Liquid transformers–the right stuff
In this installment, Guentert gives more reasons for the differences in reliability between liquid transformers and dry-types.
We left off last week discussing some of the possible reasons for differences in reliability of liquid transformers versus dry-types, when primary windings are switched by vacuum circuit breakers (all my personal opinions, again). Here are more reasons:
1. On all data center projects I’m involved with, we very carefully watch the installation details of the “system,” not just the transformer itself. Using the “loadcenter” concepts discussed earlier, moving the transformations into very close proximity to the loads has the effect of not only reducing high costs of the large secondary feeders, but generally has the added benefit of LENGTHENING the primary feeders, creating “Nature’s Own Snubber” by adding shunt capacitance to the shielded MV cables.
2. Then, we always connect a very good set of MOV surge arrestors DIRECTLY to the primary winding terminals. The capacitance of the cables flattens out the extremes of (dv/dt) rate-of-rise, to allow the arrestors to do their job in clamping the surge voltage as the transient develops across the windings. (Many consulting engineers still place arresters on the load runback connections of upstream vacuum breakers – at the source ends of the cables – where they are ineffective. Maximum effectiveness is when the arresters are connected directly to the transformer primary windings. Remember, this is not a transient voltage entering the windings from upstream at the vacuum breaker. Instead, the transient is developed within and across the winding itself, and this event occurs with such extreme speed that the winding insulation is damaged before the voltage wave can travel all the way back upstream through the cables to the arresters at the breaker.)
3. And, also importantly, we use only vacuum breakers from manufacturers who have performed their due diligence through R&D to ensure that the metallurgy of the power contacts inside their vacuum bottles has the “right stuff” to minimize current chopping and re-ignition phenomena during switching operations.
4. Almost all modern VPI dry-type transformers are ventilated and air-cooled, usually by simple convection, sometimes with supplemental forced air from cooling fans. Either way, air is drawn into the lower vents, then either convects or is blown by fans directly over the transformer windings. The problem is, as that happens, microscopic particles of airborne contaminants are deposited on the primary windings, reducing the overall dielectric strength of the insulation system over time, and encouraging partial discharges during large transient events. If a ventilated transformer begins its life with a primary BIL of 110 kV, the BIL can degrade significantly within a few years, unless there is very frequent cleaning and maintenance. (It’s my experience that the majority of the failures I investigated occurred between one and three years after the failed unit had been initially placed into service. When I asked the question, I learned that practically none of them had been ever been maintained and cleaned.)
Liquid transformers, on the other hand, have their windings immersed in clean fluid within a sealed tank, so external airborne contaminants and moisture can’t get to the windings. (Use of an epoxy cast resin transformer also eliminates much of this problem.)