Saturable Transformer with Continuously Variable Turn-ratio Using SSN External Group Feature

The circuit shows the capability of ARTEMiS-SSN to interface with the user's custom nodal code with a Simulink S-function builder block. The case of a saturable transformer with a continuously variable turn-ratio is studied here.

Model Explanation

This model implements a single-phase saturable transformer using SSN external group feature used for current transformer applications.

Because the application is of a current transformer, the user should note the very low core resistance and inductance.

  Figure 1: Current transformer application

The transformer is coded as an SSN group using the latest SSN methodology without S-function. It is coded using Simulink blocks to enable the continuous variation of secondary resistance and inductance as well as the transformer turn-ratio. This feature can be very useful in many applications like On Load Tap Changing (OLTC) transformers.

Inside the transformer block, you can study the SSN code used to implement this feature. It is based on the state-space representation of a transformer.

The Custom User Coded SSN model can be compared with an OLTC made with regular transformer and switches to implement the various taps. Common 3-phase OLTC can have several taps that would require hundreds of switches to implement them; too many switches can handicap the real-time performance of the model. With the SSN Custom Coded model, there are no switches, thus no limitation for real-time.

How to convert saturation data from the SSN model to SPS model?

Question: the OLTC SSN model specifies the transformer saturation with parameters: 'Non-saturated magnetization inductance' (nsmi), 'Saturated magnetization inductance' (smi) and 'Flux saturation threshold' (ft) while the SPS models usually specify the saturation as pairs of (current, flux) points. How do we converter from SSN to SPS?

Answer: the SPS model saturation data mimic the SSN data with 3 pairs of (current, flux) points. The first pair is (0,0).

The second pair must be set to (ft/nsmi, ft)

The third pair can be set to (ft/smi*k, ft*k) with k>>ft (ex:k=1000)

This is an approximation that produces very good accuracy.

Model notes