Computing Hvdc line loss in AC emulation #1062

SylvestreSakti · 2024-07-10T09:08:53Z

Please check if the PR fulfills these requirements

The commit message follows our guidelines
Tests for the changes have been added (for bug fixes / features)
Docs have been added / updated (for bug fixes / features)

Does this PR already have an issue describing the problem?
No

What kind of change does this PR introduce?
The computation of power loss due to resistance in HVDC line in AC emulation has been added.

What is the current behavior?
The resistance R of an HVDC line is not taken into account, the result of this approximation is that active power transit in the HVDC line can be overestimated.

What is the new behavior (if this is a feature change)?
The power loss formula has been introduced in the computation of power in AC emulation and is given in the figure below.

The Joule loss is minor compared to the power transit so its derivative has been neglected in the computation of the Jacobian matrix.

Does this PR introduce a breaking change or deprecate an API?

Yes
No

Other information:

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

… Emulation Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill · 2024-07-10T11:48:34Z

src/main/java/com/powsybl/openloadflow/network/impl/LfHvdcImpl.java

@@ -52,6 +54,7 @@ public LfHvdcImpl(String id, LfBus bus1, LfBus bus2, LfNetwork network, HvdcLine
 this.id = Objects.requireNonNull(id);
 this.bus1 = bus1;
 this.bus2 = bus2;
+ this.r = hvdcLine.getR() / PerUnit.zb(hvdcLine.getNominalV()); // Storing R directly in per unit


Can you remain consistent with other treatment. P0 is not "stored" in per-unit.

Yes, we had this question of consistency with two possible options :

Being consistent with the other treatments so introducing an additional property nominalV in LfHvdc

Changing the other treatment to "store" directly p0 and droop in per unit (but it may need some other modifications and maybe another PR)

I don't know which one is the best, but the easiest seems to be the first option

annetill · 2024-07-10T11:55:04Z

...main/java/com/powsybl/openloadflow/ac/equations/AbstractHvdcAcEmulationFlowEquationTerm.java

@@ -55,6 +58,20 @@ protected double rawP(double p0, double k, double ph1, double ph2) {
 return p0 + k * (ph1 - ph2);
 }

+ protected double rawPWithLoss(double rawP, double loss1, double loss2, double v, double r) {


We already have in powsybl-core the method in HvdcUtils.getHvdcLineLosses(...). Can we imagine reusing it here because it is what we use when the hvdc is operated in active power set point mode.

Thank you for the idea to reuse it (The formula is simpler with the approximation that the rectifier's output is at nominal voltage. When testing, there is no difference in the obtained results).

But for the moment I can't use this method because it is private

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

…-core Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill · 2024-07-10T13:37:42Z

...main/java/com/powsybl/openloadflow/ac/equations/AbstractHvdcAcEmulationFlowEquationTerm.java

@@ -78,6 +81,17 @@ protected double getVscLossMultiplier() {
 return (1 - lossFactor1) * (1 - lossFactor2);
 }

+ // FIXME: This method is a copy of HvdcUtils.getHvdcLineLosses() from powsybl-core but it cannot be used yet because this static method is private
+ protected static double getHvdcLineLosses(double rectifierPDc, double nominalV, double r) {


Thanks, should be fixed soon with powsybl/powsybl-core#3098, that should be included in powsbl-core 6.4.1.

You can upgrade the version, powsybl-core 6.4.1 available.

Thanks, the code now uses the method getHvdcLineLosses().

…the losses Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill · 2024-07-15T08:52:20Z

...n/java/com/powsybl/openloadflow/ac/equations/HvdcAcEmulationSide1ActiveFlowEquationTerm.java

+ if (isController(rawP)) {
+ return rawP;
+ } else {
+ double rectifierPDc = (1 - lossFactor1) * rawP;


Can you try to put more commun code in the abstract?

Some modifications have been done, especially by creating a method getAbsActivePowerWithLosses() that computes the absolute value of the active power taking into account loss of converters (in the correct order depending on the direction of the transit) and the line loss.

annetill · 2024-07-15T08:54:01Z

src/test/java/com/powsybl/openloadflow/ac/AcLoadFlowVscTest.java

@@ -450,6 +481,7 @@ void testAcEmuWithOperationalLimits() {
 Network network = HvdcNetworkFactory.createHvdcLinkedByTwoLinesAndSwitch(HvdcConverterStation.HvdcType.VSC);
 // without limit p=195
 network.getHvdcLine("hvdc23")
+ .setR(0d) //Removing resistance to ignore cable loss


As a general comment, if unit tests changed because we improve/fix something, we prefer to fix them than to retry to code the old behavior.

Ok, then the values of the tests have been updated with the new behavior (that have been verified). A new test had been added for the case without R to have an independant test ignoring line loss.

…ionTerm.java Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill · 2024-07-16T10:36:51Z

...main/java/com/powsybl/openloadflow/ac/equations/AbstractHvdcAcEmulationFlowEquationTerm.java

@@ -78,6 +83,15 @@ protected double getVscLossMultiplier() {
 return (1 - lossFactor1) * (1 - lossFactor2);
 }

+ protected double getAbsActivePowerWithLosses(double boundedP, double lossRectifier, double lossInverter) {


Just a detail here: rectifier and inverter terms seems to be used only for hvdc operating in active power set point (where you have which converter station is rectifier and which one is inverter). That is why, we prefer "controller" (see isController) for Ac Emulation because we only rely on sign of P0 + K delta angle.

Thank you for this detail, I understand now. The orders of the losses are important in this formula so the name are important. On one side there is the "controller", and which name on the other side ? What about "controlled" station ?

I think that in a previous version, I used controller vs non controller :-)

annetill · 2024-07-16T10:38:40Z

src/test/java/com/powsybl/openloadflow/sa/OpenSecurityAnalysisWithActionsTest.java

@@ -1284,6 +1284,7 @@ void testVSCLossAcEmulation() {
 // but other converter station keeps its voltage control capability.
 // remedial action re-enables the ac emulation of the hvdc line.
 Network network = HvdcNetworkFactory.createHvdcLinkedByTwoLinesAndSwitch(HvdcConverterStation.HvdcType.VSC);
+ network.getHvdcLine("hvdc23");


Line to be removed?

Yes, it had to be removed.

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill · 2024-07-16T12:51:52Z

...n/java/com/powsybl/openloadflow/ac/equations/HvdcAcEmulationSide1ActiveFlowEquationTerm.java

- return (isController(rawP) ? 1 : getVscLossMultiplier()) * boundedP;
+ double boundedP = boundedP(rawP(ph1, ph2));
+ // if converterStation1 is controller, then p1 is positive, otherwise it is negative
+ return isController(boundedP) ? boundedP : -getAbsActivePowerWithLosses(boundedP, lossFactor1, lossFactor2);


Why did you change the check isController from rawP to boundedP?

Because isController() only checks the sign of the power transit. I thought there was no need to store rawP instead of directly checking the sign of boundedP.

I think it is equivalent but as Didier is not here (these lines about boundedP are from him) maybe keep former lines.

Yes it is equivalent (unless bounds of P are negative which they shouldn't). But I modified in order to keep the former lines.

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

sonarcloud · 2024-07-16T14:16:38Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
100.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud

annetill

Thanks for your first contribution!

SylvestreSakti added 4 commits July 10, 2024 10:14

Added HVDC line loss in AC emulation

67346d7

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Added newline for checkstyle

7b60f08

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Added newline for checkstyle

a15c4e7

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Modified tests to keep same behaviour with HVDC line resistance in AC…

f6c740b

… Emulation Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

SylvestreSakti force-pushed the hvdc-line-loss-ac-emulation branch from 00ec727 to f6c740b Compare July 10, 2024 09:23

annetill reviewed Jul 10, 2024

View reviewed changes

SylvestreSakti added 2 commits July 10, 2024 14:39

Keeps r in ohms and its getter returns value in per unit

cf2eb5f

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Modified implementation of line losses with the method getHvdcLineLosses

fba1fc4

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill mentioned this pull request Jul 10, 2024

Make method to compute hvdc line losses public powsybl/powsybl-core#3098

Merged

7 tasks

Added FIXME comment to use HvdcUtils.getHvdcLineLosses() from powsybl…

ba21575

…-core Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill reviewed Jul 10, 2024

View reviewed changes

SylvestreSakti and others added 6 commits July 10, 2024 17:25

Modified one test to check the correct sign of the influence of R in …

86189c4

…the losses Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Checkstyle corrections

a7e9dac

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Added comments to explain test of difference between R=0 and R!=0

b6707f5

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Use of HvdcUtils.getHvdcLineLosses with new release of powsybl-core

65069d7

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Fixed for checkstyle

90809f6

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Merge branch 'main' into hvdc-line-loss-ac-emulation

35bdfbb

SylvestreSakti changed the title ~~Draft: Computing Hvdc line loss in AC emulation~~ Computing Hvdc line loss in AC emulation Jul 15, 2024

annetill reviewed Jul 15, 2024

View reviewed changes

SylvestreSakti added 2 commits July 15, 2024 11:45

Modified to have more common code in AbstractHvdcAcEmulationFlowEquat…

860ee50

…ionTerm.java Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Fixed tests to take cable loss into account

e9cb339

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

SylvestreSakti changed the title ~~Computing Hvdc line loss in AC emulation~~ Draft: Computing Hvdc line loss in AC emulation Jul 15, 2024

SylvestreSakti added 2 commits July 15, 2024 14:57

Changed signs to avoid confusion

a2fb18c

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Added assert to check sign of powers

74dface

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

SylvestreSakti changed the title ~~Draft: Computing Hvdc line loss in AC emulation~~ Computing Hvdc line loss in AC emulation Jul 15, 2024

SylvestreSakti and others added 3 commits July 15, 2024 17:05

Rename methods and added comments

50ecc4d

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Fixed error in checking sign of losses

8c2af3f

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Merge branch 'main' into hvdc-line-loss-ac-emulation

423aeeb

annetill reviewed Jul 16, 2024

View reviewed changes

SylvestreSakti added 3 commits July 16, 2024 13:43

Removed unused line

b8dc6dd

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Renamed variables

6fc2c5a

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

Renamed variable for better comprehension

47a137e

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill reviewed Jul 16, 2024

View reviewed changes

Changed boundedP to rawP in checking controller station

e8fd904

Signed-off-by: PRABAKARAN Sylvestre <sylvestre.prabakaran@rte-france.com>

annetill approved these changes Jul 18, 2024

View reviewed changes

annetill merged commit f785efd into main Jul 18, 2024
7 checks passed

annetill deleted the hvdc-line-loss-ac-emulation branch July 18, 2024 12:59

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Computing Hvdc line loss in AC emulation #1062

Computing Hvdc line loss in AC emulation #1062

SylvestreSakti commented Jul 10, 2024 •

edited

Loading

annetill Jul 10, 2024

SylvestreSakti Jul 10, 2024

annetill Jul 10, 2024

SylvestreSakti Jul 10, 2024 •

edited

Loading

annetill Jul 10, 2024

annetill Jul 12, 2024

SylvestreSakti Jul 12, 2024

annetill Jul 15, 2024

SylvestreSakti Jul 15, 2024

annetill Jul 15, 2024

SylvestreSakti Jul 15, 2024

annetill Jul 16, 2024 •

edited

Loading

SylvestreSakti Jul 16, 2024

annetill Jul 16, 2024

SylvestreSakti Jul 16, 2024

annetill Jul 16, 2024

SylvestreSakti Jul 16, 2024

annetill Jul 16, 2024

SylvestreSakti Jul 16, 2024

annetill Jul 16, 2024

SylvestreSakti Jul 16, 2024

sonarcloud bot commented Jul 16, 2024

annetill left a comment

Computing Hvdc line loss in AC emulation #1062

Computing Hvdc line loss in AC emulation #1062

Conversation

SylvestreSakti commented Jul 10, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

SylvestreSakti Jul 10, 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

annetill Jul 16, 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

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sonarcloud bot commented Jul 16, 2024

Quality Gate passed

annetill left a comment

Choose a reason for hiding this comment

SylvestreSakti commented Jul 10, 2024 •

edited

Loading

SylvestreSakti Jul 10, 2024 •

edited

Loading

annetill Jul 16, 2024 •

edited

Loading