Improve DC sensitivity analysis with HVDC in AC emulation #1293
Description
Describe the current behaviour
Currently, DC sensitivity analysis silently ignores the hvdcAcEmulation.
This can be misleading. Because:
- The HVDC injection is its fixed point and does not depend on the angle
- The fact that the HVDC injections varies with a perturbation is not taken into account in the sensitivites
This issue is related to #1016 but in the scope of sensitivity analysis.
Describe the expected behaviour
The purpose of this issue is to define the desired behaviour. Some possible solutions that have been described:
1 raising an error if DC analysis is run with hvdcAcEmulation set to True and if the network contains HVDC with AC emulation enabled.
Or
2 Considering HVDC in AC emulations as lines, with a flow proportional to the angle differenc. Ignoring the Pmax and Pmin. (For use cases where HVDC are not saturated)
Or
3 Computing the DC flow for HVDC lines in AC emulations. Then consider them in fixed point mode when computing the other sensitivites (for use cases where the sensitivity is desired before the HVDC emulation stabilization)
Or
4 Comptuting the DC flow of HVDC lines in AC emulation, taking into account that they can be saturated at PMax. And using this state to compute the sensitivity (for use cases that may have a large HVDC flow, that sill want the use the small angle approximation where DC is valid, and that want to distinguish cases where the HVDC is saturated form cases where it is not)
Describe the motivation
Avoid misleading results. If possible take into account HVDC emulation in DC sensitivity analysis.
Extra Information
No response