The validations should state:

• What are blockers and won't let the simulation run in any form
• What are reasonable upper and lower limits for loads (i.e. LPD 0 W/ft2 - 5 W/ft2)

Based on #57 there doesn't need to be a "true" value that load calcs are compared against. There is however some interest to compare the results produced by Revit compared to the commercial reference buildings.

• Build geometries for commercial reference buildings in Revit
• Translate using out of the box functionality to OpenStudio and compare results
• Customize the measures to be as close to possible as the commercial reference buildings and compare again.

The OpenStudio CLI doesn't currently support different character sets that Autodesk needs (i.e. Chinese characters) and they will need this as their product is international. There is the ITF event at the end of October that I've communciated we won't be able to meet, but would Autodesk University in mid-november be doable? If not, we really need it done by the mid december code freeze.

Check if E+ supports IP units natively for reporting and if it does, add it as a measure to check the attribute in the gbXML file.

Have two dedicated OSWs; one for sizing runs and one for annual simulation.

this may be part of the ci workflow. we may also want to include user acceptance testing in our release pipeline. per noah, main issue is consistency of paths.

Given a focus on sizing, develop a plan in this thread initially for how to appropriately implement passive chilled beams in OpenStudio. There is no radiant vs convective component specified in E+ so if we decide to incorporate a radiant component then it'll have to be a guess based on typical equipment.

EnergyPlus has some restrictions on which configurations of HVAC systems can be on a zone. For example, two air based systems can't serve the same zone. We need to determine what combinations are and are not valid to hand-off to Autodesk to do UI validation.

This can be done by setting the heating coil capacity to 0 after instantiation. Use an electric or furnace coil so there are no further dependencies.

Need to support various character sets to enable support for international users better.

Add a measure that runs at the end of translation to handle any cleanup that may be necessary. There is a short term need for this to fix the interior openingType="air" issue, but others may arise.

Double check that the solution implemeneted for PFPBs and SFPBs behave appropriately and don't leave behind any orphan objects.

Currently the chilled water loop configuration is pretty basic simply using a chiller to provide 44F water. However, the loop should follow the following logic:

• If there are non-radiant components (anything but radiant panels and chilled beams) then supply 44F water
• If there are only radiant components then the loop should only provide 57F water
• If there is a mix, then a higher temp loop using heat exchangers should use the 44F loops to create a loop that supplies 57F water.

Add environment variables to the MSI we generate

The supply air temperature needs to be inferred based on the type of equipment that is present in a zone.

1. Discuss in this thread the proposed logic
2. Create an implementation design in the code
3. Write tests
4. Write code

Help Autodesk perform UI validation by:

• This should indicate which combinations of mechanical systems can be present on the same zone. The strict E+ requirements should be communicated as opposed to limiting it to what makes sense from an engineering standpoint. This may require running various combinations on a one zone E+ model or by some other method.
• We should also indicate any system configurations that produce situations that are nonsensical for simulation. For example, if a space is served by multiple zone equipment (ZE) and all those spaces don't have the same signature of what ZE serve them, then a zone can't be created. Ask Noah for more details on this if needed.

Until mapping is complete, the interim solution relies on the Space Type and Construction Set Wizard measure. This measure doesn't assign constructions for interior glass doors, overhead doors, skylights, tubular daylight dome and diffusers.

It's looking like the default construction information won't be populated in the gbXML, but rather the data sources are going to be provided and we will implement it via a measure. At least, that is what is proposed and we need to investigate it's feasibility and how much effort we think it would be.

At the end of this task we should have an understanding of what is needed to be able to implement applying default constructions based on the string in constructionIdRef, but we should also support overriding for when construction/material information is available in the gbXML.

The current logic and data sources for the implementation can be found here.

Currently it's implemented as an FPFC while QA/QC details of true radiant systems for this purpose are ironed out. The FPFC object requires a heating and a cooling coil, which may not work for models that don't have a chilled or hot water loop. Therefore it would perhaps be better to use a Unit Ventilator object instead.

we may share incremental releases with ad. major releases will be committed to git lfs and pushed.

Change annual simulation OSW from "Annual simulation" workfile to "Annual systems & building energy simulation"

Autodesk said they'd ideally like to support UTF-8 with and without BOM as their clients may run custom OSWs and they may not know what the customer files will be saved as. Tasks:

• Verify that both UTF-8 w and w/o BOM is supported
• If not, provide an estimate of hours and reasonable deadline and create a new task to do so, referencing this one and closing it out.

current manual build using windows advanced installer \ program. will also require signing, but this has physical component and may happen manually after automated build.

Update the valid UI enumerations for each analytical system and send to Liang

The current radiant panel implementation is simply a FPFC with parameters to not use fan energy, which isn't the best assumption for radiant systems when operative temperature is considered. In this thread, propose an implementation before writing tests and finally implementing the code.

The build_installer rake task doesn't seem to override the appropriate fields in the .aip file. The aip file seems to hard code the file locations when it is opened in advanced installer and run preventing updates when the 'build_installer' rake task is run.

Encoding problems could arise using Schedule:File, the multivariable table lookup and the Window5 calculation input. Develop tests for these.

this should run tests on commits / PRs to master and develop branches.

The loop temperature needs to be inferred based on the type of equipment that is on the loop. It is important to note that no objects that are available in Revit (i.e. a chilled water loop) should be created behind the scenes to improve the engineering quality. For example, if high and low temp chilled water components are on the same loop then an additional loop for the higher temp, perhaps served by an HX from the low temp loop, should not be created. Currently the proposed logic for the loops are:

• Hot water loops: Fixed 140F supply with 30F delta T
• Chilled water loop: If a low temp component (i.e. FPFC, air handler cooling coil, etc.) is present then 44F is supplied, otherwise 57F is supplied.
• Condenser loop: 70F fixed supply temp. If a dx heating coil is present then a boiler should be added to keep temp above 50F (not implemented yet)

Steps to complete issue are:

1. Discuss in this thread the final logic
2. Create an implementation design in the code
3. Write tests
4. Write code

Investigate and propose a fix as to why the unit heater test is undersizing the heating coil of the unit heater when it is hydronic.

At the moment, the handoff from Autodesk to OpenStudio is done via gbXML with additional elements to support HVAC systems. However, there is nothing to validate against during hand off. Defining a validation mechanism, perhaps in the form of a schema, would be helpful to ensure specific development changes aren't breaking the handoff.

Currently air systems don't support WC DX units. It simply creates an AC DX unit, but doesn't connect to the condenser loop at the moment

Simulation control fields, by order:
Timesteps per hour = 4
Do Zone Sizing Calculation ? "Yes"
Do System Sizing Calculation ? "Yes"
Do Plant Sizing Calculation ? "No"
Run Simulation for Sizing Periods ? "Yes"
Run Simulation for Weather File Run Periods ? "No"
Shading calcs: frequency = every 27 days, and implementation = "AverageOverDaysInFrequency"
Output control fields: "All"
IDF report Table specification: "InputVerificationandResultsSummary, ZoneComponentLoadSummary"
For annual simulation runs:

Simulation Control fields, by order:
Timesteps per hour = 4
Do Zone Sizing Calculation ?"Yes"
Do System Sizing Calculation ?"Yes"
Do Plant Sizing Calculation ?"No"
Run Simulation for Sizing Periods ?"No"
Run Simulation for Weather File Run Periods ?"Yes"
Shading calcs: frequency = every 27 days, and implementation = "AverageOverDaysInFrequency"
Output control fields: "XMLandHTML"
IDF report Table specification: "AnnualBuildingUtilityPerformanceSummary, SensibleHeatGainSummary"
General adjustments in all modes:

Convergence limits: 4 per hour, Max. HVAC iterations = 10
Building parameters fields customization:
Loads Convergence Tolerance Value = 0.1
Temperature Convergence Tolerance Value {deltaC} = 0.5
Full shading mode? Use "FullExterior" or "MinimalShadowing"
Maximum Number of Warmup Days = 8
Minimum Number of Warmup Days = 3
Output variable fields: currently just set to empty
Design day data definition - currently design day data read in from the epw file, but does not need to be forced
Geometry adjustments in all IDF file parsing modes:

Window processing:
For rectangles: ignore windows with areas < 0.0025 sq. meters
For any polygon with more than 3 vertices:
Iteratively clip to sub-surfaces of simple triangles
Ignore any window triangle area < 0.002525 sq. meters
Surface processing:
Ignore any surface < 0.0025 sq. meters
Scan the MATERIAL:NoMass field(s) to check the minimum resistance value >=0.001

Change the MSI to read "OpenStudio CLI For Revit 2020" to enable side by side installation with Revit.

For objects not currently named in Revit (i.e. reheat coil of a VAV Box), provide a smart name based on it's most relevant parent.

The PTAC object needs to be updated to support the way heating coils are specified now. First step is to update the test model on a newer build of Revit. Then update the tests to support no heating coil before changing the implementation to support heating coils appropriately in addition to a lack of heating coil (set capacity to 0).

There are certain scenarios where the workflow from Revit may need to spin up a subsequent OpenStudio run (i.e. hybrid DOAS systems or baseline automation). A test should be performed to see if there are any issues in doing so. If there are, report to the Autodesk team, if not, pass the test OSW + measures to their team for more extensive testing.

• What are blockers that don't let the simulation run (i.e. a material with no mass)
• What are reasonable upper and lower limit values

Changes for loads-only mode:

Simulation control fields, by order:
Timesteps per hour = 4
Do Zone Sizing Calculation ? "Yes"
Do System Sizing Calculation ? "Yes"
Do Plant Sizing Calculation ? "No"
Run Simulation for Sizing Periods ? "Yes"
Run Simulation for Weather File Run Periods ? "No"
Shading calcs: frequency = every 27 days, and implementation = "AverageOverDaysInFrequency"
Output control fields: "All"
IDF report Table specification: "InputVerificationandResultsSummary, ZoneComponentLoadSummary"
For annual simulation runs:

Simulation Control fields, by order:
Timesteps per hour = 4
Do Zone Sizing Calculation ?"Yes"
Do System Sizing Calculation ?"Yes"
Do Plant Sizing Calculation ?"No"
Run Simulation for Sizing Periods ?"No"
Run Simulation for Weather File Run Periods ?"Yes"
Shading calcs: frequency = every 27 days, and implementation = "AverageOverDaysInFrequency"
Output control fields: "XMLandHTML"
IDF report Table specification: "AnnualBuildingUtilityPerformanceSummary, SensibleHeatGainSummary"
General adjustments in all modes:

Convergence limits: 4 per hour, Max. HVAC iterations = 10
Building parameters fields customization:
Loads Convergence Tolerance Value = 0.1
Temperature Convergence Tolerance Value {deltaC} = 0.5
Full shading mode? Use "FullExterior" or "MinimalShadowing"
Maximum Number of Warmup Days = 8
Minimum Number of Warmup Days = 3
Output variable fields: currently just set to empty
Design day data definition - currently design day data read in from the epw file, but does not need to be forced
Geometry adjustments in all IDF file parsing modes:

Window processing:
For rectangles: ignore windows with areas < 0.0025 sq. meters
For any polygon with more than 3 vertices:
Iteratively clip to sub-surfaces of simple triangles
Ignore any window triangle area < 0.002525 sq. meters
Surface processing:
Ignore any surface < 0.0025 sq. meters
Scan the MATERIAL:NoMass field(s) to check the minimum resistance value >=0.001

in particular, we may want to run tests for windows 7/8/10.

Air systems currently don't have any economization when the majority of air handlers do nowadays. Adjust the behavior to apply an economizer by default. Discuss the proposed implementation here relating to control, lockout temperatures and so on here first. After that, create tests and adjust the implementation.

Speaking with Ian at Autodesk he doesn't seem too worried about having a "truth" to compare against despite the desire from the dev team. Kind of hard to define what's the true sizing value when engineers approach the same problem differently.

Working branch for this feature

• Add these objects to the data model (Revit team):

InfiltrationFlowPerSpace
InfiltrationFlowPerSpaceArea
InfiltrationFlowPerExteriorSurfaceArea
InfiltrationFlowPerExteriorWallArea
InfiltrationFlowAirChangesPerHour
OAFlowPerPerson
OAFlowPerArea
OAFlowPerSpace
OAFlowAirChangesPerHour

• Add methods to advanced_import_gbxml measure Complete
• Add tests to import measure Complete

Add a rake task to call 'openstudio measure --update_all'. The installer should do this for sure so that measures are versioned.

We need to provide to Autodesk a comprehensive list of what is necessary for them to have in their data model in order to enable baseline automation. For example, tagging of spaces with specific space type enumerations. Let's focus on what standards are available in openstudio-standards for now, but it may grow to include the needs of energy codes not currently in the standards.

This will probably require some spiking via modeling and looking at output reports, but the idea is to determine how to handle multiple pieces of HVAC equipment serving a space. For example, a space served by an air system and a WSHP is most likely a WSHP + DOAS with no OA drawn by the WSHP. However, WSHP + baseboard would suggest OA comes from the WSHP and the baseboard is just supplemental heat. Many variations to consider.

Generate test cases for non-ascii character sets. Predominantly focus on Japanese, German and Chinese in the short term

Set OutputControl:Table:Style to output XMLandHTML

Add support for unit ventilators without a heating coil and/or a cooling coil. This object doesn't require a heating and cooling coil so no need to set capacities, simply just don't add the coil if it isn't meant to be present. That said, the coil option needs to be set appropriately (None, Heating, Cooling or HeatingAndCooling)