Introduction
The office building located at 1 Shore Circle, Sparkle Bay, CA has experienced a number of main circuit breakers operations (trips) whenever either of the two roof top chillers is turned-on. PowerCET Corporation was contacted to assists in determining if the breaker tripping was the result of equipment (chiller) problems, electrical wiring issues or circuit breaker problems.
Methodology
A Dranetz-BMI PP4300 equipped with a Multi-DAQ Taskcard and 3000A flexible current probes was connected to the load side of the main 2000A breaker located in the basement area adjacent to the elevators. (Note: The voltage probes were connected to the line (utility) side of the breaker to maintain synchronization for the power monitor.)
The PP4300 was programmed as follows:
Multi-DAQ Instrument configuration
---------------------------------------------------
Site/Filename = 1 Shore Circle
Measured From = 06/14/2004 10:30:25
Measured To = 06/14/2004 10:41:32
Configuration = 4 WIRE / 3 PROBE
Monitoring Type = STANDARD PQ
Nominal Voltage = 277.0 V
Nominal Current = 146.3 A
Nominal Frequency = 60.0 Hz
Trigger Response Setups
---------------------------------------------------
Pre-trigger length = 9 cycles
Post-trigger length = 60 cycles
On Voltage Trigg = Record all active
voltage and current channels
On Current Trigg = Record all active
voltage and current channels + D
After recording = REARM
Limit Setups
---------------------------------------------------
Voltages AV BV CV DV
High: 304.7 304.7 304.7 0.0
Low: 249.3 249.3 249.3 0.0
Wave: 19.4 19.4 19.4 0.0
Crest: 470.9 470.9 470.9 0.0
---------------------------------------------------
Currents AI BI CI DI
High: 500.0 500.0 500.0 100.0
Low: 0.0 0.0 0.0 0.0
Wave: 400.0 400.0 400.0 100.0
Crest: 0.0 0.0 0.0 0.0
Two separate tests were performed, one with each chiller being cycled.
Key Findings
Test #1
Figure 1 is the voltage and current waveform associated with the initial test—amperage scale in Figure 1 represents peak current levels and not RMS. The current waveforms document the breaker operation approximately 50 milli-seconds after turn-on of the chiller. Figure 2 is the time plot data showing a maximum RMS current of 860A on phase B.
Test #2
Figures 3 and 4 show the waveform and time plot data from the second chiller test. Note in this case there is consistent inrush on all 3-phases. The maximum current for the second test was approximately 936A RMS on phase C.
Figures 5 shows the voltage and current waveform envelop as the main breaker was reset after the initial chiller test. There is a second current inrush that occurs approximately 1-second after the reset of the breaker. This most likely some of the building’s equipment with a time delay after restoral of power.
Photo 3 shows the adjustment controls for the main 2000A breaker. The Ground Fault Trip is set for a 0.3-second (300ms, 18-cycle) delay with a Pick-up Amperes setting of 800A. The cover panel sticker in Photo 4 shows that the Pick-up Amperes setting was set to 400A on 7/23/2002.
The Short Time Trip level, Photo 3, is set at 8. To obtain the actual ampere rating the continuous Amp Rating (probably 80% of the 2000A) is multiplied by the setting. The setting of 8 most likely corresponds to an ampere setting of 12,800A. (Note: Detailed specifications of the breaker were not available for this report.)
Recommendations
1. The chiller cycled during the initial test should be tested and inspected. Current waveforms indicate that there may be a single-phasing condition—refer to Figure 1—as no inrush current was recorded on phase A. The chiller cycled during the second test appears to be normal—refer to Figure 3.
2. Current levels recorded during the two chiller tests should not be sufficient to trip the main circuit breaker. The existing breaker should be tested (calibrated) and replaced if it can not be adjusted (repaired) in the field.
3. It appears that the Pick-up Amperes setting for the ground fault section of the breaker was adjusted sometime after testing completed on 7/23/2002. Ground fault settings should be verified and the circuit breaker readjusted to the correct setting.
4. A battery powered emergency light should be installed in the main power room.
Summary
The nominal load on the building during the tests, with the chillers off, was approximately 170A per phase which seems quite low. The amperage recorded at the end of the initial test, Figures 5 and 6 show a much larger current in the 400 to 500 ampere range. There is only 2-sconds of data and it may simply be inrush associated with equipment in the building restarting.
[Note: The 2000A breaker was replaced 2-days later and there have been no other reported problems.]
Epilogue
I was hired to figure out if the breaker or load was at fault...clearly the breaker should not have tripped at the recorded current levels. I also happened to notice during the testing that the one chiller unit was single phasing.
One of the problems with being a consultant is that once we identify the problem they get us out quite quickly...kink of like a taxi cab ... when your are in the cab the meter is running.
What we think happened with the older main breaker was that the trip module was either damaged (stressed) when the chiller initially failed or the resistor values changed (aging)...we find this with older breakers. By the way...we are probably talking about a breaker close to 30 years old.
Report prepared by:
Bruce Lonie
President, PowerCET Corporation
PowerCET Corporation (www.powercet.com) is an independent consulting company specializing in power quality and electromagnetic environment.
