Lighting troubleshooting

[wirenut1980]

One final thought..have any of the ballasts been replaced to see if they are the problem? Maybe try replacing one ballast and bulb with new, different manufacturer. Otherwise if you have exhausted all troubleshooting with the tools you have...

Unfortunately in problems like yours, unless you have specifically experienced the same problem before and already know the solution, you pretty much have to get a long term (at least one week) look at the voltage and current. You also need a monitor that can pick up and record short term (waveforms, looking at cycle by cycle) anomalies in the voltage and current. Anything else and you would be guessing. When "selling" a power quality monitor to management you can put it like this:

There are a few options:

1) Spend $5,000 now on a monitor and have a quality, valuable tool that can be used for years.*

2) Pay thousands of dollars over the next few years on replacement light bulbs, since they keep failing all the time. If there are any other problems that are unknown with the power, add in cost of downtime to this.

3) Pay someone else thousands of dollars to come in and monitor for you with their power quality analyzer.


*Caveat to number 1...it will take some time for you to learn to properly set up a power quality monitor so that the monitor will retain the information you want and not the added stuff you do not care about.

 

[drbond24]

One final thought..have any of the ballasts been replaced to see if they are the problem?

I asked the purchasing guys yesterday to buy 3 or 4 ballasts. We'll see what happens with that.

I actually formulated a theory yesterday that I think explains both problems in the OP. When our plant is operating, we run right around 480 volts phase to phase. However, when the plant goes down the voltage on the line side of our main breaker goes up to 500 or more. When I was measuring voltages in the warehouse yesterday I noticed that the phase to phase voltage is 500 over there under normal circumstances. I bet that when the plant goes offline the voltage over there jumps up to at least 520 phase to phase. Considering that would have been happening for around 7 years, we've probably just slow cooked all of the ballasts. It also explains why the warehouse goes off eventually after the plant does; when the weakest ballast finally fails it trips the breaker. I could be something as simple as poor coordination between protective devices that takes down the whole building instead of one circuit.

Might not be right, but gives me somewhere to start. Next time the power goes off in the plant I'm going to run to the warehouse and measure voltage instead of running to the plant switchgear and trying to get the power back on immediately. :smile:

 

[wirenut1980]

The utility is required to supply voltage in a certain range from nominal. It would be a good idea to check with them and see what is the upper range of voltage supply. For my company the upper range on a 480 V supply is 504 Volts. If they are exceeding their upper range of acceptable voltage, they probably need to at least make some plans for correcting it. You might ask them to help monitor voltage at the point of common coupling and try to test voltage while the factory is running and also when it is down. No fancy monitor is needed for this...a simple voltage RMS monitor will do.