Fixing cooling mode cooling coil cooling rate much higher than heat pump cooling rate and OU capacity #10557

yujiex · 2024-06-11T00:56:30Z

Pull request overview

Fixes VRF fluid control model cooling rate at terminal unit is much higher than cooling rate at system level #10187
Fixes sensible + latent output doesn't match total output energy in VRF #10607

For both the original and the fluid-control VRF model, terminal unit capacity should have been constrained by condenser capacity, where a maximum allowable terminal unit capacity is calculated in LimitTUCapacity(.) for each VRF system. The maximum allowable terminal unit capacity is stored in state.dataHVACVarRefFlow->MaxCoolingCapacity(VRFCond) and passed into SimDXCoil(.) which is NOT passed into CalcVRFCoolingCoil_FluidTCtrl(.), as compared with the non-fluid-control VRF model, CalcVRFCoolingCoil(.)

This PR will add an optional argument MaxCoolCap to CalcVRFCoolingCoil_FluidTCtrl(.) as is in CalcVRFCoolingCoil(.) and use this quantity to constrain the cooling coil capacity.

The following shows the effect of the fix: before the fix, cooling coil cooling rate is much higher than heat pump cooling rate; after the fix, the two cooling rate are similar. The test file used here is the US+SF+CZ4A+hp+crawlspace+IECC_2006_VRF.idf example file, using NY weather epw input. The "Rated Evaporative Capacity" of the "AirConditioner:VariableRefrigerantFlow:FluidTemperatureControl" is adjusted to 500W.

Regression diff

The following files have regression diff

DOASDXCOIL_wADPBFMethod
DOASDXCOIL_wADPBFMethod_NoReturnPath
DOAToVRF
HVACTemplate-5ZoneVRF
US+SF+CZ4A+hp+crawlspace+IECC_2006_VRF
VRFMultispeedFan
VariableRefrigerantFlow_5Zone
VariableRefrigerantFlow_5Zone_wAirloop
VariableRefrigerantFlow_FluidTCtrl_5Zone
VariableRefrigerantFlow_FluidTCtrl_HR_5Zone
VariableRefrigerantFlow_FluidTCtrl_wSuppHeater_5Zone
VariableRefrigerantFlow_wSuppHeater_5Zone
_DOASDXCOIL_wUserSHRMethod

ESO difference

The following files have ESO diffs in Zone VRF Air Terminal Total(Latent) Cooling(Heating) Rate. These are expected as a result of the changes in the total and latent system output calculation which fixed the problem of total != sensible + latent output.

DOASDXCOIL_wADPBFMethod
DOASDXCOIL_wADPBFMethod_NoReturnPath
DOAToVRF
HVACTemplate-5ZoneVRF
VRFMultispeedFan
VariableRefrigerantFlow_5Zone
VariableRefrigerantFlow_5Zone_wAirloop
VariableRefrigerantFlow_wSuppHeater_5Zone
_DOASDXCOIL_wUserSHRMethod

The 4 files with fluidTCtrl VRF model (
US+SF+CZ4A+hp+crawlspace+IECC_2006_VRF, VariableRefrigerantFlow_FluidTCtrl_5Zone, VariableRefrigerantFlow_FluidTCtrl_HR_5Zone, VariableRefrigerantFlow_FluidTCtrl_wSuppHeater_5Zone
) have difference in many more fields as follows. The core difference is the coil cooling rate and the VRF air terminal cooling rate. The cooling coil cooling rate difference (due to a capacity limit added in this feature) will also cause difference in compressor side cooling output, such as cooling rate, cooling electricity rate, etc. Piping correction is also affected as it's calculated as (coil cooling rate) / (coil cooling rate + piping loss). The coil cooling rate difference will also cause the temperature and air flow difference in corresponding nodes. The coil cooling rate difference will affect zone temperature as well (In the extreme case, when OU capacity is very small, the zone might not be able to control to setpoint temperature any more. This won't happen if coil has unlimited capacity like before the fix). Zone temperature difference would result in slight difference in heating coil and supplemental heating coil heating rate as well.

Zone Air Temperature
Zone Air Relative Humidity
Zone Air Temperature
Zone Air Relative Humidity
Zone Air System Sensible Heating Rate
Zone Air System Sensible Cooling Rate
Zone Predicted Sensible Load to Setpoint Heat Transfer Rate
Zone Predicted Sensible Load to Heating Setpoint Heat Transfer Rate
Zone Predicted Sensible Load to Cooling Setpoint Heat Transfer Rate
Fan Electricity Rate
Cooling Coil Total Cooling Rate
Cooling Coil Sensible Cooling Rate
Cooling Coil Latent Cooling Rate
Cooling Coil Runtime Fraction
Zone VRF Air Terminal Cooling Electricity Rate
Zone VRF Air Terminal Cooling Electricity Energy
Zone VRF Air Terminal Total Cooling Rate
Zone VRF Air Terminal Sensible Cooling Rate
Zone VRF Air Terminal Latent Cooling Rate
Zone VRF Air Terminal Heating Electricity Rate
Zone VRF Air Terminal Heating Electricity Energy
Zone VRF Air Terminal Total Heating Rate
Zone VRF Air Terminal Sensible Heating Rate
Zone VRF Air Terminal Latent Heating Rate
VRF Heat Pump Total Cooling Rate
VRF Heat Pump Total Heating Rate
VRF Heat Pump Cooling Electricity Rate
VRF Heat Pump Cooling Electricity Energy
VRF Heat Pump Heating Electricity Rate
VRF Heat Pump Heating Electricity Energy
VRF Heat Pump Compressor Electricity Rate
VRF Heat Pump Compressor Rotating Speed
VRF Heat Pump Outdoor Unit Condensing Temperature
VRF Heat Pump Outdoor Unit Evaporating Temperature
System Node Standard Density Volume Flow Rate
System Node Wetbulb Temperature
System Node Standard Density Volume Flow Rate

EIO difference

EIO difference are in the four idfs with FluidTCtrl model. There are some warmup differences in each zone in the model. This is expected as the coil cooling rate difference will change the zone temperature, which affects the warmup convergence.

Meter diff

Meter diffs happen in two of the idfs with VRF FluidTCtrl object. It is expected as the cooling coil cooling rate difference will also affect its electricity rate, and corresponding HVAC, Facility electricity rate part of which is coil electricity rate.

Table string diffs

These table diffs happen in the four idfs with VRF FluidTCtrl objects. The diffs are in warmup statistics and electricity rate, which are explained in the eio diff and meter diff sections.

NOTE: ENHANCEMENTS MUST FOLLOW A SUBMISSION PROCESS INCLUDING A FEATURE PROPOSAL AND DESIGN DOCUMENT PRIOR TO SUBMITTING CODE

Pull Request Author

Add to this list or remove from it as applicable. This is a simple templated set of guidelines.

Reviewer

This will not be exhaustively relevant to every PR.

Follow the heating mode case to use the air flow rate calculated from the ControlVRFIUCoil, this makes coil cooling rate bounded by the OU capacity, although the heat pump cooling rate and the coil cooling rate are still not exactly the same. Piping correction doesn't need to bound the load either, similar to the case in heating mode

…TUCap

rraustad · 2024-06-27T23:04:32Z

src/EnergyPlus/DXCoils.cc

@@ -231,7 +231,7 @@ void SimDXCoil(EnergyPlusData &state,
        CalcDXHeatingCoil(state, DXCoilNum, PartLoadRatio, fanOp, AirFlowRatio, MaxCap);
    } break;
    case HVAC::CoilVRF_FluidTCtrl_Cooling: {
-        CalcVRFCoolingCoil_FluidTCtrl(state, DXCoilNum, HVAC::CompressorOp::On, FirstHVACIteration, PartLoadRatio, fanOp, CompCycRatio, _, _);
+        CalcVRFCoolingCoil_FluidTCtrl(state, DXCoilNum, HVAC::CompressorOp::On, FirstHVACIteration, PartLoadRatio, fanOp, CompCycRatio, _, _, MaxCap);


I'm not sure why the max capacity is not always passed into the coil. It's either 1.0e+20 or it's a value. That's for another day.

I just changed MaxCoolCap to regular non-optional arguments. @rraustad

rraustad · 2024-06-27T23:11:18Z

src/EnergyPlus/HVACVariableRefrigerantFlow.cc

@@ -11374,9 +11374,6 @@ void VRFCondenserEquipment::CalcVRFCondenser_FluidTCtrl(EnergyPlusData &state)
        if (Q_c_TU_PL > CompEvaporatingCAPSpdMax) {
            // Required load is beyond the max system capacity

-            Q_c_TU_PL = CompEvaporatingCAPSpdMax;
-            TU_CoolingLoad = CompEvaporatingCAPSpdMax;
-            this->TUCoolingLoad = TU_CoolingLoad;


Should TU_CoolingLoad be reset here or not? This var is used at line 11385 below. For a VRF system with 1 TU this might make sense but for multiple TUs this will likely have no impact.

It shouldn't be reset here. Similar to the situation in the heating side, which we fixed a while ago: #10154
I copied in the relative discussion here:

This upper bounding will cause the compressor to reduce its output at more extreme weather, which is bad. So this upper bounding needs to be removed.

rraustad · 2024-06-27T23:15:27Z

src/EnergyPlus/HVACVariableRefrigerantFlow.cc

-        state.dataHVACVarRefFlow->MaxHeatingCapacity(VRFCond) = 0.0;
+        this->HeatingCapacity = 0.0;                                 // Include the piping loss
+        this->PipingCorrectionHeating = 1.0;                         // 1 means no piping loss
+        state.dataHVACVarRefFlow->MaxHeatingCapacity(VRFCond) = 0.0; // yujie: might be wrong here too, should be MaxCap = 1e+20


Please check where MaxHeatingCapacity and MaxCoolingCapacity get set to MaxCap(1.0e+20). That will answer your question. It should either be set to MaxCap or not set at all. Maybe this is the reason that MaxCap is an optional argument, because it's 0 at times. If this var were never set to 0, and only set to MaxCap when reset is needed, and only reduced when the system is overloaded, then MaxCap could always be passed to the coil model and NOT be an optional argument.

I just changed MaxCoolCap to non-optional.
I think this initialization here (state.dataHVACVarRefFlow->MaxHeatingCapacity(VRFCond) = 0.0) might be just an error. The default for capacity limit should be a really large number, so that no limit is placed.

Have you checked where MaxHeatingCapacity and MaxCoolingCapacity are set to MaxCap(1.0e+20)? Or if you believe this initialization is an error, I think you can correct the code line and delete the comment in this code.

It should be 1.0e+20. I'll change it and remove the comment.

@dareumnam. I corrected the code line and removed the comment on the 4th. Sorry I forgot to ping you. Please see if it looks okay. Thanks

rraustad · 2024-06-27T23:39:04Z

This change did improve the results based on the data shown in the description. This may be for a different issue, but how can coil capacity ever exceed HP capacity? Isn't it sum of coil capacity + piping losses = HP capacity? So coil capacity should always be less than or equal to HP capacity.

yujiex · 2024-06-27T23:49:00Z

This change did improve the results based on the data shown in the description. This may be for a different issue, but how can coil capacity ever exceed HP capacity? Isn't it sum of coil capacity + piping losses = HP capacity? So coil capacity should always be less than or equal to HP capacity.

Indeed, I'll look into this. There might still be something not quite right. Coil cooling rate should be no more than heat pump cooling rate

yujiex · 2024-06-28T21:21:59Z

This change did improve the results based on the data shown in the description. This may be for a different issue, but how can coil capacity ever exceed HP capacity? Isn't it sum of coil capacity + piping losses = HP capacity? So coil capacity should always be less than or equal to HP capacity.

I investigated this and found the issue might be caused by this 0.65 min fan speed ratio constraint.

    FanSpdRatioMin = min(max(OAMassFlow / Garate, 0.65), 1.0); // ensure that coil flow rate is higher than OA flow rate

After removing this constraint, the difference become smaller and the coil cooling rate is slightly smaller than HP cooling rate

@rraustad

move the MaxCap(1e+20) to DataGlobalConstants, to avoid circular inclusion issue

Myoldmopar · 2024-07-10T13:37:06Z

@yujiex in addition to the diffs in a handful of files, there are 3 unit tests that are now failing and must be dealt with before this goes much further. And at least one of the test failures is quite a significant change. We can discuss on technicalities if needed.

yujiex · 2024-07-10T17:33:48Z

@yujiex in addition to the diffs in a handful of files, there are 3 unit tests that are now failing and must be dealt with before this goes much further. And at least one of the test failures is quite a significant change. We can discuss on technicalities if needed.

Hi @Myoldmopar. The unit test failing is caused by this change (if you revert this part, the unit test can run through).

Previously the coil TotalCoolingEnergyRate and SensCoolingEnergyRate uses the InletAirMassFlowRate (shown in the following code chunk) instead of the AirMassFlow computed using the FanSpdRatio output from function ControlVRFIUCoil

        Real64 AirMassFlowRate = thisDXCoil.InletAirMassFlowRate;
        // Coil total/sensible/latent cooling rates
        CalcComponentSensibleLatentOutput(AirMassFlowRate,
                                          InletAirDryBulbTemp,
                                          InletAirHumRat,
                                          OutletAirTemp,
                                          OutletAirHumRat,
                                          thisDXCoil.SensCoolingEnergyRate,
                                          thisDXCoil.LatCoolingEnergyRate,
                                          thisDXCoil.TotalCoolingEnergyRate);

This might be wrong. The air flow should be changed to meet the cooling demand.

For example, in heating mode, the heating energy rate is actually computed using the AirMassFlow from function ControlVRFIUCoil

        ControlVRFIUCoil(state,
                         DXCoilNum,
                         QCoilReq,
                         thisDXCoil.InletAirTemp,
                         thisDXCoil.InletAirHumRat,
                         thisDXCoil.CondensingTemp,
                         AirMassFlowMin,
                         FanSpdRatio,
                         OutletAirHumRat,
                         OutletAirTemp,
                         OutletAirEnthalpy,
                         ActualSH,
                         ActualSC);
        AirMassFlow = FanSpdRatio * thisDXCoil.RatedAirMassFlowRate(Mode);
        ...
        thisDXCoil.TotalHeatingEnergyRate = AirMassFlow * (OutletAirEnthalpy - InletAirEnthalpy) * PartLoadRatio;

…TUCap

…rgyPlus into VRFFluidFixCoolingTUCap

github-actions · 2024-09-04T21:18:15Z

⚠️ Regressions detected on macos-14 for commit `ecfbb2d`

Regression Summary

ESO Big Diffs: 11
ESO Small Diffs: 2
EIO: 4
Table Big Diffs: 4
Table String Diffs: 4
MTR Big Diffs: 2

github-actions · 2024-09-05T17:42:41Z

⚠️ Regressions detected on macos-14 for commit `ae8bcbf`

Regression Summary

ESO Big Diffs: 11
ESO Small Diffs: 2
EIO: 4
Table Big Diffs: 4
Table String Diffs: 4
MTR Big Diffs: 2

…TUCap

github-actions · 2024-09-07T00:22:09Z

⚠️ Regressions detected on macos-14 for commit `7787de8`

Regression Summary

ESO Big Diffs: 11
ESO Small Diffs: 2
EIO: 4
Table Big Diffs: 4
Table String Diffs: 4
MTR Big Diffs: 2

dareumnam · 2024-09-10T21:30:44Z

@rraustad, Do you have any further comments on this? I think it's good to go after @Myoldmopar's final review, unless you have additional feedback.

Myoldmopar · 2024-09-11T15:12:55Z

This has now got conflicts from all the other merges. @yujiex could you resolve them? If so this might be ready.

yujiex · 2024-09-11T15:32:59Z

@Myoldmopar I will resolve the conflict in a bit (on bus now)

…TUCap

yujiex · 2024-09-11T16:39:05Z

@Myoldmopar I just resolved the conflict.

Myoldmopar · 2024-09-11T17:52:04Z

Thanks, let's see how CI responds in like 30 minutes.

github-actions · 2024-09-11T18:26:15Z

⚠️ Regressions detected on macos-14 for commit `e392e07`

Regression Summary

EIO: 6
ESO Big Diffs: 13
MTR Big Diffs: 4
Table Big Diffs: 6
Table String Diffs: 6
ESO Small Diffs: 2
ERR: 1

Myoldmopar · 2024-09-11T18:28:21Z

OK OK 35 minutes not 30.

So there are 2 more files now showing big diffs. I wonder what's caused that. Do you have any idea?

…TUCap

yujiex · 2024-09-11T18:38:18Z

@Myoldmopar It's strange, as the additional diffs happen in "CentralChillerHeaterSystem_Cooling_Heating" and "CentralChillerHeaterSystem_Simultaneous_Cooling_Heating", which doesn't have the VRF objects. Could it be because I'm not up-to-date with the develop?

I just pulled develop and again and run regression on those two locally, they don't show diffs.

Myoldmopar · 2024-09-11T18:43:22Z

Hmmm. Now that you mention it, maybe so. I was just bragging about how we wouldn't see diffs just from being behind develop. But maybe there's still a way. When the CI system is doing the build, it checks out the current develop as the baseline. Because CI is a bit bogged down right now, develop has definitely been moving fast. So yeah, it's possible. In most scenarios it won't be happening, just if develop moves really really fast.

yujiex · 2024-09-11T18:48:00Z

Hmmm. Now that you mention it, maybe so. I was just bragging about how we wouldn't see diffs just from being behind develop. But maybe there's still a way. When the CI system is doing the build, it checks out the current develop as the baseline. Because CI is a bit bogged down right now, develop has definitely been moving fast. So yeah, it's possible. In most scenarios it won't be happening, just if develop moves really really fast.

seems develop is getting many things merged in today

github-actions · 2024-09-11T21:06:21Z

⚠️ Regressions detected on macos-14 for commit `f4ba8d2`

Regression Summary

ESO Big Diffs: 11
ESO Small Diffs: 2
EIO: 4
ERR: 1
Table Big Diffs: 4
Table String Diffs: 4
MTR Big Diffs: 2

Myoldmopar · 2024-09-12T15:54:41Z

Back to the 11 ESO big diffs like above. Seems like that took care of it. Thanks @yujiex

yujiex · 2024-09-12T15:59:29Z

Thanks @Myoldmopar

Myoldmopar · 2024-09-12T16:39:19Z

This has a conflict in the unit test. I'm doing a final quick test on #10416 and then try to resolve the conflict. If it's minor I'll handle it, but if it's more complex, I'll toss it back over to you.

Myoldmopar · 2024-09-12T16:56:33Z

I resolved the test conflicts and everything is passing locally. I'm going to stop for lunch and let CI catch up and get fresh results to confirm I've got this ready. If CI is clean this can go in.

github-actions · 2024-09-12T17:40:27Z

⚠️ Regressions detected on macos-14 for commit `268a433`

Regression Summary

ESO Big Diffs: 11
ESO Small Diffs: 2
EIO: 4
ERR: 1
Table Big Diffs: 4
Table String Diffs: 4
MTR Big Diffs: 2

yujiex · 2024-09-12T17:57:03Z

Thanks for handling the conflict @Myoldmopar

Myoldmopar · 2024-09-12T20:23:02Z

Looks like I didn't break anything with my conflict resolution, merging this. Thanks @yujiex

yujiex and others added 5 commits August 30, 2023 15:51

initial commit of example file

cf6d16c

fix cooling with similar approach as heating

f7249d7

add testfile to CMakeList, add weather file, clang-format

1eccac5

test file for cooling not meeting demand case

0ad50a9

yujiex added the Defect Includes code to repair a defect in EnergyPlus label Jun 11, 2024

yujiex added this to the EnergyPlus 24.2 milestone Jun 11, 2024

yujiex self-assigned this Jun 11, 2024

yujiex changed the title ~~Vrf fluid fix cooling tu cap~~ Fixing cooling mode cooling coil cooling rate much higher than heat pump cooling rate and OU capacity Jun 11, 2024

Yujie Xu added 2 commits June 26, 2024 17:54

Merge remote-tracking branch 'origin/develop' into VRFFluidFixCooling…

06a0bbc

…TUCap

remove duplicated add testfile in CMakeLists

867e464

yujiex requested review from Myoldmopar, mjwitte and rraustad June 27, 2024 21:44

yujiex marked this pull request as ready for review June 27, 2024 21:44

Merge remote-tracking branch 'origin/develop' into VRFFluidFixCooling…

8f63776

…TUCap

rraustad reviewed Jun 27, 2024

View reviewed changes

Change MaxCoolCap to non-optional arg

cbaa5d5

Yujie Xu added 2 commits June 30, 2024 22:40

make MaxCap and MaxCoolCap a default argument, move constant declare

0b22d9f

move the MaxCap(1e+20) to DataGlobalConstants, to avoid circular inclusion issue

remove unwanted comment

0b62cc6

Myoldmopar assigned Myoldmopar and unassigned yujiex Jul 3, 2024

Merge remote-tracking branch 'origin/develop' into VRFFluidFixCooling…

b064b92

…TUCap

Merge branch 'VRFFluidFixCoolingTUCap' of https://github.com/NREL/Ene…

0b9e1ff

…rgyPlus into VRFFluidFixCoolingTUCap

fix unit test and clang-format

7b30ac4

Merge remote-tracking branch 'origin/develop' into VRFFluidFixCooling…

49506c8

…TUCap

Merge remote-tracking branch 'origin/develop' into VRFFluidFixCooling…

f59c122

…TUCap

Merge remote-tracking branch 'origin/develop' into VRFFluidFixCooling…

50c2647

…TUCap

Merge develop; resolve conflicts

185e5c1

Myoldmopar merged commit 1b9d24d into develop Sep 12, 2024
7 checks passed

Myoldmopar deleted the VRFFluidFixCoolingTUCap branch September 12, 2024 20:23

yujiex mentioned this pull request Oct 2, 2024

Large diffs between 24.1.0 and 24.2.0 #10778

Closed

3 tasks

Fixing cooling mode cooling coil cooling rate much higher than heat pump cooling rate and OU capacity #10557

Fixing cooling mode cooling coil cooling rate much higher than heat pump cooling rate and OU capacity #10557

Conversation

yujiex commented Jun 11, 2024 • edited Loading

Pull request overview

Regression diff

ESO difference

EIO difference

Meter diff

Table string diffs

Pull Request Author

Reviewer

Choose a reason for hiding this comment

yujiex Jun 27, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rraustad commented Jun 27, 2024

yujiex commented Jun 27, 2024

yujiex commented Jun 28, 2024

Myoldmopar commented Jul 10, 2024

yujiex commented Jul 10, 2024

github-actions bot commented Sep 4, 2024

⚠️ Regressions detected on macos-14 for commit ecfbb2d

github-actions bot commented Sep 5, 2024

⚠️ Regressions detected on macos-14 for commit ae8bcbf

github-actions bot commented Sep 7, 2024

⚠️ Regressions detected on macos-14 for commit 7787de8

dareumnam commented Sep 10, 2024

Myoldmopar commented Sep 11, 2024

yujiex commented Sep 11, 2024

yujiex commented Sep 11, 2024

Myoldmopar commented Sep 11, 2024

github-actions bot commented Sep 11, 2024

⚠️ Regressions detected on macos-14 for commit e392e07

Myoldmopar commented Sep 11, 2024

yujiex commented Sep 11, 2024 • edited Loading

Myoldmopar commented Sep 11, 2024

yujiex commented Sep 11, 2024

github-actions bot commented Sep 11, 2024

⚠️ Regressions detected on macos-14 for commit f4ba8d2

Myoldmopar commented Sep 12, 2024

yujiex commented Sep 12, 2024

Myoldmopar commented Sep 12, 2024

Myoldmopar commented Sep 12, 2024

github-actions bot commented Sep 12, 2024

⚠️ Regressions detected on macos-14 for commit 268a433

yujiex commented Sep 12, 2024

Myoldmopar commented Sep 12, 2024

yujiex commented Jun 11, 2024 •

edited

Loading

yujiex Jun 27, 2024 •

edited

Loading

⚠️ Regressions detected on macos-14 for commit `ecfbb2d`

⚠️ Regressions detected on macos-14 for commit `ae8bcbf`

⚠️ Regressions detected on macos-14 for commit `7787de8`

⚠️ Regressions detected on macos-14 for commit `e392e07`

yujiex commented Sep 11, 2024 •

edited

Loading

⚠️ Regressions detected on macos-14 for commit `f4ba8d2`

⚠️ Regressions detected on macos-14 for commit `268a433`