DSP MKII, system failure??

[sparky321]

I am trying to resolve a problem with and FPE switchgear system. The system is a high resistance grounded 3 phase delta Y. 600 / 347 VAC, the resister is sized to limit the ground current to 5 amps. We have a DSP MKII ground fault protection system that has worked great for 15 years.

For the last while the system will not indicate a ground fault, the transformer, resistor, wiring and grounding seem to be in order.

If there is a ground in system the current to ground will reach 5 amps but this does not return to the transformer via the resistor, checking the current at the resistor and it,s .3 to .5 amps. Also the normal voltage shift is not there, Eg a fault on A should have B and C at 600 volts to ground, they are still at 347 Volts. As the DSP MKII sees only the current and not the volage change it does not alarm. If there is a second fault 2 feeders both shutdown.

 

[waross]

My first thought would be to check the system to see if you have a grounded neutral bypassing the grounding resistor.
However, To limit the current to ground to 5 amps at 347 volts requires about 70 ohms. If your current to ground is 5 amps, and the resistor only passes .5 amps, that means that;
1> The total resistance to ground is still 70 ohms.
2> The values of the accidental ground path would have to be 77.1 ohms, and the value of the normal resistance has accidentally changed to 694 ohms.
We still have a problem with the voltage balance.
Also, we tend to assume that the readings are accurate as stated, but how are you measuring the current through the resistor?
When you add up the coincidences, it seems more likely that you have a problem internally in your ground protection relay or in the wiring to it. I would check the field wiring for a frayed connection with only a few strands carrying current or a wiring change.
I would suggest ensuring that the system is free from grounds and then intentionally grounding one phase. You may want to use a 10 or 20 Horsepower motor winding with a 6 or 7 amp fuse to limit the current in the event of a second accidental ground while you are testing. Use the same ammeter to measure the current through your applied ground and the through the grounding resistor.
If you have a wide discrepancy then look for a medium resistance connection from ground to neutral.
If you have 347 connected loads, examine them for bad connections. If a metalic discharge lamp ballast develops a ground near the high side of the winding and either trips the breaker or burns free but in either case leaves the winding still connected to ground, you now have an accidental high impedance from the neutral to ground through the ballast winding.
Good luck. Let us know what you find.
yours

 

[sparky321]

The 69 ohm ground resistor was replaced, The transformer was tested, including draining oil to inspect the condition of the insulators on the 3 phases and the neutral.
The Current readings were done through at the ground resistor with a true RMS clamp on meter, this meter was also used on the ground that was inserted to measure the 4.5 amps to ground. 2 other systems also measured this current at about 4.5 amps. These systems are the FPE CT and meter on the grounded feeder and A high end monitoring system we use for Harmonics, Power Factor, and trending of the entire system. I have no reason to not trust the readings as the people, meters etc. are top notch.
We have a spare ground fault resistor that we use for testing, this resistor gives us the same results as the one ground that I found that did not trip 2 feeders.

Note: The system seems to work fine with most everything shut down, (so I hear) although I was not present at that time.

One thing that seems to be common is that one particular feeder always is one of the two to trip. WW1 has tripped with 3 other feeders at different times over the past year.

My next steps will include inserted a 69 ohm ground and powering things down until the system starts to work. The only thing that makes sense to me is that somehow there is a solid ground that is of less resistance than the ground resistor and also provides a path back to the main transformer. One thought that I have had is that if say a large Y connected motor or transformer had a solid ground at or very close to the Y point could this explain events I have listed if a second Phase to ground occurred on another main feeder and tripped the 2 feeders.

Most of the time it is not possible to shutdown the majority of the equipment as we run 24 / 7. At Easter I will have 2 days to work this out so I want to be prepared. Any ideas will help.

Thanks Again

 

[waross]

Do you have a neutral carried out to the panels and 347 volt loads?
I am trying to visualize your system and your test connections.

If there is a ground in system the current to ground will reach 5 amps but this does not return to the transformer via the resistor, checking the current at the resistor and it,s .3 to .5 amps. Also the normal voltage shift is not there, Eg a fault on A should have B and C at 600 volts to ground, they are still at 347 Volts. As the DSP MKII sees only the current and not the volage change it does not alarm. If there is a second fault 2 feeders both shutdown.

sparky321, in the passage that I have quoted, do you mean a solid ground, or a test ground with a 69 ohm resistor?
If you are getting 5 amps or 4.5 amps with a 69 ohm test resistor and .5 amps through the service ground resistor then you very likely have a grounded neutral. Were any new panels installed on feeder WW1? I would check to see if you have a panel with the bonding screw installed. That would be a good candidate to cause your problems.
How about a "thumper" test? You can build a thumper with a contactor and a couple of timers or you can use a switch and an electrician.
The idea of the thumper is to inject a pulsing current into the system that you can trace with an ammeter.
Connect your test resistor to ground and to the thumper. Connect the other side of the thumper to a hot line.
Set the timers or instruct the electrician to close the circuit for 1/2 to 1 second and then open the circuit for 1/2 to one second. As long as the pulses are reasonably regular and the tester knows what timing to expect. I understand that your test resistor draws about 5 amps. You should be able to see 5 amp pulses even against a background neutral current of 100 or 200 amps or more. If you can safely meter the feeder neutrals, you should be able to find the feeder that has the neutral problem, and then follow the pulses to the source of the problem.
Let me know what you think and what you find if you do any testing.

If you don't find the current returning on the neutral, you have to look for it returning on the ungrounded feeders.
I have no idea what your feeder current is, and your "thumper" current will probably divide between the other two phases. You may have to increase your ground current to get a reading.
You may also have better luck "thumping" at individual panels and/or switches where the line current is less and the pulses are easier to see.
Yes it's possible that a motor or transformer has a fault near the neutral point and is acting as a grounding transformer. for the system. If you can see the "thumper" pulses against the normal current you should be able to follow them to the problem.
good luck
respectfully

 

[sparky321]

"Do you have a neutral carried out to the panels and 347 volt loads?"
No there is no neutral, the main transformer is a 600 volt “Y” secondary, the neutral is wired directly into a 69 ohm resistor this in turn is wired directly to the ground grid.
- All of the 347 loads are on isolation transformers, so only the primary (delta side is on the ground detection portion, the secondary sides are Y with a ground and a neutral. This is used for all of the lighting and office type loads.
- There are about 10 MCC centers with mostly 600 volt motors, transformers and controls, these loads are all wired as delta. Also there is a power factor correct unit RTRC (Real time reactive compensation Model AV9000 from Schneider Electric)
- The MCC’s feed motors, Frequency Drives, transformers etc. The wrapping room has a larger number of robot, servo drives etc. (high speed packaging) Most of this equipment is USA standards (480) so this load is mostly isolation transformers.
- Everything in the plant on the ground fault monitored 600 volts is done in a 3 wire format, there is no neutral wire, system ground wiring is used as a return path in the event of a ground.

"I am trying to visualize your system and your test connections."
From the rest of your ideas I feel that you have the right idea of how the system is setup.

"Quote:
If there is a ground in system the current to ground will reach 5 amps but this does not return to the transformer via the resistor, checking the current at the resistor and it,s .3 to .5 amps. Also the normal voltage shift is not there, Eg a fault on A should have B and C at 600 volts to ground, they are still at 347 Volts. As the DSP MKII sees only the current and not the voltage change it does not alarm. If there is a second fault 2 feeders both shutdown.
By this I meant a test ground of 69 ohm’s, with this in series with the ground resistor I should have seen 2.5 amps on both resistors."

The rest of your comments about having a grounded neutral point are the same as mine. Some of the equipment may have Y configurations (eg. transformers) and one of these may have been grounded at the neutral point. I am looking into this possibility, some new equipment was installed including the RTRC I mentioned. The drawings we have are mostly one line and leave much to the imagination so will require the hands on approach.

I would like to use the thumper idea, this might make it easier to locate the problem. I am familiar with using thumpers, never made one before though. Hopefully this will work on the phase wires as we do not have a neutral wire. The main feeders have 100 to 300 amps load during normal operation, most individual loads are 30 amps or less. The MCC that I suspect having a ground is for wrapping, again this is high speed packaging and mostly 480 volt so checking the main feeds to each piece should be fairly easy at each bucket.

The plant has been numerous times with this problem including a couple of times when someone was doing testing (not by me), that being said I feel that any future testing will not be allowed unless the plant is not running.

Jim

 

[waross]

Hello sparky321
It may be well to disable the RTRC during testing. Alternatively, you may want to simulate a ground with the RTRC connected and then disconnected. I don't think that it will be a problem, but the schematic shows a neutral connection. I am more worried about complications that may arise if the RTRC is monitoring voltage phase to neutral rather than phase to phase.
Manual;

http://www.schneider-electric.ca/

http://www/en/products/pqc/AV9000_Instr_Bulletin.pdf

I would survey the transformers prior to the shut-down. The delta primary is prefered as it distributes unbalanced loads over two phases rather than one phase.
If you have any wye primary transformers check that the primary wye point has not been grounded.
Particularly bad is a wye delta with the wye point grounded. An electrician may think that he is doing a good job by grounding this point when he is installing the transformer. Not so. The grounded wye/delta configuration will draw excessive current with either a missing phase or a grounded phase. It will try to balance the primary voltages. With a missing phase, the grounded wye/delta bank will try to supply enough energy to replace the missing phase. If the transformer bank is large in relation to the system, it may succeed. With a ground the transformer (or bank) will still try to pass enough current to balance the voltages.

Summary, check your transformers for wye primary windings. If you have any transformers with wye primary windings, check for grounded wye points.
If you find a grounded wye point, you have found the problem.
Delta primary transformers. Possibly a problem but not a high priority. A ground on a delta primary winding should display a shift in voltage to ground. The best case of a grounded delta winding should show ground voltages of 520V, 300V and 300V.
Motors. This is a strong possibility.
I'm out of time this morning.
I'll add more later.
What part of the world are you in?
respectfully

 

[sparky321]

Thanks again for your input, just to clarify who I am. I am an Electrican with 25 years experience, working in central Ontario. Several unsucessful attempts have been made to find this problem, I have been tasked with finding the solution.

I agree with this approach to finding the solution, with only a 2 day window until likely next Oct. I want to be able to do as much as possible on Easter weekend.

The RTRC has already been identified as a possible contributer to this problem and was on my list of items to check, this is why I mentioned that we had this unit.
Most of the engineers and electrcians on site are in the dark pertaining to the Ground Fault system and how it works. So anything is possible, grounded neutrals etc! as a peon I struggle to get management to allow time and resourses to keep everyone up to speed on how this system functions.

A number of the machines here are refurbished equipment, some with 480 volt as the norm so without checking I can't say if these are delta or wye configuration. Documentation is poor at best on the 600 volt side.

What part of the world are you in?

Thanks again
Jim

 

[sparky321]

Thanks for the help Waross,

We checked the main 600 VAC systems out using the ground resistor and found 8 seperate auto transformers that someone had added a Y point ground. These grounds were removed and the DSP MKII works perfectly.

Thanks again for your support.

Jim

 

[waross]

It's good to know that I was able to help.
I've been in Central America for quite some time, but am heading back to Vancouver next week.
Are the autotransformers used to convert 600 volts to 480 and vice versa? I have had good luck many times using two small dry-type transformers in open delta to adjust voltages. We used to service a plant that had both 480 volt and 600 volt systems. Also 460 volt and 575 volt motors. They would move a machine or change a machine and we would have the wrong voltage. We had pairs of dry type transformers all over the plant. As long as the transformer secondary has a current rating equal to or a little more than the motor current it is big enough for open delta for the motor.
respectfully

 

[sparky321]

This is a great time of the year to head back to Vancouver, I lived in Williams Lake for 7 years and the spring is great in BC. I hope to semi retire to British Columbia ar Alberta in 5 years or so. When I retire I am hoping to use my Electrical experience to work in interesting places on a contract basis.

These transformers are used to convert 575 volt to 480 volt, the end use is for standard DKC servo drives, we have positional control (CNC), speed control and some with 2 step positional control. None of our equipment is moved around as a norm so flexibility in the power distribution is not important.

Thanks Jim