Monday, February 04, 2013

Guest Repost: The Night the Lights Went Out on the Super Bowl

We got an email this morning from the Richard Cadena's Technopolis newsletter which had a great article about the Superbowl power outage last night.  For more info or to subscribe check out the Academy of Production Technology website.

The Night the Lights Went Out on the Super Bowl   by Richard Cadena

In case you missed it, nothing much happened during the Super Bowl after Jacoby Jones ran back the kickoff that started the third quarter. That's because the electricity in the Super Dome failed, leaving two teams and 100,000 fans mostly in the dark for about half an hour. What caused the failure?

I have no specific knowledge of the reason behind the event, but I do know that just a few minutes prior to the failure, Beyonce was rocking the 50 yard line with an impressive array of lighting and video displays. Could the failure be linked to the halftime show? Perhaps.

While massive Super Bowl halftime shows are nothing new, what is new is the predominance of LEDs, switch-mode power supplies, and pulse-width modulation. What do all of these have in common? They are all non-linear loads, which produce harmonics that can cause excessive current to flow in the neutral conductor of a 3-phase system. If that's not taken into account, then it can lead to failures of components like feeder transformers.

What's a non-linear load? It's anything that's plugged into the power distribution system that draws current in a waveform that's different than the applied voltage waveform. LEDs are non-linear loads because they convert AC to a pulse-width modulated current in order to drive the emitters and change their intensity. When the waveform changes from a sine wave to any other waveform, it produces harmonics, which are whole number multiples of the fundamental frequency. In North America, the fundamental frequency is 60 Hz and in Europe and Australia it's 50 Hz. Harmonics are currents, just like the fundamental current, except they have frequencies of 2 × the fundamental, 3 × the fundamental, 4 × the fundamental, etc. And just like any other current, they follow the same path as the fundamental, combining in the neutral conductor of a 3-phase system.

When the three phase currents return to the supply, they do so through the neutral conductor. Since the fundamental phase currents are 120˚ out of phase with each other, in a balanced 3-phase system they exactly cancel. The same is true of the second harmonic, except that when the frequency doubles, so does the phase angle between them. 

But the third harmonic is the villain. The third harmonic is three times the fundamental frequency, or 180 Hz in North America, and the phase angle is 120 Hz × 3, or 360˚. But since we're dealing with sine waves, 360˚ is exactly the same as 0˚, which means the phase current in every third harmonic is in phase with each other, so they reinforce in the neutral conductor. And that can cause a huge amount of current to flow in the only conductor that has no overcurrent protection.

What's puzzling is why the failure occurred after the halftime performance and not during it. It could be that the failure was unrelated to the halftime show, or it could be that the heat produced by the current draw from the show took some time to cause the failure. Any conclusion is purely speculation at this point, but my money is on a feeder transformer failure due to miscalculation of current draw.

The nature of our loads in the entertainment industry is changing. Unless we learn to deal with a largely non-linear system, we're doomed to repeat our failures.