Failing u-values after upgrade

[Terence]

Sorry, when I wrote that I was thinking about spaces that are indirectly heated from adjacent heated spaces; I Wasn't thinking about totally enclosed spaces. In that case set the HVAC system to 'None' and the VE will write the rooms out as 'Zones without HVAC' and it will adjust the U-values accordingly.

[jerry]

Terence
In which case, as you say, the U-values are modified and the model fails Criterion 2, since the separating walls between e.g. a heated corridor and an unheated storeroom will not be insulated. It seems the only way to get around this daft situation is to assign an HVAC system to such rooms even though they are not directly conditioned and even though this contradicts the SBEM guidance, that way the separating wall isn't tested against Criterion 2. This didn't used to be a problem when such rooms were set up as "warm roof" as they were written out as conditioned, I don't understand why this needed to change.
Jerry

[Complex Potential]

I only ever use DSM but as far as I know, setting an enclosed perimeter room as a "storage" space with no HVAC system will result in the thermal boundary being defined on the outside wall. The same is true of setting a space type as warm roof. It's only when a space type is defined as something like "unheated buffer" that the internal wall is then subject to the external boundary criterion 2 check.

DSM still appears to work this way but if SBEM is doing otherwise now that sounds a bit odd. I'll admit it has been a while since I checked this specifally in DSM though....

[jerry]

Terence,
I've just tested it with VE_SBEM and VE_DSM
With my store rooms assigned an HVAC system all is fine, i.e. building passes criterion 2, and the BRUKL table reports all the U_values to be unmodified, this is all as I would expect.
If I set the store rooms to system=none and re-run, the building fails criterion 2 and the BRUKL shows much higher average U_values which I assume now include the U-values of the uninsulated walls which separate heated rooms from the adjacent unheated stores rooms.Isn't this what you said would happen ? ("In that case set the HVAC system to 'None' and the VE will write the rooms out as 'Zones without HVAC' and it will adjust the U-values accordingly.")

DSM behaves differently. In both tests, the BRUKL U-value table is the same (i.e. displays the U-values of the envelope) and the building passes criterion 2.

Jerry

[Terence]

This is all down to how the SBEM (not VE-SBEM) calculation handles void spaces. With ApacheSim, a user can say a space inside the buildings thermal envelope is an 'Internal Void or Warm Roof' so Apache uses the correct logic and ensures it is handled correctly for the Criterion 2 check...

Although this option can also be selected when using VE-SBEM it does not mean it is treated in the same manner. When you select this room type selection in SBEM, only the adjacencies of this space to the nearest heated space are considered.

How the model with a void space looks to the user



How the model looks from the SBEM input file. Void space is dropped but its adjacencies are still considered



The SBEM engine accepts the following options when defining wall types for rooms.

Exterior
Strongly ventilated spaces
Unheated adjoining space
Conditioned adjoining space

So when you say a room is 'Internal Void or Warm Roof' one of above wall types has to be applied when describing the adjacencies of the room. A space defined as 'Internal Void or Warm roof' won't be conditioned or strongly ventilated, it won't be 'Exterior' so the only option is 'Unheated adjoining space'.

[PCully]

I had meant to chime in on this myself but looks like Terence has provided a nice summary.

I think in reality if you were working in iSBEM you wouldn't model these rooms this way at all because it provides no such option. We are spoiled by the interface for ApacheSim because it can sensibly handle such a scenario.

I think for SBEM the best advice is to absorb those rooms into nearby conditioned rooms and not have them modelled separately when they are internal? I think the iSBEM User Guide has a few notes on such unheated spaces dealing for exmaple with the case of roofs ("don't model them") and internal spaces such as corridors ("consider these as being conditioned by the same systems as they are adjoining since they will most likely be indirectly heated/cooled by these systems anyway")

Phil

[jerry]

Thanks Phil and Terence for the replies.
iSBEM does allow you to set "Zones without HVAC system", which is the same as setting a room in VE_SBEM, to "System=None", so I don't understand you when you say this isn't possible in iSBEM. I also don't understand why the same building would be handled differently in regard to unheated rooms depending on whether you are using VE_SBEM or VE_DSM as surely the same conventions must apply. Obviously VE_DSM and VE_SBEM analysis a building differently but both write out to SBEM at the end of the day and I would have thought that IES having BRE approval for VE_DSM as a compliance tool would requires that it observes the same conventions as iSBEM.

Jerry

[Terence]

Hi Jerry,

Although both are approved for PartL2, VE-SBEM and DSM are very different calculations. DSM does not use the SBEM engine to carry out its calculations, it uses Apache. They both use BRUKL.exe though which is used after their respective simulations.

With Apache (DSM) it knows to treat a space as being inside a buildings thermal envelope when it is set to "Internal Void or Warm Roof". You may expect SBEM to treat rooms classified as "Zones without HVAC system" the same way but it doesn't. This is because SBEM doesn't have any overall concept of the 3-d geometry, so there can be only one thermal envelope - therefore everything unconditioned is "outside" as far as criterion 2 is concerned.

[jerry]

Thanks Terence, I do understand your explanation as to *why* the Criterion 2 results in the BRUKL reports differ for a building produced by DSM compared to the same building produced by SBEM - but I can't quite accept that it is *correct* that they do differ. To my mind the same compliance "rules" (i.e. the way the simulation tool treats a building for the purposes of compliance should be the same, irrespective of which method (DSM or SBEM) is used to analyse it, I can't see how the alternative can be acceptable to the BRE. Anyway, what this ultimately means is that an assessor has to decide at the outset whether a building is going to be modelled using SBEM or DSM, and build it differently dependant on the answer to that question. If he starts off with the intention to use SBEM on the basis that it is a simple building that doesn't merit full DSM then he will need to absorb any voids such as ceiling voids into the zones otherwise he will be screwed on the Criteria 2 tests. If it subsequently comes to light that he needs to use DSM on the same building then he will need to spend time drawing ceiling voids and altering zone heights etc to accommodate the different approach that DSM takes which leads to a different way of reporting the adjacencies of unheated voids on the BRUKL.
Jerry