New attempt at fixing #7464: outlet water temperature control issue for variable speed cooling tower #10248

lymereJ · 2023-09-27T23:06:43Z

Pull request overview

Fixes CoolingTower:VariableSpeed control problems #7464

#7702 provided a fix for the original defect file from #7464 by increasing the upper bound used to solve for the range to determine the outlet water temperature of the tower. Tests have shown that increasing further the upper bound would solve the issue for the new defect file, however, it might still not be a solution that is general enough . The proposed changes include a variable upper bound in order to avoid this type of error before calling regula falsi.

A warning (and recurring as well) was added to identify instances where no outlet temperature control is attempted due potential bad intervals passed to regula falsi.

Pull Request Author

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

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

Reviewer

This will not be exhaustively relevant to every PR.

…ases where the tower does no outlet temperature control but run the fan at full speed.

lymereJ · 2023-09-27T23:27:44Z

Results for the defect file:

With the proposed changes:

The figure shows that the cooling tower is now meeting the setpoint.

lymereJ · 2023-09-27T23:41:14Z

The proposed changes have a large impact on the heat rejection energy use because the tower is now running appropriately.

Current:

Proposed changes:

In the current case, the tower is running more because the following code snippet sets the airflow ratio, attempts to calculate the outlet and if the outlet is greater than the setpoint it assumes that the unit is running at full capacity, however, if the combination of lower and upper bound passed to regula falsi lead to a returned value of -2 the outlet water temperature is not actually controlled (i.e. not calculated) so the tower energy use is overestimated.

EnergyPlus/src/EnergyPlus/CondenserLoopTowers.cc

Lines 5223 to 5244 in 8f094c8

    
               this->airFlowRateRatio = 1.0; 
        
               OutletWaterTempOFF = state.dataLoopNodes->Node(this->WaterInletNodeNum).Temp; 
        
               this->OutletWaterTemp = OutletWaterTempOFF; 
        
               FreeConvectionCapFrac = this->FreeConvectionCapacityFraction; 
        
               OutletWaterTempON = this->calculateVariableTowerOutletTemp(state, WaterFlowRateRatioCapped, this->airFlowRateRatio, TwbCapped); 
        
               if (OutletWaterTempON > TempSetPoint) { 
        
                   this->FanCyclingRatio = 1.0; 
        
                   this->airFlowRateRatio = 1.0; 
        
                   this->FanPower = this->HighSpeedFanPower * this->NumCellOn / this->NumCell; 
        
                   this->OutletWaterTemp = OutletWaterTempON; 
        
                   // if possible increase the number of cells and do the calculations again with the new water mass flow rate per cell 
        
                   if (this->NumCellOn < this->NumCell && (this->WaterMassFlowRate / (this->NumCellOn + 1)) > WaterMassFlowRatePerCellMin) { 
        
                       ++this->NumCellOn; 
        
                       WaterMassFlowRatePerCell = this->WaterMassFlowRate / this->NumCellOn; 
        
                       IncrNumCellFlag = true; 
        
                   } 
        
               } 
        
           } 
        
           // find the correct air ratio only if full flow is  too much 
        
           if (OutletWaterTempON < TempSetPoint) {

src/EnergyPlus/CondenserLoopTowers.cc

lymereJ · 2023-09-27T23:49:41Z

The defect file uses the YorkCalc model results using the CoolToolsCrossFlow model are the same before and after the fix but are more in part with the new results for the YorkCalc model:

src/EnergyPlus/CondenserLoopTowers.cc

rraustad · 2023-09-28T04:44:06Z

So if the SP is 30C, why is the tower outlet temp now showing 32.2C?

Schedule:Day:Interval,
    New HP-Loop-Temp-Schedule 2 All Days,  !- Name
    Temperature 44,          !- Schedule Type Limits Name
    No,                      !- Interpolate to Timestep
    24:00,                   !- Time 1 {hh:mm}
    30;                      !- Value Until Time 1

I don't see anything in the tower inputs that would explain that. But a new loop on range temp may be influencing the results? But then I do see a min range (1.11) and approach (1.11) temperature that would explain that (30 + 2.22 = 32.22)? Too hard to know how these inputs affect results without spending a lot of time with this model.
And on that note, how do these inputs relate to what is happening now with a dynamic range temperature?

CoolingTowerPerformance:YorkCalc,
    YorkCalc Default Tower Model,  !- Name
    -34.4,                   !- Minimum Inlet Air Wet-Bulb Temperature {C}
    26.6667,                 !- Maximum Inlet Air Wet-Bulb Temperature {C}
    1.1111,                  !- Minimum Range Temperature {deltaC}
    22.2222,                 !- Maximum Range Temperature {deltaC}
    1.1111,                  !- Minimum Approach Temperature {deltaC}
    40.0,                    !- Maximum Approach Temperature {deltaC}

rraustad · 2023-09-28T04:57:04Z

What energy results from a single-speed and two-speed tower? Those results could be compared to these.

lymereJ · 2023-09-28T05:01:42Z

So if the SP is 30C, why is the tower outlet temp now showing 32.2C?
Schedule:Day:Interval,
    New HP-Loop-Temp-Schedule 2 All Days,  !- Name
    Temperature 44,          !- Schedule Type Limits Name
    No,                      !- Interpolate to Timestep
    24:00,                   !- Time 1 {hh:mm}
    30;                      !- Value Until Time 1
I don't see anything in the tower inputs that would explain that. But a new loop on range temp may be influencing the results? But then I do see a min range (1.11) and approach (1.11) temperature that would explain that (30 + 2.22 = 30.22)? Too hard to know how these inputs affect results without spending a lot of time with this model. And on that note, how do these inputs relate to what is happening now with a dynamic range temperature?
CoolingTowerPerformance:YorkCalc,
    YorkCalc Default Tower Model,  !- Name
    -34.4,                   !- Minimum Inlet Air Wet-Bulb Temperature {C}
    26.6667,                 !- Maximum Inlet Air Wet-Bulb Temperature {C}
    1.1111,                  !- Minimum Range Temperature {deltaC}
    22.2222,                 !- Maximum Range Temperature {deltaC}
    1.1111,                  !- Minimum Approach Temperature {deltaC}
    40.0,                    !- Maximum Approach Temperature {deltaC}

That schedule is not in the defect file. The defect file uses 32.22.

  Schedule:Day:Interval,
    Min Max Temp w Deadband Max-90.0C Day,  !- Name
    Min Max Temp w Deadband Max-90.0C Limits,  !- Schedule Type Limits Name
    No,                      !- Interpolate to Timestep
    24:00,                   !- Time 1
    32.22;                   !- Value Until Time 1

lymereJ · 2023-09-28T05:03:18Z

What energy results from a single-speed and two-speed tower? Those results could be compared to these.

I have results for the variable speed Merkel model but I'll generate these as well.

rraustad · 2023-09-28T05:09:10Z

OK, I am looking at in-moreoutput-2xTower.idf from #7464. Now I see there is defect_file.zip which I assume you are using. Good, that answers the SP issue.

lymereJ · 2023-09-28T16:53:17Z

What energy results from a single-speed and two-speed tower? Those results could be compared to these.

I have results for the variable speed Merkel model but I'll generate these as well.

Here are the results for the other models:

Single speed:

Two speed (Merkel):

Variable speed (Merkel):

So the current case (YorkCalc model) would perform worse than the single speed tower. The results for the variable speed merkel model are in par with the YorkCalc and CoolToolsCrossFlow.

rraustad · 2023-09-28T17:01:25Z

So with cooling energy roughly the same at 280 GJ, the difference in Heat Rejection is due to a reduction in fan/pump power. This trend (heat rejection energy falling as tower speeds increase) should provide a good basis to validate all tower models. Now the results of this branch make more sense.

tst/EnergyPlus/unit/CondenserLoopTowers.unit.cc

src/EnergyPlus/CondenserLoopTowers.cc

nrel-bot-3 · 2023-10-31T00:15:22Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

nrel-bot-2 · 2023-11-28T00:06:20Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

nrel-bot-2 · 2023-12-26T00:08:24Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

nrel-bot · 2024-01-23T00:10:06Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

nrel-bot · 2024-02-20T00:16:30Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

nrel-bot · 2024-04-12T00:11:18Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

nrel-bot-2b · 2024-04-12T00:14:36Z

@lymereJ @Myoldmopar it has been 28 days since this pull request was last updated.

lymereJ

Walkthrough of the changes.

lymereJ · 2024-05-06T16:41:56Z

src/EnergyPlus/CondenserLoopTowers.cc

+        SetupOutputVariable(state,
+                            "Cooling Tower Approach",
+                            Constant::Units::C,
+                            this->coolingTowerApproach,
+                            OutputProcessor::TimeStepType::System,
+                            OutputProcessor::StoreType::Average,
+                            this->Name);
+        SetupOutputVariable(state,
+                            "Cooling Tower Range",
+                            Constant::Units::C,
+                            this->coolingTowerRange,
+                            OutputProcessor::TimeStepType::System,
+                            OutputProcessor::StoreType::Average,
+                            this->Name);


Add range and approach as outputs for all cooling tower objects.

Looks good.

lymereJ · 2024-05-06T16:56:38Z

src/EnergyPlus/CondenserLoopTowers.cc


-        Real64 OutletWaterTempLocal = this->WaterTemp - Tr;
+        // calculate outlet temperature
+        Real64 outletWaterTempLocal = this->WaterTemp - min(Tr, this->MaxRangeTemp);


Make sure that the range doesn't exceeds the model's maximum range.

lymereJ · 2024-05-06T17:30:44Z

src/EnergyPlus/CondenserLoopTowers.cc

+        } else if (SolFla == -2) {
+            // bad starting value means that solution corresponds to a range that is beyond
+            // the bounds of the model; The maximum range is used here
+            outletWaterTempLocal = this->WaterTemp - this->MaxRangeTemp;
+            ++this->VSErrorCountTRCalc;
+            if (this->VSErrorCountTRCalc < 2) {
+                ShowWarningError(
+                    state,
+                    format("The range for the cooling tower {} likely exceeds the bounds of the model. The maximum range of the model is used {}.",
+                           this->Name,
+                           this->MaxRangeTemp));
+                ShowContinueError(state,
+                                  format(" ... Occurrence info = {}, {} {}",
+                                         state.dataEnvrn->EnvironmentName,
+                                         state.dataEnvrn->CurMnDy,
+                                         General::CreateSysTimeIntervalString(state)));
+            } else {
+                ShowRecurringWarningErrorAtEnd(
+                    state,
+                    format("The range for the cooling tower {} likely exceeds the bounds of the model. The maximum range of the model is used {}",
+                           this->Name,
+                           this->MaxRangeTemp),
+                    this->ErrIndexTRCalc);


Previously, if a "bad starting value" error was returned by Regula Falsi (SolFla == -2, i.e., both the output for the function for the lower and upper bounds are of the same sign), the outlet temperature was only recalculated if the inlet temperature minus the range were greater than the setpoint. As described in the issue, this occurs for the YorkCalc model. Looking at that model closer shows that this typically occurs when the range returned by the model is significantly greater than the maximum range. In the first figure below, the solution is a range of about 63 deg. C which is way beyond the limit of the model (22.2222).

Inlet water temperature: 38.75 deg. C; Wet-bulb temperature: 15.4 deg. C; Flow ratio: 2

Varying wet-bulb temperature

Varying air flow ratio

Varying water flow ratio

Increasing further the upper bound used with Regula Falsi works but doesn't make sense since the solutions are outside the limits of the model. The proposed changes apply the maximum range every time SolFla == 2. As shown in the simulation results above, this approach leads to the correct control of the outlet of the tower and reasonable fan power unlike the status quo.

Currently, when the outlet is not recalculated, no warning is issued, The proposed changes include a (recurring) warning every time this happens.

lymereJ · 2024-05-06T17:54:32Z

tst/EnergyPlus/unit/CondenserLoopTowers.unit.cc

    VSTower.calculateVariableSpeedTower(*state);
    EXPECT_DOUBLE_EQ(30.0, VSTower.OutletWaterTemp);
-    EXPECT_NEAR(0.5213, VSTower.FanCyclingRatio, 0.0001);
+    EXPECT_NEAR(0.5424, VSTower.FanCyclingRatio, 0.0001);


This change to an existing test was necessary because prior to the proposed changes the range used in CoolingTower::calculateVariableTowerOutletTemp was not capped to the maximum range for a selected model which lead to this test using a range greater than the maximum, see screenshot below.

Interesting. And makes sense.

lymereJ · 2024-05-06T17:54:37Z

tst/EnergyPlus/unit/CondenserLoopTowers.unit.cc

    EXPECT_EQ(vSpdMerkel_tower.HeatRejectCapNomCapSizingRatio, 1.25);
 }

+TEST_F(EnergyPlusFixture, CondenserLoopTowers_CalculateVariableTowerOutletTemp)


New unit test for defect.

lymereJ · 2024-05-06T18:59:15Z

The RDD/AUD diffs are due to the new outputs. Most ERR files are due to the new warning. The two EIO diffs are very small convergence value changes. The ESO files diffs appears very small.

Myoldmopar · 2024-05-08T16:20:03Z

CI results do look OK based on the intended changes. I'm building and testing locally and looking over the changes now.

Myoldmopar

No major red flags here, I think this is good to go in.

Myoldmopar · 2024-05-08T16:32:17Z

src/EnergyPlus/CondenserLoopTowers.cc

+        SetupOutputVariable(state,
+                            "Cooling Tower Approach",
+                            Constant::Units::C,
+                            this->coolingTowerApproach,
+                            OutputProcessor::TimeStepType::System,
+                            OutputProcessor::StoreType::Average,
+                            this->Name);
+        SetupOutputVariable(state,
+                            "Cooling Tower Range",
+                            Constant::Units::C,
+                            this->coolingTowerRange,
+                            OutputProcessor::TimeStepType::System,
+                            OutputProcessor::StoreType::Average,
+                            this->Name);


Looks good.

Myoldmopar · 2024-05-08T16:37:19Z

tst/EnergyPlus/unit/CondenserLoopTowers.unit.cc

    VSTower.calculateVariableSpeedTower(*state);
    EXPECT_DOUBLE_EQ(30.0, VSTower.OutletWaterTemp);
-    EXPECT_NEAR(0.5213, VSTower.FanCyclingRatio, 0.0001);
+    EXPECT_NEAR(0.5424, VSTower.FanCyclingRatio, 0.0001);


Interesting. And makes sense.

Myoldmopar · 2024-05-08T16:40:37Z

@rraustad these changes seem fine, let me know soon if you want anything else done here, otherwise I'll merge.

rraustad · 2024-05-08T16:53:41Z

I'm good.

Myoldmopar · 2024-05-08T17:29:43Z

I'm good.

Thanks @rraustad , I'm going to merge this then. Thanks @lymereJ

Add variable upper bound for reg falsi and warning message to avoid c…

a442532

…ases where the tower does no outlet temperature control but run the fan at full speed.

lymereJ added the Defect Includes code to repair a defect in EnergyPlus label Sep 27, 2023

rraustad reviewed Sep 27, 2023

View reviewed changes

src/EnergyPlus/CondenserLoopTowers.cc Outdated Show resolved Hide resolved

lymereJ marked this pull request as draft September 27, 2023 23:43

rraustad reviewed Sep 27, 2023

View reviewed changes

src/EnergyPlus/CondenserLoopTowers.cc Outdated Show resolved Hide resolved

Address comments.

a0927b4

rraustad reviewed Sep 27, 2023

View reviewed changes

src/EnergyPlus/CondenserLoopTowers.cc Outdated Show resolved Hide resolved

lymereJ requested review from Nigusse and mjwitte September 28, 2023 00:01

lymereJ added 2 commits September 28, 2023 11:18

Add unit test.

08e42ed

Merge branch 'develop' into clgtwr_ext_rng

c525790

lymereJ commented Sep 28, 2023

View reviewed changes

tst/EnergyPlus/unit/CondenserLoopTowers.unit.cc Show resolved Hide resolved

lymereJ commented Sep 28, 2023

View reviewed changes

src/EnergyPlus/CondenserLoopTowers.cc Outdated Show resolved Hide resolved

lymereJ marked this pull request as ready for review September 28, 2023 18:29

Merge branch 'develop' into clgtwr_ext_rng

0513e38

lymereJ added the DoNotMerge Code that requires additional attention and investigation label Sep 29, 2023

lymereJ marked this pull request as draft September 29, 2023 00:26

lymereJ removed request for Nigusse and mjwitte September 29, 2023 17:20

Add cap.

97a4b63

Merge branch 'develop' into clgtwr_ext_rng

d18d43d

lymereJ removed the DoNotMerge Code that requires additional attention and investigation label Oct 2, 2023

Myoldmopar added this to the EnergyPlus 24.2 milestone Mar 14, 2024

lymereJ added 8 commits April 30, 2024 11:52

Merge branch 'develop' into clgtwr_ext_rng

0ec2d12

Revise fix and add outputs for range and approach.

48c95dc

Merge branch 'develop' into clgtwr_ext_rng

c0d414e

Fix reference.

321a20f

Address unit test failure.

98d8eb0

Edit warning.

7a2ea43

Revert unecessary change.

55e6b7c

Cleanup...

f80e5bf

lymereJ commented May 6, 2024

View reviewed changes

lymereJ marked this pull request as ready for review May 6, 2024 18:59

Myoldmopar approved these changes May 8, 2024

View reviewed changes

Myoldmopar merged commit 4d39dff into develop May 8, 2024

Myoldmopar deleted the clgtwr_ext_rng branch May 8, 2024 17:29

New attempt at fixing #7464: outlet water temperature control issue for variable speed cooling tower #10248

New attempt at fixing #7464: outlet water temperature control issue for variable speed cooling tower #10248

Uh oh!

Conversation

lymereJ commented Sep 27, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull request overview

Pull Request Author

Reviewer

Uh oh!

lymereJ commented Sep 27, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lymereJ commented Sep 27, 2023

Uh oh!

Uh oh!

lymereJ commented Sep 27, 2023

Uh oh!

Uh oh!

Uh oh!

rraustad commented Sep 28, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rraustad commented Sep 28, 2023

Uh oh!

lymereJ commented Sep 28, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lymereJ commented Sep 28, 2023

Uh oh!

rraustad commented Sep 28, 2023

Uh oh!

lymereJ commented Sep 28, 2023

Uh oh!

rraustad commented Sep 28, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

nrel-bot-3 commented Oct 31, 2023

Uh oh!

nrel-bot-2 commented Nov 28, 2023

Uh oh!

nrel-bot-2 commented Dec 26, 2023

Uh oh!

nrel-bot commented Jan 23, 2024

Uh oh!

nrel-bot commented Feb 20, 2024

Uh oh!

nrel-bot commented Apr 12, 2024

Uh oh!

nrel-bot-2b commented Apr 12, 2024

Uh oh!

lymereJ left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lymereJ commented May 6, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Myoldmopar commented May 8, 2024

Uh oh!

Myoldmopar left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lymereJ commented Sep 27, 2023 •

edited

Loading

lymereJ commented Sep 27, 2023 •

edited

Loading

rraustad commented Sep 28, 2023 •

edited

Loading

lymereJ commented Sep 28, 2023 •

edited

Loading

rraustad commented Sep 28, 2023 •

edited

Loading

lymereJ commented May 6, 2024 •

edited

Loading