutions.

One additional difficult issue is that many of the traditional courtyard
buildings have amounts of thermal mass that go way beyond the capabilities
of today?s whole-building thermal simulation programs, which can require a
rather long period to converge (i.e., weeks or months).

Finally, although there has been some previous work in this area, most
creditable efforts attempt to resolve one of more of the above issues.

Jeff

8=! 8=) :=) 8=) ;=) 8=) 8=( 8=) 8=() 8=) 8=| 8=) :=') 8=) 8=?
Jeff S. Haberl, Ph.D.,P.E.inactive,FASHRAE,FIBPSA,......jhaberl at tamu.edu
<........jhaberl at tamu.edu>
Professor........................................................................Office
Ph: 979-845-6507
Department of Architecture............................................Lab
Ph:979-845-6065
Energy Systems Laboratory...........................................FAX:
979-862-2457
Texas A&M
University...................................................77843-3581
College Station, Texas, USA, 77843..............................URL:
www.esl.tamu.edu
8=/ 8=) :=) 8=) ;=) 8=) 8=() 8=) :=) 8=) 8=! 8=) 8=? 8=) 8=0

*From:* Equest-users [mailto:equest-users-bounces at lists.onebuilding.org
] *On Behalf Of *Ng, Jason
*Sent:* Monday, July 20, 2015 4:01 PM
*To:* equest-users at lists.onebuilding.org
*Subject:* [Equest-users] Interior courtyards

Of course, I forgot to attach my image??

[image: cid:image001.jpg at 01D0C3C7.9449B0F0]

*Jason* *Ng**_ing. jr. M.Sc.A., LEED green associate*

jng at bpa.ca | www.bpa.ca | t: 5143833747x2813

Les informations contenues dans le courriel que vous venez de recevoir, y
compris les pi?ces jointes, sont destin?es ? l?usage exclusif de la (ou
des) personne(s) identifi?e(s) comme destinataires et sont confidentielles.
Si vous n?en ?tes pas le destinataire, soyez avis? que tout usage en est
interdit. Si vous avez re?u ce courriel par erreur, veuillez le retourner ?
l?exp?diteur et le supprimer compl?tement de votre syst?me informatique.

AT:2|

------------------------------

*From:* Equest-users [mailto:equest-users-bounces at lists.onebuilding.org] *On
Behalf Of *Ng, Jason
*Sent:* Monday, July 20, 2015 3:57 PM
*To:* equest-users at lists.onebuilding.org
*Subject:* [Equest-users] Interior courtyards

Hello all,

I have a particular building geometry that I am unsure of how to simulate.
I know that you cannot simulate an exterior space that is completely
surrounded by interior zones. In the past when I have come across an
interior courtyard type of geometry I was able to conveniently split my
shell in two parts with minimal effort with the split occurring along the
edges of the interior courtyard.

This time, however, my interior courtyard (which is actually more like a
large light well than an interior courtyard) is relatively small compared
to my building, but not small enough to ignore, and spans across three
floors, including the interior parking garage in the basement. Before
taking the time to split my shells like I did in the past, I was wondering
if anyone has a more clever way of approaching this particular geometry?

One option I thought of was to separate the spaces surrounding the
courtyard on each floor into their own zones and simply adding
appropriately sized external walls and fenestration to these zones in
detailed mode, while simulating the courtyard itself as non-conditioned.
Could it really be so simple?

The attached image is of the second floor of the building. The courtyard is
in the red circle.

*Jason* *Ng**_ing. jr. M.Sc.A., LEED green associate*

jng at bpa.ca | www.bpa.ca | t: 5143833747x2813

Les informations contenues dans le courriel que vous venez de recevoir, y
compris les pi?ces jointes, sont destin?es ? l?usage exclusif de la (ou
des) personne(s) identifi?e(s) comme destinataires et sont confidentielles.
Si vous n?en ?tes pas le destinataire, soyez avis? que tout usage en est
interdit. Si vous avez re?u ce courriel par erreur, veuillez le retourner ?
l?exp?diteur et le supprimer compl?tement de votre syst?me informatique.

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