Expression Evaluation Error

9 posts / 0 new
Last post

This one has me stumped, and it looks like it's been an issue before, so I'm hoping someone has an idea. I was building a model in the wizard and went to save-as before something got screwy, and it appeared I was too late. I got an error for multiple zones that read "MULTIPLIER in XXXzone has Expression Evaluation Error: Invalid BDL reference", followed up by a bunch of zone errors stating that they were "referenced but not defined". I did some searching and came up with a few hits, but no solutions (that I could see). I bit the bullet and went back and tried to rezone it from my last save, and while the error seems to occur for different zones, I still got it. I went through and tried to purge out some AHU's and spaces that may have pushed me up against some limit I'm not aware of, but nothing's helped so far.

Below is what I'm currently zoning, and you can see that in the building footprint screen, there are zones that I've defined that aren't showing up. I'm attaching the pd2. Any thoughts or ideas would be appreciated!

[cid:image007.jpg at 01CE9F26.1DB79FF0][cid:image008.png at 01CE9F21.73113DD0]

[cid:image011.png at 01CE9F26.1DB79FF0]Eric L. O'Neill, P.E., LEED AP

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

When you are entering your zones try to use an autocad background with few lines and change your properties to polygon 1st cad 2nd and eliminate the zone overlap and gaps you have in your model. Hopefully that helps...
[Capture.PNG]

Thanks,
Matthew J. Baron

Matthew Baron's picture
Offline
Joined: 2013-07-30
Reputation: 0

An HTML attachment was scrubbed...
URL:

Bruce Easterbrook's picture
Joined: 2011-09-30
Reputation: 0

Bruce and Matthew,

My understanding of defining the vertices is that the way they line up only affects how the wall constructions are defined, and shouldn't result in an error. For instance, for zone boundaries defined near an exterior wall where the vertices don't line up exactly, eQUEST may define them as an interior wall, or possibly an exterior wall with no windows (I haven't taken the time to discern what the ruleset is for that). Internal zones that don't line up perfectly are usually defined as adiabatic rather than as an interior wall construction, but the space and zones themselves don't really care if they line up or even overlap. For my purposes, I took the time to do the exterior walls correctly, but it's not worth the additional time to get the interior walls perfect (this isn't for LEED or anything quite to that level of detail) for every single space.

As an example, you can set up and run a brief model on a building like below, and it will fully simulate without errors, despite obvious overlapping floorspaces and walls that don't line up.

Normally I take the time to line things up well, so I could be wrong on this. Am I misunderstanding what you guys are trying to say, or (more importantly) am I missing something wrong with this approach?

I tried Robby's suggestion - I duplicated that shell, and on one copy I deleted the north zones, and the other I deleted the south zones, and the errors have disappeared.

Thanks everyone!

Eric L. O'Neill, P.E., LEED AP

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

I poked around in your file and couldn't see an obvious answer to your
problem. I was able to eliminate the errors for the referenced plenum
spaces being called out but not existing by changing your floor to ceiling
height to match the floor to floor, which essentially deleted the plenum
spaces.

Though I can't say for sure that there is a maximum number of spaces that
can be defined per floor in the wizard, that would be my guess as to what's
happening here. If I were in your shoes I'd break the floor into two
floors at the same elevation and see if that works out. I took a crack at
doing that and it seems to work. I've run into a similar issue where I
exceeded the number of allowed vertices for a single floor polygon and had
to break it in two.

May not be the answer you're looking for. Hope this helps.

[image: Inline image 1]

Robby Oylear's picture
Offline
Joined: 2011-09-30
Reputation: 202

An HTML attachment was scrubbed...
URL:

Bruce Easterbrook's picture
Joined: 2011-09-30
Reputation: 0

Thanks Bruce, I hear ya. In this particular case, this model is for an existing building energy study, where the owner actually has nearly up to date as-built dwgs (!), so I think we're talking apples and oranges. It's a hospital (nearly all clinic and office spaces are out of scope and have been excluded from the model) and it seems that almost all AHU's have spaces that require continuous space pressure relationship requirements, so one of the measures we're looking at is doing some unoccupied airflow reductions in specific areas, which is why I've done a more detailed zoning. However, zones don't currently have temperature setbacks and on this project we're not planning on recommending that, so my gut says there's not enough heat transfer between spaces at similar temperatures to justify spending the time making sure everything lines up perfectly. Bad assumption? Maybe I'll test it out at some point when I get a chance.

Eric L. O'Neill, P.E., LEED AP

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

In that case I would do a major simplification of the zoning based on
the AHU systems. Interior walls inside a zone won't have much impact and
I would leave them out. What I have found with the zones not lining up
properly is in some cases the model will freeze and not even run. Or it
generates pages of errors. Once the points/walls are lined up the
errors go away. What is left is criticaland the noise is gone.
Energy studies are a totally different ball game on existing buildings.
Typically the first thing you will have to do is get your model running
similar to the buildings current performance. You will probably find
that even more difficult. There is lots of discussion on that as well
in the archives. Have fun.
Bruce

Bruce Easterbrook's picture
Joined: 2011-09-30
Reputation: 0

I might round back and reinforce part of Bruce's initial assessment: When shared vertices aren't perfectly locked together all sorts of havoc can ensue during or well after the wizards, and I also see some traces in that original screengrab suggesting *slightly* askew (non-orthagonal) walls which may or may not have been intentional...

Some high-level simplification may well be in order as has been discussed, but should you decide to flesh out a full accounting of the interior zones you might try this strategy: For more complex and large CAD tracings, rather than reference the actual plan backgrounds I will instead first draft up my own simplified perimeter/zone map for the energy model in CAD using the actual HVAC/background xref'ed in for reference. What follows in equest is a much simpler to use single-line reference with a magnitude fewer CAD vertices. This really cuts down on potential for human-error - usually it's plain as day with such a simplified reference when I miss a vertex snap. Such a simplified reference will dramatically cut down on your time spent on those zoning screens as well which I'm sure has mental health benefits.

As always with large projects, save your progress early and often!

~Nick

[cid:489575314 at 22072009-0ABB]

NICK CATON, P.E.

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