Fix 10802 - Divide by zero error

[kevin-moos]

Pull request overview

• Fixes Unitary system div by zero in controlCoolingSystemToSP #10802

Description of the purpose of this PR

I was able to recreate this in unit tests for DX coils and variable speed DX coils using values from the original defect file. It looks like PartLoadFrac already defaults to 0 so I just added some divide by zero checks.

Pull Request Author

• Title of PR should be user-synopsis style (clearly understandable in a standalone changelog context)
• Label the PR with at least one of: Defect, Refactoring, NewFeature, Performance, and/or DoNoPublish
• Pull requests that impact EnergyPlus code must also include unit tests to cover enhancement or defect repair
• Author should provide a "walkthrough" of relevant code changes using a GitHub code review comment process
• If any diffs are expected, author must demonstrate they are justified using plots and descriptions
• If changes fix a defect, the fix should be demonstrated in plots and descriptions
• If any defect files are updated to a more recent version, upload new versions here or on DevSupport
• If IDD requires transition, transition source, rules, ExpandObjects, and IDFs must be updated, and add IDDChange label
• If structural output changes, add to output rules file and add OutputChange label
• If adding/removing any LaTeX docs or figures, update that document's CMakeLists file dependencies
• If adding/removing any output files (e.g., eplustbl.*)
  • Update ..\scripts\Epl-run.bat
  • Update ..\scripts\RunEPlus.bat
  • Update ..\src\EPLaunch\ MainModule.bas, epl-ui.frm, and epl.vbp (VersionComments)
  • Update ...github\workflows\energyplus.py

Reviewer

• Perform a Code Review on GitHub
• If branch is behind develop, merge develop and build locally to check for side effects of the merge
• If defect, verify by running develop branch and reproducing defect, then running PR and reproducing fix
• If feature, test running new feature, try creative ways to break it
• CI status: all green or justified
• Check that performance is not impacted (CI Linux results include performance check)
• Run Unit Test(s) locally
• Check any new function arguments for performance impacts
• Verify IDF naming conventions and styles, memos and notes and defaults
• If new idf included, locally check the err file and other outputs

[mitchute]

This looks right.

[rraustad]

There is likely a reason (a developer oversight) that FullOutput has not been calculated. Usually FullOutput is set near the beginning of this function when PLR is set to 1 and the system capacity is checked to see if that capacity exceeds the load, otherwise PLR = 1 and more calculations are not needed. See line 13249, 13681 and 13732. One thing is for sure, if this function is iterating with SolveRoot then the system does have a non-zero FullOutput (or maybe the coil is scheduled off? but I would think that would be caught early). To figure out the why you would need to step through this function at the time the error occurred.

[mitchute]

@rraustad I believe this catches the issue noted, but I'll hold this open for a little bit in case you want to take another quick look for any other higher level issues.

[rraustad]

I added at line 14133:

if (OutletHumRatLS > DesOutHumRat) {
    CycRatio = 1.0;
    this->m_CoolingSpeedNum = std::max(1, this->m_CoolingSpeedNum);  <-- new line since speedNum was 0

    for (int speedNum = this->m_CoolingSpeedNum; speedNum <= this->m_NumOfSpeedCooling; ++speedNum) {

The reason SolveRoot was failing was that a sensible load did not exist and FullOutput was not calculated and remained 0 from initialization (line 12999). There was, however, a latent load when there should NOT have been a latent load given these inputs. Adding the line above did allow SolveRoot to find the correct PLR and no longer fails with SolFla = -2. But there is another logic issue somewhere, I have not found that yet. I guess this change is OK but avoiding this calculation if FullOutput = 0 is not really the correct thing to do (because calling SolveRoot means the system is trying to meet a load and therefore the system should have a non-zero capacity) since if SolveRoot does fail with -2 then PLR will default to 0 and it would be prudent to at least try to find an operating PLR, as faulty as this could be (i.e., trying to find PLR based on PartLoadFrac = ReqOutput / FullOutput).

CoilSystem:Cooling:DX,
  GUID11_SYS0SCC-1System,
  None,   !- Dehumidification Control Type
  Yes,    !- Run on Sensible Load
  No,     !- Run on Latent Load

CoilSystem:Cooling:DX,
  A10,  \field Run on Latent Load
    \type choice
    \key Yes
    \key No
    \default No
    \note If Yes, unit will run if there is a latent load.
    \note even if there is no sensible load.
    \note If No, unit will not run only if there is a latent load.
    \note Dehumidification controls will be active if specified.

[rraustad]

A more important point is that SolveRoot should never fail with -2. That in itself is a developer error.

[rraustad]

Ah, there are 2 CoilSystem objects in this file. One uses humidity control and the other does not. So my comment about a logic issue was incorrect. I would still add the new line in this comment so that SolveRoot does not fail with -2.

CoilSystem:Cooling:DX,
  GUID11_SYS1SCC-1System,
  ,
  GUID11_SYS1SHC-1Outlet Node,
  GUID11_SYS1SCC-1Outlet Node,
  GUID11_SYS1SCC-1Outlet Node,
  Coil:Cooling:DX:VariableSpeed,
  GUID11_SYS1SCC-1,
  CoolReheat,
  Yes,
  Yes,
  Yes,
  2.22222222222222;

CoilSystem:Cooling:DX,
  GUID11_SYS0SCC-1System,
  ,
  GUID11_SYS0SHC-1Outlet Node,
  GUID11_SYS0SCC-1Outlet Node,
  GUID11_SYS0SCC-1Outlet Node,
  Coil:Cooling:DX:VariableSpeed,
  GUID11_SYS0SCC-1,
  None,
  Yes,
  No,
  No,
  2.22222222222222;

[kevin-moos]

Added the new line from this comment

[mitchute]

Added the new line from this comment

@rraustad