EMS incorrectly reads output variable

NREL / EnergyPlus

EnergyPlus™ is a whole building energy simulation program that engineers, architects, and researchers use to model both energy consumption and water use in buildings.

https://energyplus.net

Other

1.09k stars 382 forks source link

EMS incorrectly reads output variable #7563

Open mldichter opened 4 years ago

mldichter commented 4 years ago

Issue overview

EMS incorrectly reads the Facility Net Purchased Electric Energy output variable. The value being read from the output variable and rewritten to a meter is not the value of the output variable in the CSV file with errors on the order of 10^7.

Details

Some additional details for this issue (if relevant):

Windows 10
EnergyPlus 9.2 (originally found in 9.0)
Unmethours link https://unmethours.com/question/40236/ems-incorrectly-reads-output-variable/

Checklist

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

[x] Defect file added (https://drive.google.com/file/d/1nx0lCBeOn6qOWiUVw2NzMs3LB9dYoAHI/view?usp=sharing)
[ ] Ticket added to Pivotal for defect (development team task)
[ ] Pull request created (the pull request will have additional tasks related to reviewing changes that fix this defect)

LorenzoBottaccioli commented 3 years ago

Having the same issue with DistrictHeating:Facilty the EMS sensor reports different values from the one I have in the output variables files.

mjwitte commented 3 years ago

This is a matter of selecting the correct calling point. These are the pertinent parts of the idf file linked to this issue:

EnergyManagementSystem:Sensor,
    Pre_EMS_Net_Electricity, !- Name
    Whole Building,          !- Output:Variable or Output:Meter Index Key Name
    Facility Net Purchased Electric Energy;  !- Output:Variable or Output:Meter Name

EnergyManagementSystem:ProgramCallingManager,
    Read_EMS_Write_Caller,   !- Name
    AfterPredictorAfterHVACManagers,  !- EnergyPlus Model Calling Point
    Read_Write_Net_Electricity;  !- Program Name 1

EnergyManagementSystem:Program,
    Read_Write_Net_Electricity,  !- Name
    SET EMS_Net_Electricity = Pre_EMS_Net_Electricity;  !- Program Line 1

EnergyManagementSystem:OutputVariable,
    EMS_Net_Electricity,     !- Name
    EMS_Net_Electricity,     !- EMS Variable Name
    Summed,                  !- Type of Data in Variable
    SystemTimestep,          !- Update Frequency
    Read_Write_Net_Electricity,  !- EMS Program or Subroutine Name
    J;                       !- Units

Meter:Custom,
    EMS Net Facility Electricity Meter,  !- Name
    Electricity,             !- Fuel Type
    ,                        !- Key Name 1
    EMS_Net_Electricity;     !- Output Variable or Meter Name 1

Output:Variable,
    Whole Building,          !- Key Value
    Facility Net Purchased Electric Energy,  !- Variable Name
    Hourly;                  !- Reporting Frequency

Output:Meter,
    EMS Net Facility Electricity Meter,  !- Key Name
    Hourly;                  !- Reporting Frequency

The idf file linked above is using calling point AfterPredictorAfterHVACManagers. The "Managers" in this calling point name are the setpoint managers and availability managers, so this is before the HVAC systems and electric power manager have been simulated. Using the "detailed" reporting frequency shows how the values lag by one system timestep.

The only calling point that will work is EndOfSystemTimestepBeforeHVACReporting. This sets the EMS variable after the electric power manager has been simulated, but before the output variables and meters are accumulated during HVACReporting. One might also think that an EndOfZoneTimestep* calling point would work, but that is not the case since the new EMS output variable has been specified as a system timestep variable.

LorenzoBottaccioli commented 3 years ago

@mjwitte I have tried with all the combination none of them is working. MY EMS configuration below with Ideal Loads.

Whene I compare DisrictHeating:Facility with my heat_power I get completely different values. I also attach my idf. test_idf_epw.zip

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:SENSOR ===========

EnergyManagementSystem:Sensor,
    T_in,                    !- Name
    Block1:ZoneWest,         !- Output:Variable or Output:Meter Index Key Name
    Zone Mean Air Temperature;  !- Output:Variable or Output:Meter Name

EnergyManagementSystem:Sensor,
    mdot,                    !- Name
    BLOCK1:ZONEWEST IDEAL LOADS AIR,  !- Output:Variable or Output:Meter Index Key Name
    Zone Ideal Loads Supply Air Mass Flow Rate;  !- Output:Variable or Output:Meter Name

EnergyManagementSystem:Sensor,
    heat_power,              !- Name
    DistrictHeating:Facility,!- Output:Variable or Output:Meter Index Key Name
    DistrictHeating:Facility;!- Output:Variable or Output:Meter Name

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:PROGRAMCALLINGMANAGER ===========

EnergyManagementSystem:ProgramCallingManager,
    sensors,                 !- Name
    EndOfSystemTimestepBeforeHVACReporting,  !- EnergyPlus Model Calling Point
    getSensors;                  !- Program Name 1

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:PROGRAM ===========

EnergyManagementSystem:Program,
    getSensors,                  !- Name
    ;                        !- Program Line 1

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:OUTPUTVARIABLE ===========

EnergyManagementSystem:OutputVariable,
    T_in,                    !- Name
    T_in,                    !- EMS Variable Name
    Averaged,                !- Type of Data in Variable
    ZoneTimestep,            !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    C;                       !- Units

EnergyManagementSystem:OutputVariable,
    heat_power,              !- Name
    heat_power,              !- EMS Variable Name
    Averaged,                !- Type of Data in Variable
    SystemTimestep,          !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    J;                       !- Units

EnergyManagementSystem:OutputVariable,
    mdot,                    !- Name
    mdot,                    !- EMS Variable Name
    Summed,                  !- Type of Data in Variable
    ZoneTimestep,            !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    kg/s;                    !- Units

bonnema commented 3 years ago

Does this work? You could look at the EDD with this, maybe get more insight.

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:SENSOR ===========

EnergyManagementSystem:Sensor,
    T_in_sen,                !- Name
    Block1:ZoneWest,         !- Output:Variable or Output:Meter Index Key Name
    Zone Mean Air Temperature;  !- Output:Variable or Output:Meter Name

EnergyManagementSystem:Sensor,
    mdot_sen,                !- Name
    BLOCK1:ZONEWEST IDEAL LOADS AIR,  !- Output:Variable or Output:Meter Index Key Name
    Zone Ideal Loads Supply Air Mass Flow Rate;  !- Output:Variable or Output:Meter Name

EnergyManagementSystem:Sensor,
    heat_power_sen,          !- Name
    DistrictHeating:Facility,!- Output:Variable or Output:Meter Index Key Name
    DistrictHeating:Facility;!- Output:Variable or Output:Meter Name

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:PROGRAMCALLINGMANAGER ===========

EnergyManagementSystem:ProgramCallingManager,
    sensors,                 !- Name
    EndOfSystemTimestepBeforeHVACReporting,  !- EnergyPlus Model Calling Point
    getSensors;              !- Program Name 1

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:PROGRAM ===========

EnergyManagementSystem:Program,
    getSensors,              !- Name
    SET T_in_var = T_in_sen, !- Program Line 1
    SET mdot_var = mdot_sen, !- Program Line 2
    SET T_in_var = T_in_sen; !- Program Line 3

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:OUTPUTVARIABLE ===========

EnergyManagementSystem:OutputVariable,
    T_in,                    !- Name
    T_in_var,                !- EMS Variable Name
    Averaged,                !- Type of Data in Variable
    ZoneTimestep,            !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    C;                       !- Units

EnergyManagementSystem:OutputVariable,
    heat_power,              !- Name
    heat_power_var,          !- EMS Variable Name
    Averaged,                !- Type of Data in Variable
    SystemTimestep,          !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    J;                       !- Units

EnergyManagementSystem:OutputVariable,
    mdot,                    !- Name
    mdot_var,                !- EMS Variable Name
    Summed,                  !- Type of Data in Variable
    ZoneTimestep,            !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    kg/s;                    !- Units

LorenzoBottaccioli commented 3 years ago

@bonnema my problem is with DistrictHeating:Facility that if compared with the one in the output:variable file they mismatch. I can read the others without writing down the program.

bonnema commented 3 years ago

@LorenzoBottaccioli, in your heat_power EnergyManagementSystem:OutputVariable you have set the Type of Data in Variable field to Averaged. In order to compare to the DistrictHeating:Facility meter, you should set this field to Summed.

@mjwitte, I think your calling point comment above still applies and you can reclose this issue.

LorenzoBottaccioli commented 3 years ago

@bonnema I tried as you said. Still not working. There is one time step difference now. The one of EMS seems to be reporting the previous one.

Date/Time         | EMS:heat_power [J](TimeStep)  | DistrictHeating:Facility [J](TimeStep)
01/01    00:15:00 | 0                             | 2890621.648
01/01    00:30:00 | 2890621.648                   | 2940704.733
01/01    00:45:00 | 2940704.733                   | 2999981.2
01/01    01:00:00 | 2999981.2                     | 3070781.035

Here part of the IDF


!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:SENSOR ===========

EnergyManagementSystem:Sensor,
    T_in,                    !- Name
    Block1:ZoneWest,         !- Output:Variable or Output:Meter Index Key Name
    Zone Mean Air Temperature;  !- Output:Variable or Output:Meter Name

EnergyManagementSystem:Sensor,
    heat_power,              !- Name
    DistrictHeating:Facility,!- Output:Variable or Output:Meter Index Key Name
    DistrictHeating:Facility;!- Output:Variable or Output:Meter Name

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:PROGRAMCALLINGMANAGER ===========

EnergyManagementSystem:ProgramCallingManager,
    sensorsTemp,             !- Name
    EndOfSystemTimestepBeforeHVACReporting,  !- EnergyPlus Model Calling Point
    getSensors;              !- Program Name 1

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:PROGRAM ===========

EnergyManagementSystem:Program,
    getSensors,              !- Name
    ;                        !- Program Line 1

!-   ===========  ALL OBJECTS IN CLASS: ENERGYMANAGEMENTSYSTEM:OUTPUTVARIABLE ===========

EnergyManagementSystem:OutputVariable,
    T_in,                    !- Name
    T_in,                    !- EMS Variable Name
    Averaged,                !- Type of Data in Variable
    ZoneTimestep,            !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    C;                       !- Units

EnergyManagementSystem:OutputVariable,
    heat_power,              !- Name
    heat_power,              !- EMS Variable Name
    Summed,                  !- Type of Data in Variable
    ZoneTimestep,            !- Update Frequency
    ,                        !- EMS Program or Subroutine Name
    J;                       !- Units

bonnema commented 3 years ago

@LorenzoBottaccioli, a one timestep lag is relatively common in EMS programs. You may be able to try different calling points or the new(ish) python plugin system and see if they provide the solution you're looking for. I am not entirely sure of the intent of your program and as such it is hard to comment in more detail.

LorenzoBottaccioli commented 3 years ago

@bonnema I'm trying to understand the EMS usage to create an std link with the gym environment in order to develop Reinforcement learning algorithms for Model Predictive control. Tnx for your suggestion.

bonnema commented 3 years ago

Thanks @LorenzoBottaccioli. FWIW, adding the Output:EnergyManagementSystem object to your IDF with the last field set to Verbose will give line-by-line execution of the EMS program(s) in the .edd file that will then be generated. That can be helpful for debugging, but can lead to large files if turned on for annual simulations.

mjwitte commented 3 years ago

I'm keeping this open. I've been doing some experimenting with this file. Output variable sensors align when the appropriate calling point is used, but there is a real lag when using meters as a sensor.

Myoldmopar commented 3 years ago

@mjwitte any updates on this? Just tidying up some quiet issues. Thanks!

mjwitte commented 3 years ago

No updates from me. It appears to be a real issue for meters. Fair game for anyone who wants to try fixing it.