Describe the current behavior
IIDM Non Linear Shunt Compensator always have an "implicit" zero sectionCount, meaning "no compensation" (docs).
Simulators (e.g. powsybl-open-loadflow) simulate this zero sectionCount: (either zero sectionCount is provided in input and no voltage regulation, or voltage regulation is enabled and may move solvedSectionCount to zero).
Describe the expected behavior
IIDM Non Linear Shunt Compensator:
- do not have an implicit zero
sectionCount
- zero
sectionCount and zero solvedSectionCount are invalid values. Only valid values are [1 ; maximumSectionCount]
- to model zero compensation as part of the table, a section with zero B/G values can still be defined.
The change is only for IIDM Non Linear Shunt Compensator.
For IIDM Linear Shunt Compensator, a zero sectionCount/solvedSectionCount is still valid.
Impact on:
- IIDM importer: older xIIDM serialized models with zero
sectionCount or zero solvedSectionCount would become rejected - unless we agree on a fix/warn strategy (e.g. find the section closest to zero B value + warn)
- CGMES importer: older incorrect models previously accepted may become rejected (models with cim:NonLinearShuntCompensator having
<cim:ShuntCompensator.sections>0</cim:ShuntCompensator.sections> in SSH or <cim:SvShuntCompensatorSections.sections>0</cim:SvShuntCompensatorSections.sections> in SV)
- PSS/E importer: no impact foreseen - already importing sections with zero B/G values, see e.g. here
- Simulators: (powsybl-open-loadflow, etc...): must update to the modified model.
To be discussed/validated in TSC.
A "less breaking change" could be to have a flag indicating if the zero sectionCount is allowed in the model. But this might be complicating the IIDM model unnecessarily.
Describe the motivation
Align with recent CIM/CGMES clarifications. The change applies to all CGMES versions (2.4.15, 3.x), see here lines 2431-2432. Although this extract is a bit confusing and contradictory, I got final confirmation that Non-Linear Shunt Compensators never have a zero section, for all CGMES versions.
Extra Information
No response
Describe the current behavior
IIDM Non Linear Shunt Compensator always have an "implicit" zero
sectionCount, meaning "no compensation" (docs).Simulators (e.g. powsybl-open-loadflow) simulate this zero sectionCount: (either zero sectionCount is provided in input and no voltage regulation, or voltage regulation is enabled and may move solvedSectionCount to zero).
Describe the expected behavior
IIDM Non Linear Shunt Compensator:
sectionCountsectionCountand zerosolvedSectionCountare invalid values. Only valid values are [1 ;maximumSectionCount]The change is only for IIDM Non Linear Shunt Compensator.
For IIDM Linear Shunt Compensator, a zero
sectionCount/solvedSectionCountis still valid.Impact on:
sectionCountor zerosolvedSectionCountwould become rejected - unless we agree on a fix/warn strategy (e.g. find the section closest to zero B value + warn)<cim:ShuntCompensator.sections>0</cim:ShuntCompensator.sections>in SSH or<cim:SvShuntCompensatorSections.sections>0</cim:SvShuntCompensatorSections.sections>in SV)To be discussed/validated in TSC.
A "less breaking change" could be to have a flag indicating if the zero sectionCount is allowed in the model. But this might be complicating the IIDM model unnecessarily.
Describe the motivation
Align with recent CIM/CGMES clarifications. The change applies to all CGMES versions (2.4.15, 3.x), see here lines 2431-2432. Although this extract is a bit confusing and contradictory, I got final confirmation that Non-Linear Shunt Compensators never have a zero section, for all CGMES versions.
Extra Information
No response