Really need some deep DOE-2 theology on this one..

7 posts / 0 new
Last post

I'm with Karen - I don't think you can simply explain this one away to a LEED reviewer and will need to address/correct the issues you've identified

If I'm not mistaken, "adiabatic" isn't an option for exterior surfaces however, at least through the eQuest detailed mode interface. You could make a distinct roof construction which has an impossibly low U-value to severely limit heat transfer (to/from exterior conditions), but that won't correct the roof area figures your reviewer is tuned into.

In my opinion, the easiest reasonable thing to do would to assert (if you can) the thermostat setpoints are the identical between floors and therefore it's reasonable to not model heat transfer between floors, which means you can simply delete those "non-exposed" roofs. Re-defining those surfaces to be adiabatic ceiling/floor (using an interior wall construction) would have the same net effect except to define additional thermal mass for the associated spaces.

The "partially correct" roof surface(s) on the other hand, which are both exposed and sandwiched, ought to be modified equally in both models to correct the exposed portion. How simple/involved this is depends on your project's geometries.

Here's a procedure for a correcting a partially sandwiched, single surface roof. This is done in detailed mode and assumes wizard-based model assembly:

- Navigate to 3D view, highlight the roof surface to be edited

- Open properties window for that surface via the component tree or right-click context menu.

- Identify the referenced space polygon (green text) in the surface properties

- Scroll down to that polygon in the component tree, right click, create a new polygon, copy the one you identified - name it anything you'll easily find in a long list.

- Re-highlight that roof surface in the 3D view and open its properties window

- Re-assign the polygon (green text) to the copied polygon with a unique name (will now be red text). Click Done. 3D view should look like nothing has changed at this point.

o Save a copy of your project separately before proceeding!

- Keeping a clear view of the roof section to be edited, now scroll down and double click on the new copied polygon to pull up its vertices.

- From here, play around with manually removing vertices to identify which consecutive series bounds the sandwiched portion of the surface. You will observe which vertices go away or move in the 3D view behind if you delete/edit these inputs.

- Determine whether you need to define any new vertices, to define the "seam" edge where exposed roof meets exposed wall. If so, experiment with editing a vertex to determine appropriate X/Y coordinates.

- Once you've identified the vertices to remove and add, for eQuest stability reasons I try to edit existing polygon vertices before removing anything from the list (you'll typically end up with fewer vertices than you started). This is hard to put to words but bear with me: From the first vertex, identify the range which should remain (if any), then edit the following vertex points to any new coordinates you may have determined for additional points, then copy the series of vertices to remain from the end of the list up, then finally right-click and restore default to remove all vertices remaining in the list, starting from the bottom.

- For context, I personally do not treat new/edited vertex coordinates super-precisely typically - anything that looks close to right in the 3D view is good enough for LEED. Remember you'll be doing the same edits to both models. Don't get worked up over small seams/holes in the picture - they don't cause roof leaks in in the model!

Procedure for a multi-surface roof (one per space/plenum) is the similar except you do it for each partially sandwiched roof section (don't worry this gets quicker/easier after the first couple). Those fully-sandwiched sections can either be fully deleted per the above logic, or else re-defined as interior floor/ceiling surfaces if deemed necessary.

Until you wrap up such polygon geometry edits, save separately and often, and run a simulation to test everything is working okay. I've had some mixed results doing such edits with eQuest interface open. You may see errors/cautions pop up that can be simply clicked through without consequence.

When all is said and done, one piece of good news is you can save, close, and copy/paste all modified custom polygons between your baseline/proposed models, which makes doing the same edits on the other model as simple as re-assigning polygons (with 3D view up to check your work).

NICK CATON, P.E.

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

Maybe I'm missing something here, but why not just change those roofs that you want
adiabatic to be Interior-walls and then change the INT-WALL-TYPE to be ADIABATIC ?
Wouldn't that take care of both the heat flow problem as well as the area numbers ?

Joe Huang

Joe Huang's picture
Offline
Joined: 2011-09-30
Reputation: 406

Hi Joe,

I briefly addressed that possibility in the highlighted line below. I've done the same in the past, but I think often converting surface types can be more effort than necessary (I'd just delete them if it made sense for the circumstances).

That said, it's no effort to share a link for completeness =). For a procedure detailing how to make an exterior surface an interior surface, here's an illustrated guide I posted a while ago: http://onebuilding.org/wiki/doku.php/convert_an_exterior_surface_to_an_interior_surface_detailed

Most of my reply below is detailing how I deal with partially sandwiched surfaces through the eQuest interface, which requires polygon edits for correction. I don't think I ever shared that previously on the lists...

NICK CATON, P.E.

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

Nick,

I have to keep reminding myself that I'm one of the few oldies left who still works
exclusively with the BDL file and the vi editor.

Joe Huang

Joe Huang's picture
Offline
Joined: 2011-09-30
Reputation: 406

I would recommend deleting the roofs where roofs are actually not and create a floor and a ceiling. In order for two spaces to most accurately model heat transfer between them, each space needs to have a construction at the joining surface.

For example, a single space per floor that is two storey's will need a ceiling construction on the first floor and a floor construction on the second floor.

Also, as mentioned...a construction will need to be created for all spaces that have adjoining surfaces.
Take the example above and make the second storey two spaces, the single space on the first floor will need two separate ceiling constructions as the second floor spaces will have two floor constructions.

This is the kind of stuff they don't talk about at home 'building energy modeling' school. I suppose either way might not matter much for green building rating systems but I have seen it make a difference in unmet hours, usage, etc.

Thanks,
Kevin

Kevin Kyte2's picture
Offline
Joined: 2011-09-30
Reputation: 2

To be clear, for an eQuest/DOE2 model I'm pretty sure only a single surface need be defined to transfer heat between two zones. It can be defined under either space and thermally linked to the other space via the "NEXT-TO" parameter.

I believe defining a surface in both spaces, each linking the other space, for the same real-world construction plane would in effect double the surface area for transfer between the spaces.

In the common scenario of a plenum in between floors (meaning three zones), a minimum of two surfaces (representative of the "ceiling" and "floor/deck" constructions) are required to model heat transfer between all three zones.

I agree there are cases where inter-floor heat transfer cannot be dismissed as negligible, I simply mean to highlight there are scenarios where this case can be made, which can simplify how to address the posed issue of incorrect roof surface areas. As with any consideration of simplification in an energy model, if one is ever unsure it's usually better to err on the side of "more accurate," time allowing. This allows you to explore either option and gain first-hand experience that will help you make such judgements in the future.

You could say I'm one of those self-taught/home-schooled energy modelers however (many of us on these lists are), so please do not hesitate to correct or inform me if I'm missing anything commonly taught in the formal programs! One of the most valuable lessons I've learned to apply to the field of energy modeling, paraphrasing Donald Rumsfeld: There are always things you know, things you don't know, and things you don't know you don't know.

Sincerest regards,

NICK CATON, P.E.

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

Ok, that is right. The INTERIOR-WALL dictionary discussion appears to address this. Maybe its Eplus that needs walls for each space that is adjacent.
Thanks for pointing out the correction.

Thanks,
Kevin

Kevin Kyte2's picture
Offline
Joined: 2011-09-30
Reputation: 2