Support of BusVoltagePerTargetV relalted to a transformer #299

annetill · 2021-04-29T15:02:42Z

Please check if the PR fulfills these requirements (please use '[x]' to check the checkboxes, or submit the PR and then click the checkboxes)

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 ? If so, link to this issue using '#XXX' and skip the rest

We don't support it.

What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)

We support it.

What is the current behavior? (You can also link to an open issue here)

We through an exception.

What is the new behavior (if this is a feature change)?

We support it. We first compute the reference voltages like in a classical load flow with the transformer control outer loop active. Then we have to rebuild the equation system, as it was before the transformer steps were rounded and the PV buses deactivated. From that equation system, we compute the sensitivity values.

But:

The equation system is not updated, but rebuild from scratch ;
I don't use the equation system listener.
So it could be improved.

Does this PR introduce a breaking change or deprecate an API? If yes, check the following:

The Breaking Change or Deprecated label has been added
The migration guide has been updated in the github wiki (What changes might users need to make in their application due to this PR?)

Other information:

(if any of the questions/checkboxes don't apply, please delete them entirely)

…o tap changers. Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

src/main/java/com/powsybl/openloadflow/sensi/AcSensitivityAnalysis.java

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

src/main/java/com/powsybl/openloadflow/ac/equations/AcEquationSystemUpdater.java

src/main/java/com/powsybl/openloadflow/sensi/AbstractSensitivityAnalysis.java

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

sylvlecl

Nice!
I think we need to add some test/handling for post contingencies sensitivities though.

sylvlecl · 2021-06-01T11:36:07Z

src/main/java/com/powsybl/openloadflow/sensi/AcSensitivityAnalysis.java

@@ -180,6 +180,12 @@ public void analyse(Network network, List<PropagatedContingency> contingencies,
                .collect(Collectors.toSet());

        // create AC engine
+        boolean hasTransformerBusTargetVoltage = hasTransformerBusTargetVoltage(factorHolder, network);
+        if (hasTransformerBusTargetVoltage) {
+            // if we have at least one bus target voltage linked to a ratio tap changer, we activate the transformer


Maybe we should do the opposite, I mean compute the sensitivities ONLY if voltage control has been turned on.

Otherwise, the behaviour can be weird, because taps will be changed whereas we did not explicitly ask them to, we only asked to compute sensitivities. The resulting voltages will be different from a loadflow with the same parameters.

So you think that the user should give a bus target voltage related to a transformer AND explicitly says that transformer control is enabled?
What I don't get in your point is: what is for you the starting point to compute sensitivity values? In AC, it is quite important. In the code, the starting point is after transformer rounding. You prefer having both modes: after the rounding if we have the control enabled in the parameters or just from the initial network?

No no I think it's perfectly OK to compute sensitivities after the rounding.

What I mean is that:

if you run a simple loadflow with voltage control disabled, you will get voltage values without tap changes, OK

if you run a sensitivity analysis with the same parameters (=voltage control disabled), now you will get different values for voltages (and sensitivities around those different values), because taps will have moved

I think it's weird, if you run the computation with the same parameters, I would expect to obtain the same voltage values.
If voltage control is disabled, I think I would expect to have 0 as sensitivity value (if we do this, maybe we could add a WARN log ?).

Yes I can do that.

Second thought: for sensitivity calculation linked to a phase shifter, we don’t use the loadflow parameters. We can compute the sensitivity even if the regulation is not active. So it is still coherent to rely on the loadflow parameter on our use case, I think yes but what is our opinion?

sylvlecl · 2021-06-01T12:06:01Z

src/main/java/com/powsybl/openloadflow/sensi/AcSensitivityAnalysis.java

+            // if we have at least one bus target voltage linked to a ratio tap changer, we have to rebuild the AC equation
+            // system obtained just before the transformer steps rounding.
+            if (hasTransformerBusTargetVoltage) {
+                for (LfBus bus : lfNetwork.getBuses()) {


I am not sure how it will behave on contingencies :
I guess the mode will be turned OFF again ? so maybe it does not work for the post contingency sensitivities ?

I will add a test but it will be similar to the generator behaviour.

sylvlecl · 2021-06-01T12:08:53Z

src/test/java/com/powsybl/openloadflow/sensi/ac/AcSensitivityAnalysisTest.java

@@ -347,6 +346,41 @@ void testBusVoltagePerTargetV() {
    @Test
    void testBusVoltagePerTargetVTwt() {


Nice :)
I think we should add a test with a contingency

sylvlecl · 2021-06-01T12:46:16Z

src/test/java/com/powsybl/openloadflow/sensi/ac/AcSensitivityAnalysisTest.java

+        assertEquals(4, result.getSensitivityValues().size());
+        assertEquals(0d, getValue(result, "T2wT", "BUS_1"), LoadFlowAssert.DELTA_V);
+        assertEquals(0.035205d,  getValue(result, "T2wT", "BUS_2"), LoadFlowAssert.DELTA_V);
+        assertEquals(1d,  getValue(result, "T2wT", "BUS_3"), LoadFlowAssert.DELTA_V);


Just a thought:
I had a doubt (not obvious when reading the code), that sensitivities are correctly "scaled back" from per-unit.
Finally it seems correct, but it looks like the logic is handled directly in the equation we solve for computing the sensitivity : I think at some point we should make more clear where is the limit between per-united values and "nominal" values (probably having a kind of conversion after having computed the sensitivities in per-unit).

It is done in calculateSensitivityValues but it is not well done for BUS_VOLTAGE and I have some doubt about sensi += Math.toRadians(factor.getEquationTerm().der(phi1Var)); line but no link with that PR :-)

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

sonarqubecloud · 2021-06-03T13:38:24Z

Kudos, SonarCloud Quality Gate passed!

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

97.0% Coverage
0.0% Duplication

First import of BusVoltagePerTargetV linked to transformers with rati…

e9a15ae

…o tap changers. Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

annetill commented Apr 29, 2021

View reviewed changes

src/main/java/com/powsybl/openloadflow/sensi/AcSensitivityAnalysis.java Outdated Show resolved Hide resolved

annetill requested a review from flo-dup April 29, 2021 15:06

geofjamg self-requested a review May 17, 2021 20:40

annetill added 3 commits May 28, 2021 11:34

Merge.

20a1cd3

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

Use the equation system listener instead of outerloop active/inactive.

4e10e0c

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

Remove useless code.

d427134

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

annetill requested a review from sylvlecl May 28, 2021 11:58

Increase unit test.

d92a656

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

geofjamg approved these changes Jun 1, 2021

View reviewed changes

Merge branch 'master' into sensi_v_tap_changer_2

cf84372

flo-dup approved these changes Jun 1, 2021

View reviewed changes

src/main/java/com/powsybl/openloadflow/ac/equations/AcEquationSystemUpdater.java Outdated Show resolved Hide resolved

src/main/java/com/powsybl/openloadflow/sensi/AbstractSensitivityAnalysis.java Outdated Show resolved Hide resolved

Small changes after review.

beedf37

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

sylvlecl reviewed Jun 1, 2021

View reviewed changes

Support of contingencies.

1935e3e

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

sylvlecl approved these changes Jun 3, 2021

View reviewed changes

annetill and others added 2 commits June 3, 2021 15:09

Fix code smell.

2f4607c

Signed-off-by: Anne Tilloy <anne.tilloy@rte-france.com>

Merge branch 'master' into sensi_v_tap_changer_2

ffc0403

annetill merged commit 6f7a2d1 into master Jun 3, 2021

annetill deleted the sensi_v_tap_changer_2 branch June 3, 2021 13:44

annetill mentioned this pull request Jun 17, 2021

AC sensitivity analysis: support of a targetV increase of a ratio tap changer #292

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Support of BusVoltagePerTargetV relalted to a transformer #299

Support of BusVoltagePerTargetV relalted to a transformer #299

annetill commented Apr 29, 2021

sylvlecl left a comment

sylvlecl Jun 1, 2021

annetill Jun 1, 2021

sylvlecl Jun 1, 2021

annetill Jun 1, 2021

annetill Jun 1, 2021

sylvlecl Jun 1, 2021

annetill Jun 1, 2021

sylvlecl Jun 1, 2021

sylvlecl Jun 1, 2021

annetill Jun 1, 2021

sonarqubecloud bot commented Jun 3, 2021

		@@ -347,6 +346,41 @@ void testBusVoltagePerTargetV() {
		@Test
		void testBusVoltagePerTargetVTwt() {

Support of BusVoltagePerTargetV relalted to a transformer #299

Support of BusVoltagePerTargetV relalted to a transformer #299

Conversation

annetill commented Apr 29, 2021

sylvlecl left a 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

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sonarqubecloud bot commented Jun 3, 2021