One of the optimisation contraint of an SSN model is the number of switches per SSN group. User will want this number of switch per SSN group to be below 12 typically. If not, some additional SSN group have to be creates, possibly increasing the number of SSN node and simulation time-step.
One optimisation that can be made in this regard is to embed switches or breakers into an existing device like the SSN Induction machine (SSN-IM).
Image 1. Comparison of two models: SSN-IM with embedded breaker (top), standard SSN-IM (bottom)
In Image 1, we have compare the standard SSN-IM with the one with embedded breakers. The impact is that the SSN group connected with the SSN-IM with embedded breaker has less switches, because it was 'transfered' inside the SSN-IM.
The effect of this is that the user can add 3 more switches into the SSN group left of the SSN-IM without negative impact on performance.
Methodology
The SSN-IM uses a stator fixed reference frame D-Q transform internally. What was done here is that we added the breakers resistance in series with the stator resistance inside the D-Q transform. The breaker resistances (3 resistances) are transformed into the D-Q frame with P*Rbrk*Ptr, where P is the fixed-referecne frame D-Q matrix. (Ptr is the transposed of P) and Rbrk is the vector 3 breaker resistance vector.
The methodology is quite simple and efficient numerically.
By comparison, if we were to apply this technique to Distributed Parameter Line (DPL) model, the approach could be less effective because DPL have an impedance that is reference to ground and computing the model+breaker equivalent impedance can require some inverse computation.
Availability
This SSN-IM with embedded breaker model will be available in early 2020 in the next release of ARTEMiS.