Allow modeling whole SFA/MF buildings

[shorowit]

Currently OS-HPXML only allows modeling individual dwelling units of a SFA/MF building (including workarounds for MF common spaces adjacent to the unit or units with shared HVAC/DHW/ventilation systems, per ANSI 301).

However, there are use cases where it would be highly beneficial to create whole MF building models. The easiest path from here to there would be to allow a HPXML file where each dwelling unit is described as an individual Building element. The HPXMLtoOpenStudio measure could be updated to create an individual model for each Building element (as it does now) and then merge all the models into a single whole-building model (see example code from URBANopt) at the end. Ideally, the BuildResidentialHPXML measure would also be updated to produce HPXML files like this.

(There could be another use case where someone wants to provide an HPXML in which a single Building element describes the entire MF building -- e.g., 30 units, 3 stories, 60 bedrooms, 20000 sqft, etc. -- but that would be a greater departure from how OS-HPXML works and far more effort.)

Some things to consider for this approach:

• Common spaces and surface adjacency
• Shared/central systems
• Unit multipliers
• Coordination w/ openstudio-standards and other existing products
• Output reporting

[joseph-robertson]

I think your "30 units, 3 stories, 60 bedrooms, 20000 sqft, etc." example is somewhat close to how URBANopt operates. Entire MF building "features" are described at the whole-building level, along with "number of dwelling units represented". The measure supporting the merge of all models starts by scaling all building-level properties down according to the number of dwelling units represented value.

[shorowit]

But URBANopt never ends up with an HPXML file that describes the entire MF building like that -- it fundamentally generates a single HPXML Building element for each dwelling unit.

I'm trying to make the point that it's a not-too-big effort to support the type of HPXML files that URBANopt produces, but if URBANopt were to provide a HPXML file that fundamentally reflected its higher-level inputs, that would be a much bigger effort.

[joseph-robertson]

Yeah I just meant in terms of collecting inputs at the whole-building level, not in terms of storing them in HPXML.

[DavidGoldwasser]

I do like the idea of providing high level description with a a mix of unit types (5 studio, 10 1br, 20br, 5 3br, 3 story, 20000 sqft)

[joseph-robertson]

I've been thinking a bit about this issue. Perhaps one way we could update the BuildResidentialHPXML measure to produce HPXML files with multiple Building elements, would be to introduce a new optional hpxml_path_in argument (or change current hpxml_path argument to hpxml_output_path, and then add new hpxml_path). When provided, the measure appends a new Building element; when not provided the HPXML output file has exactly one Building element. Users of the measure could then apply BuildResidentialHPXML in series to describe a SFA/MF building.

[shorowit]

Work plan:

• @shorowit + @joseph-robertson: Update HPXMLtoOpenStudio measure to support creation of single OSM
• ??: Update ReportSimOutput measure as needed
• @joseph-robertson: Update BuildResHPXML measure to support creation of single HPXML w/ multiple Building elements
• @shorowit + @joseph-robertson: Somehow allow unit multipliers (e.g., E+ zone multipliers w/ additional logic for any gaps, changes to reporting measure, etc.)
• ??: Update ReportUtilityBills as needed (Don't support whole MF buildings? Use average unit consumption for bill calcs? Or do we need consumption for each unit? What about central meters?)
• @shorowit: Allow modeling of common spaces as E+ thermal zones

[shorowit]

• Give a warning if unit multipliers w/ stochastic schedules?

[shorowit]

Nice to have, but not critical:

• Combine unconditioned zones into a single zone for each type (e.g., attic, unconditioned basement, etc.)? Shouldn't affect energy use that much (?) but could improve runtime performance a bit.
• Model single E+ interzonal walls for walls between two conditioned thermal zones (instead of two E+ adiabatic walls)?