Protection, arc flash mitigation using internal VFIs in liquid substation transformers

From our previous discussions about the Cooper HDC liquid transformer with internal, under-liquid VFI’s, expanding the use of the VFI takes this concept a step further, and makes a great way to mitigate arc flash hazards on the secondary.

By Joe Guentert August 29, 2012

David Durocher of Eaton (assisted by Warren Hopper of Weyerhaeuser and Dave Shipp of Eaton), published an article in June 2011 describing a fairly unusual case of secondary unit substations without secondary main breakers (which was common 30 to 40 years ago under NEC’s “Six-Handle Rule”), and how overcurrent relaying could be applied on the secondary bus to trip a protector on the primary side of the transformer. The result was that very effective and sensitive protection could be provided for both secondary faults and internal winding faults, even without having a secondary main breaker.

I love this protection concept (works just as well – maybe even better – if you DO have a secondary main breaker). I’ve been doing essentially the same thing in most data center projects for about 10 years now, having employed this scheme on approximately one-hundred 5-MVA data center substations that had large fault duties on 600-V secondary buses.

David’s article described VPI dry-type transformers in unit substation configurations, in which a compact VCP-TL linear actuator breaker with an isolating switch in the same section were added on the transformer primary to provide good isolation and excellent protection, replacing the primary load interrupter switches and CL fuses. An elegant, smart, simple, and very safe scheme, I think. I’m happy to see that Eaton thought well enough of the approach to develop into a standard product offering called “MSB.” I believe that this product can literally save lives, and prevent serious injuries and major equipment damage at large data centers and other facilities.

From our previous discussions about the Cooper HDC liquid transformer with internal, under-liquid VFI’s, expanding the use of the VFI takes this concept a step further, and makes a great way to mitigate arc flash hazards on the secondary.

Here’s an example (Figure 1), based upon a real data center project design I saw this earlier year, that used 2.5 mVA 115 C rise VPI dry-type substation transformers, 24.9 kV primary, close-coupled to 480 V secondary switchgear, with solidly-grounded 480 Y/277 V secondary windings. The design used two interlocked 600-amp primary air switches, with a set of 100E current-limiting fuses to protect the transformer against primary and secondary faults, with upstream 27 kV metalclad switchgear feeding transformer primary loops.

Let’s say that the transformer is energized, and operators wish to rack the 4,000-amp secondary main breaker, and that as they do so, one of the drawout finger clusters on the line side of the breaker misaligns, and creates a line-to-ground fault. (This can and DOES happen – not frequently, but often enough to be of concern, and always with EXTREMELY serious consequences. It’s everyone’s worst nightmare, because there is no way to shut it down until after a lot of equipment has been destroyed, or personnel injured. Generally, it also happens to be the very location of the greatest arc-flash hazard in the entire data center.)

The time-current curves below show the result. The 100E primary fuses provide virtually NO protection against this condition. The fuses would not clear this fault until after the transformer had been destroyed. The arcing secondary ground fault would burn for more than 4 min (about 250 sec, in fact, unless another phase became involved), and in this case, the arc-flash hazard boundary on the secondary makes the secondary switchgear bus “unapproachable”. No PPE presently available would adequately protect operators from personal injury from this fault, and the 100E fuses are the only protectors anywhere in the system that could possibly clear this fault.

If these were instead the Cooper HDC liquid transformers with an internal VFI we’ve been discussing, the VFI would clear this same fault within about 3 sec on its normal curve (Red curve on Figure 2).

Now, consider an overcurrent relay with a selectable Maintenance Mode settings group on the secondary, supplied from current transformers slipped over the transformer’s secondary bushings, and connected to trip the primary VFI (the very same concept as described in the Dave Durocher article). With the Maintenance Mode settings group activated by a selector switch, this same fault would be cleared in 0.080 sec (Green curve on Figure 2), and the arc flash hazard would be reduced by more than 99.9% from the hazard level of the 100E fuses.

All in all, it’s the same great result described in Dave Durocher’s article, except that the liquid transformer option allows the VFI to be located entirely inside the transformer tank, where it adds not much cost, and adds zero extra footprint.

The VPI dry-type design example used in Dave’s article requires addition of an external, floor-standing primary breaker and an isolating switch, requiring additional floor space, and some additional cost, per unit substation. It’s a perfect solution for a dry-type or cast-coil transformer, but the same approach becomes much simpler, more compact and much less expensive in a liquid transformer having a primary VFI in the tank – and the VFI is available in transformers with primary voltages up through 35 kV.