EMT tool integration to co-simulation

[Kali-Nath]

Hello Everyone,
I have a quick question to clarify from other HELICS users, has anyone here used HELICS to integrate an EMT tool (RTDS/PSCAD or Simulink) with an RMS tool (PSS/E or equivalent)? If you did, how did you addressed the time step issues and difference in measured quantities? do you apply FFT inside EMT tool and convert it into a phasor value before sending it to RMS? does HELICS has any native methods to implement the connection seamlessly? Any input on this is greatly appreciate.

[phlptp]

helics_PSCAD.pdf
In HELICS the federates don't need to have the same time stepping, all federates control their own time so that isn't much of an issue as far as HELICS is concerned. The attached document looked at connecting PSCAD to a communication federate. HELICS also handles unit translation so that is some of it. Otherwise there was either an intermediate federate that did data computation or just done in the federates themselves. We also have a simHELICS which has simulink integration, though testing and use has been limited, it has been used for some applications. In a couple cases we also coupled to an FMU which was doing an EMT simulation of some components. I also believe RTDS has been connected to other components via HELICS, though I don't know the details.

There are lots of possibilities, I don't think there has been a standardization on the best practices for this kind of thing yet, but HELICS has the mechanics to test out a bunch of different methods.

[Kali-Nath]

Thank you very much for your detailed explanation, I really appreciate it.
A quick follow up question, if we are integrating EMT via an intermediate federate (which I believe is what the document has shown with UDP interface instead of ZMQ), how do we ensure that the simulation control from HELICS. From what I understand, the strong aspect of HELICS is the simulation control, where all the tools are synchronously executed by requesting the simulation time from HELICS, if we allow EMT to do their own thing and extract and pump values using intermediary, will that still yield co-simulation purpose? just thinking out loud. What's your take on this?

[phlptp]

When we have done intermediaries before they act as additional federates so are fully synchronized with the other federates. So in this case the EMT would be a federate, so would an intermediary. The intermediary only publishes at a much lower rate but runs in synchronization with the EMT pulling in the data and doing the computation necessary for coordination with slower federates. The timing on this can be coordinated so they effectively can run in parallel. You wouldn't let the EMT federate free run it would still be synchronized to the extent needed. For some real time applications the cosim could be linked to wall clock, then you could let the EMT simulator "free run" This is sometimes how the coordination with and RTDS or OpalRT is handled.

[Kali-Nath]

Thank you very much for sharing your knowledge, it's really helpful.