By David Windley, P.Eng.
Determining the cause of recurring board failures on a new piece of equipment proved to be quite a challenge in a medium-sized metal stamping plant. After the discussion of equipment quality and warranties, the power system was monitored for anything which might be contributing to the problem. A very fast, recurring transient was uncovered using high speed capture monitoring equipment and eventually traced back to a completely unrelated piece of equipment. This case study will show that power quality problems may be present without even showing significant short term effects. Care taken in power distribution installation methods and periodic monitoring can minimize future problems.
DISCOVERING THE PROBLEM
Situation
The plant was experiencing board failures on two press material positioning systems. These positioning systems were new and had been recently installed. The plant suspected that power quality might be a factor in these failures and requested a power quality audit be performed. The plant electrical load was divided between two main bus ducts fed from a newly installed substation. Each of the presses in question, A and B, was installed on a separate bus duct.
Monitoring Procedure
As the most recent and frequent failures occurred on Press B, the monitoring equipment was connected to the 'B' feeder duct disconnect switch in the plant substation.
A power monitoring system was used and set to record data for 7 days. This device was selected because it has very fast event capture capabilities and can collect complete information regarding harmonics and other system characteristics.
A visual check of equipment was performed to determine if any installation problems were evident. The wiring and grounding methods on 'A' and 'B' bus ducts were inspected.
Specific machine monitoring was performed on Press B Controls, Press B Positioning Controls, and another Press C connected to the 'B' duct. It was suggested that this third press, as it was old and had experienced some failures, may be the culprit.
Observations
General Equipment and Wiring
The wiring to 'A' feeder duct had been installed using two parallel cable runs in conduits, with the conduits serving as the ground path. The wiring to 'B' feeder duct was installed in a similar manner to 'A' except that one of the parallel conduits carried a ground conductor although it was connected to the switchgear frame rather than to the ground bus.
There were four loads connected to the 'B' Feeder Duct:
1. Press B Controls
2. Press B Positioning Controls
3. Press C Controls
4. Maintenance Shop
Items 1 and 3 have 600:480 volt auto transformers to step down the incoming voltage.
Item 2 has a 600:208/120 volt distribution transformer.
Monitoring Results
Voltage Transients
Four types of voltage transients were observed:
a) Type A: Large magnitude, short duration. These appeared as spikes on the voltage waveform. Magnitudes were as much as 150 volts but lasted only 6-10 microseconds.
b) Type B: Large magnitude, long duration. On Sunday morning, a large sag occurred at approximately 6:30 am.
c) Type c: Small magnitude, medium duration. At points where there was a large current inrush (e.g. where a motor started), a small voltage sag occurred.
Total Harmonic Distortion
Total Harmonic Distortion (or THD) is a measure of the total harmonics generated by the load being measured. It is representative of the distortion observed in the sinusoidal voltage and current waveforms and is expressed as a percentage of the fundamental.
In the case of 'B' feeder duct, significant current THD (up to 50 per cent) was observed when the Press B was running. The voltage THD was normal at less then 2.5 per cent.
Specific Machine Monitoring
Press B Controls: High levels of current harmonics were observed when the press was running. A significant number of Type A voltage transients were observed.
Press B Positioning Controls: High levels of current harmonics were observed. Current imbalance and leading power factor situations were observed. A reduced number of Type A transients were observed.
Press C Controls: Large current inrushes and the resulting Type D transients were observed. Many Type A transients were also observed but didn't appear to correlate with the running of this press.
Conclusion
Voltage Transients
The Type B voltage transient could be attributed to an incoming effect due to the utility experiencing a voltage sag. There was significant lightning activity at the time. The current inrushes which created the Type D transients were due to motors starting in the plant and are explainable and normal.
The short duration Type A and medium duration Type C transients were unusual and were being generated by a piece of equipment in the plant. Although short in duration the Type A transients were large in magnitude and were likely to cause cumulative damage in electronic equipment within the plant. However, the specific machine monitoring seemed to indicate that another piece of equipment -- which was turned off during the second and third shifts -- was responsible.
During the specific monitoring on the Press B Positioning Controls, the Type A transients were getting through the distribution transformer and being applied directly to the Press B positioning control equipment. It is likely that this was directly contributing to the board failures.
It was also noted during the study that the ground fault relay on the plant main switchgear oscillated between a reading of 0.00, 0.12 and 0.14 indicating a fluctuating ground fault on the system of up to 28 amperes. It is possible that this was related to these transients.
Harmonics
Significantly high harmonics were being generated by the positioning controls. Also, the input filtering on this equipment is creating a leading power factor situation. These are not healthy conditions and may be creating some problems themselves within the positioning controls.
The low level of voltage distortion was an indication that the system is healthy and that severe problems should not occur in unrelated equipment elsewhere in the plant.
Equipment and Wiring
Grounding practices were inadequate for sensitive electronic equipment. Auto transformers had been used in lieu of isolation transformers to step down the voltage to the electronic equipment. This condition, while not causing the situation, accentuated any system transient problems.
TRACKING DOWN THE PROBLEM
Situation
Severe voltage transients on the power system were discovered during previous monitoring. As these transients could have damaging effects to the Positioning Controls and other sensitive equipment, the source needed to be located and corrected.
Monitoring Procedure
The power monitoring system was connected to the Bus Duct 'A' disconnect at the plant substation.
The occurrence of transients was monitored and compared against operation of equipment in the plant.
Observations
There are seven loads connected to the 'A' Feeder Duct:
1) Press A Controls
2) Press A Positioning Controls
3) Press D Controls
4) Press E Controls
5) Press F Controls
6) Press G Controls
7) Press H Controls
After approximately five days of monitoring, the transients were seen to occur during normal plant production times, with very little activity during times when machinery was not running. The Press A runs quite often and the transients were not observed when it was running during the off shifts.
After reviewing various possibilities with production staff, short monitoring studies were performed to verify equipment such as a surface grinder in the tool room, the welding equipment and some small presses in the back. No conclusive results were obtained.
On Wednesday during short time monitoring, the transients stopped abruptly at break time at 1400 hours and started again after break. Because of the timing, it was felt that the Press E might be contributing. The whole front line of presses was kept off for 5 minutes. During the five minutes rest, no transients were recorded. When Press E was started, transient activity started immediately. However, the transients occurring were of the type 'C' variety, already noted in the previous power quality study. The severe transients had not re-occurred.
The Press D was started and stopped to see if it contributed. A direct correlation was found. Additional short term trials were performed to eliminate the press auxiliary equipment as sources.
The source of the severe transients appeared to be the Press D Controls. The next day the controls were checked and connections tightened. The press was run in continuous mode and the transients were still observed. The source was further narrowed to the press DC drive controls. An oscilloscope was connected to the output of the DC drive during idling conditions to look at the output waveform. It was observed that one of the SCR's was misfiring under low load conditions. Spikes were observed under heavy load. It was noted that the voltage reduction on this press was performed with an auto-transformer.
Conclusions
The source of severe transients was traced to Press D Controls. The DC drive was found to be mis-operating in that only two of the three phase SCR's were firing correctly.
Under heavily loaded conditions (such as the press stroke) the circuitry associated with the malfunctioning SCR was breaking down and creating the severe voltage transient.
Press E was responsible for the Type C transients. Because of the characteristics, this is likely due to effects related to the ground faults observed.
Recommendations
a) Replace auto transformers with isolation transformers.
b) Ensure that all electronic systems are solidly grounded to less than 1 ohm. Install separate ground conductors in conduits.
c) Investigate reason for ground fault relay display oscillations.
d) Repair or replace the DC drive on Press D.
Points for Consideration
When it comes to protecting your facility from power quality problems, the following points should be taken into consideration when designing, specifying and maintaining equipment.
Design
a) Proper grounding for sensitive electronic equipment (IEEE STD 142, 1100)
b) Isolation of harmonic loads
c) Location in power system
Equipment Specification
a) Quality and reliability
b) Ability to accept plant power conditions
c) Effect on power system
Maintenance
a) Effective grounding
b) Immediate repair
c) Predictive maintenance: study equipment signatures before problems occur and regularly perform power system audits.
David Windley is with Wintek Engineering. He can be reached at wintek@wintek-eng.com.