What a Catastrophe! phase-to-phase fault occurs, 20 times normal current, fault lasts FOREVER- 1.7 sec, 105 cycles. 1 cycle is enough for a fire.

What a Catastrophe! phase-to-phase fault occurs, 20 times normal current, fault lasts FOREVER- 1.7 sec, 105 cycles. 1 cycle is enough for a fire.

An initial phase-to-phase fault (AB) occurred on a primary zone of feeder F. The feeder and the transformer both capture the events.

However, the event lasts forever to the point that the relays capture 4 consecutive events (SEL calls them back-to-back events) for the feeder and the transformer relay. 8 total events are captured.

What a catastrophe!

How long do we want a fault to last in any power system? 0 cycles. Yes, we never want a fault in the system.?Unfortunately, because of the breaker time to operate, we have no choice but to incur time before the fault is removed.

The industry started with 5 cycles breakers at medium to high voltage, now 3 cycles are the standard. 2 cycles breaker is the top of the line. Currently, several companies are working on creating 1 cycle breaker. The breaker industry recognizes that a difference in 1 cycle is enough to save a life, save equipment, and limit financial losses.

IEEE buff book recommends no longer than 3 cycles for the relay to remove a fault in your primary zone. Assuming 3 cycles breakers, it means the recommendation is at most 6 cycles for a fault to last.?At best, you should remove a fault within 3 cycles, meaning no delay time for the relay.

The following 4 pictures are one fault event captured in 4 back-to-back reports. I alternate colors black and white for lesser monotony.

Notice how the fault starts in the first report at 8:20:23.941000 (Tref). The format means 8 hours 20 min 23.941000 seconds.

In the second report (first white graph), the fault is not removed. IC has now increased to the point where the fault evolves to a 3 phase.

In the third report (second black graph), the fault is still there. We now have almost a 3 phase fault. IC is high but not as high as IA and IB. On the third event, at least the trip is finally initiated.

In the fourth report (second white graph), the fault is now removed. 52A is gone, meaning the breaker is opened—the time shows 8:20:25.671000 (Tref).

If we do the math to find the length of the time the fault lasted, 25.671000 – 23.941000 = 1.73 seconds. 1.73 * 60 = 103.8 cycles

It took 1.73 seconds, 103.8 cycles from when the fault occurred to when the breaker opened. RIDICULOUS!

Something Important to Notice

On the last 2 graphs, 2 inputs were not present previously, IN103 and IN104. They assert at the end of the 3rd event and drop after the breaker open. Something was burning, and those inputs were related to hot temperatures.

Keynotes

-?????Highest fault current occurs during the 3rd report. IA = 5800 A.

-?????Fault current was 23 times the normal current.

- Fault lasts 103.8 cycles. 34th times slower than the max achievable speed of 3 cycles.

-?????Fault occurs at the primary zone of Feeder F. no need for delay or coordination.

-?????There were no equipment failures. Every breaker and relay operated as designed.

This event is not a one-off, not a mistake; it is a systematic error. I have seen thousands and thousands of such events. Mistakes and errors are not synonyms. For the next 10 years, you will see numerous similar articles in this newsletter. I intend to write 1,024 articles, 1 article per week.

It is HIGHLY UNLIKELY to accurately design high complexity power systems solely based on input data in simulation software.?

It is HIGHLY UNLIKELY to accurately review the accuracy of power studies solely based on verifying if the input of the software is correct.

In power system modeling and studies, "Garbage in, Garbage out" does not mean "Not Garbage in, Not Garbage Out."?

I have seen numerous incidents caused by inaccurate power systems simulations:

1-???Arc flash incident energy rating was incorrect and led to a permanent disability of a worker during an arc flash fault

2-???Vestas V120 turbines total damage

3-???Nordex turbine N90 total damage

4-???Cogen generator windings failure

5-???Pregnant worker escaped death during arc flash incident. Label was incorrect

When the initial power system simulation of a facility is flawed, probably, the flaws would never be fixed. Every subsequent power study will assume that most of the previous simulation is correct and build on the same flaws.

When you multiply a decimal number by itself, the result is lower than the original number. 90% accuracy times 90% accuracy = 81% accuracy.

The consequences of this event were disastrous. I would not have time to discuss it in this article. See you next month for the follow-up; I will also include the 4 event reports for the 100MVA transformer.

Inadequate power simulations and inadequate power studies are the primary reason our equipment fails. Unfortunately, we are not aware, and we do not have the best practices to determine such gross flaws.

Feel free to subscribe to the newsletter and share it with your colleagues. Remember, Knowledge is Power and Power is Our ExpErtisE