Transformer de-energisation transient response behaviour

[LiteYear]

I've recently taken some CRO readings of a transformer secondary during operation of a circuit breaker on the primary. The resulting voltage transients are causing problems in the field, but I don't have enough experience in this area to know whether the results are typical.

Transformer is 11kV/690V/690V/1050V, Dzn0/Dyn11/Dyn11, 1000kVA/1000kVA/500kVA. Each secondary neutral is connected via a dedicated resistor to earth, to limit earth fault to 5A. That is, 80Ω on the 690V secondaries and 121Ω on the 1050 secondary.

Circuit breaker on high tension side is a Schneider LF1 12kV 1250A 25kA.

With no load on all of the secondaries, I measured voltage from earth to neutral, phase A to neutral, phase C to neutral and current leaving the neutral on the 690V star secondary. When the circuit breaker opens, the transformer is isolated from all loads and sources, and the connected cables are quite short (less than 5m total). Waveforms such as those attached appear at the time of opening.

As you can see, the phases start with normal 50Hz oscillation, then there's a period (about 1ms) of high frequency ringing (probably mostly noise) followed by a voltage transient with a fundamental around 1000Hz. The voltages continue for another couple of milliseconds and then there's another voltage transient before they decay to zero.

I can only guess that the high frequency ringing is to do with the mechanical sliding of the CB opening mechanism, the initial transient is the resonant system response, coupled to the secondary, to the interruption of the primary magnetising current in one phase and the final transient is the resonant response to the interruption in another phase.

Worst case voltage transient we recorded was about 1900V phase to phase on the 690V winding.

The main question I have is, should we expect occasional 1900V peak transient on a 690V winding and design accordingly, or is this unexpected and to be fixed at the source?

 

[LiteYear]

I'm curious as to why this didn't attract any responses. Was it the presentation or the content? In other words, could I have asked in a better way, or is it just something that people are unaware of?

Perhaps I could try another way:

1. Suppose you had a 11kV/690V transformer on a 11kV 20kA fault level network. What phase to phase voltage would you rate components to on the 690V side? 800V? 1000V? 20kV?

2. Is there any reason to avoid disconnecting an unloaded transformer? Is there any practice of installing a permanent load on transformer secondaries, so that interruptions on the primary don't cause inductive spikes?

 

[LiteYear]

Anyone? I've heard of interruption voltage spikes causing damage to the circuit interrupting device itself. Given the transformer is supplying that voltage spike, would it (and its secondaries) be susceptible to the same spike?

 

[DRWeig]

Consider the stored energy in the transformer. If you disconnect the primary, all of that energy has to go down the secondary to the load. The 1900V spikes you see are normal for disconnecting the primary. Appropriately sized surge protectors on the secondary will knock them down, as will connected motors and downstream transformers.

In buildings, we don't expect that the service transformer primary will be opened on a regular basis. System power outages are the main culprits in producing that spike, and sensitive equipment may be damaged. BUT, most times not. Generally, sensitive equipment has its own protection in the form of surge suppressors and/or UPS units.

If your system is 600V, then design for 600V.

The primary disconnect (if it's a load-break type) will survive just fine -- it's designed for that type of service.



Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[LiteYear]

Appreciate the response DRWeig. Makes a lot of sense. Thank you.