Electrical current leaks on 4160 volt equipment 1

[h20dr]

Can any one give the test procedure for testing for current leaks. This is 4160 voltage and may be causing corrosion in chiller condenser. Each condenser is grounded on each end.

h20dr

 

[waross]

Can you give a little more information please?
yours

 

[h20dr]

waross,

Thanks for asking. I have been in industrial water treatment 26 years and have never seen pitting corrosion so rapid and with severity in which I have encountered. I have elemined all the possibilities except for a possible electrical current leak from a chiller motor, pump motor or fan motor and the current is possibly going out to ground through copper condensers with about 1200 gpm of water moving. These are big industrial air conditioners that use water for a condening medium.

I am not electrical, but rather chemical. Can you help with basics of how to test for a current leak on 41 60 volt compressor motors?

 

[rbyrne15]

Would an appropriate test involve testing the motor's insulation resistance, from phase to ground, at 5 kV dc? This, I would think, would check the quality of the insulation which, if the resistance is too low, would cause a high resistance ground fault not picked up by the overcurrent protection.

Please correct me if I'm wrong.

Thanks,

Richard.

 

[h20dr]

The voltage may not have been clear. 4160 Volts AC compressor motor.

Dan

 

[rbyrne15]

That's what I figured h20dr.

When maintaining electrical distriubtion equipment and motors, the insulation resistance is tested (although not always), using a dc test voltage above the ac rating of the equipment under test, in order to stress the insulation temporarily.
This test voltage causes a current (again, dc in this case) to flow between the two parts under test - such as from a motor winding to ground, through the insulation. This path should ideally provide an infinite resistance to current - we don't want any current leaking directly to ground under normal operating conditions. However in practice, the materials used to insulate electrical conductors from ground are made up of atoms containing electrons (as you well know), which form the leakage current when the device is being tested. The larger this leakage current, the lower the insulation resistance, and the lower the integrity of the insulation.

 

[h20dr]

rbyrne15,

Thanks for your expertise. Will discuss this with the paper mill electricans and electrical engineers. They did not know how to perform this test.

Thank you again.

Dan

 

[rbyrne15]

You're welcome.

I would prefer that someone verify my recommendations before you approach your electrical staff, just to make sure I'm giving you the right information about which test to carry out.

After reading a section of my book "Operating, Testing, and Preventive Maintenance of Electrical Power Aparatus", by Charles H. Hubert, I would also recommend that your engineers look into "controlled-dc overvoltage test (step-voltage test) of insulation for rotating machines" - section 25.4 of the said text.

I have only seen this test performed once. The nature of the test is to apply a pre-determined voltage to the device under test, wait and record leakage current at regular time intervals, then begin to raise the applied voltage by equal amounts every minute up to the maximum allowable test voltage. This test is more likely to show you the nature of the possibly-failing insulation on your "leaky" machine, compared to the test I mentioned in my first post.

Once again, have the electricians/engineers investigate the methods of testing to ensure safety and the avoidance of unecessary damage to equipment.

Richard Byrne, EET.

 

[rbyrne15]

Another thought - could you describe the condensers? You mentioned in you first post that each condenser is grounded at each end - does this mean that each condesr has two ground connections?

Just interested...

Richard Byrne, E.E.T.

 

[electricpete]

I wouldn't do any electrical testing unless you have someone reasonably trained.

IEEE43-2000 specifies routine insulation resistance test of 4kv motors at 2500vdc or less as I remember. 5000vdc would be a dc hi-pot or step voltage test per IEEE95.

Insulation resistance test at 2500vds is a harmless and nondestructive test (but again need trained people).

Hi-pot test is a useful test but also potentially destructive. It requires careful consideration of the consequences of failure during the test. We would not perform dc hi-pot on critical motor without verifying the availability of a spare first.

I have a hard time believing that insulation resistance problem on 4kv system is causing your problem. I would spend my time looking carefully at the chemistry and operating conditions myself.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[ScottyUK]

Any leakage wfrom the motor will be AC - corrosion problems typically need a DC current source or an electrochemical cell. Such cells can be formed between structure and earth, dissimilar metals, etc. I think you're looking in the wrong place, but without more idea of the plant and surrounding structures it is hard to be more specific.

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[waross]

hello h20dr
I'm still waiting for a description of the system.
A.C. corrosion. I have seen rapid corrosion of sheet piling subjected to A.C. welding currents.
Insulation resistance vs ground currents. There is a relation but not a direct correlation.
The best way to measure ground current is to identify the ground path and measure the current in the ground path with a meter.
Where is the corrosion taking place. Is your water conductive? Are you using cooling towers or deep wells or another method of rejecting the heat of condensation. Is your piping conductive and does it have any non-conducting connections or couplings?
I am assuming that your 4160 motors, compressors, and condensors are mounted on a common skid. If not, please advise.
There would be a metalic connection between the motor, the condensor, the compressor and the skid. Any leakage would pass through the metal and return on one of the grounds you mentioned. However, if a pump or fan motor at the cooling tower or a well pump motor was poorly or improperly grounded and was starting to fail this could cause current flow in the water.
PLEASE NOTE: For electrical currents to cause corrosion, the current must be flowing in the water, not near the water.
Is the 4160 volt system delta or wye? Is it solidly grounded, high resistance grounded, low resistance grounded, Zig-zag grounded, directly grounded etc.?
More importantly, what is your pump and fan voltage and what type of system and how is the system grounded. This will be a more likely cause of corrosive ground currents.
What is the condition of your zincs? If your zincs have eroded or are fouled you can expect rapid corrosion.
yours

 

[h20dr]

This chiller plant has 3 X 500 Ton Trane Centrifugal chillers with 4160 vac compressors. Each end of the chiller condensers, compressor motors, pump motors and fan motors are grounded with straps to a common ground grid. The pump and fan motors for cooling towers are 480 VAC.
I have severe pitting type corrosion in the carbon steel piping and copper condensing tubes. Extensive work on the chemistry has been done. Anarobic bacteria, hydrogen sulfide & chlorine together, manganese pitting, LSI +2.2, pH 8.8 have all been controlled, but was once a problem.

 

[mpparent]

You could also check the NETA standard for Maintenance and Testing.

Mike

 

[Laplacian]

Does any of this piping go underground? Do you have cathodic protection rectifiers installed on your piping system?

 

[waross]

Hello h20dr
Is the pitting in just the condensor circuit. I understand it is, but please confirm this.
Has there been any change in the make-up water?
Can you ask your electricians what the grounding is on the 480 volt system.
How is the water tower grounded? eg; Feeder conduit used as a ground conductor, direct cable connection to the same grid as the chiller skids, etc?
Has there been any changes in the plant recently? Not nesessarily related to the chillers?
Do you use zincs, if not have you considered using them.
The sheet piling problem I mentioned earlier is held in check with zincs. When the zincs are completely sacrificed the sheet piling will corrode rapidly.
To check for ground current on the 4160 volt system, determinme the type of grounding.
For a high resistance grounded system, measure the voltage across the grounding resistor. beware of possible high voltages. There may be over 2000 volts present.
For a low resistance grounded system, the test is the same but the voltage will be much less.
For a solidly grounded system, find the ground connection at the 4160 V transformer. Open the connection, WITH THE CURRENT OFF, and connect a transformer in place of the grounding jumper. The primary rating should be at least 2300 volts. connect one or more lamps across the secondary of the transformer. The combined voltage rating of the lamps should be at least 50% of the expected maximum voltage. Measure the voltage across the lamps.
Motor testing. You can megger your motors to ground with a 1000 volt megger. Both your 4160 V motors and your 480 V motors.
You may decide to hold off on the ground current testing of the 4160 V motors until last if they are solidly grounded.
Megger everything, 480 V motors and 4160 volt motors and report back.
Please report back on the grounding system and transformer connections on the 480 V system.
respectfully

 

[Laplacian]

waross,

By "zincs", do you mean sacrificial anodes?

 

[waross]

Hello Laplacian
Yes. I see them most often in a marine environment where they are referred to as zincs. Probably because when they are new the word ZINC is cast into them.
respectfully

 

[h20dr]

mpparent, Laplacian, waross,

Thanks for some of your gray matter.

Mpparent,

NETA - Looking for this now.

Laplacian,

None of the piping is underground. All overhead.

waross,

The pitting is occuring in both the copper condenser tubes and carbon steel piping. (1,000 feet total of 6" and 18" carbon steel) The towers have a Balticoat coating so they are not affected.

Regarding the water change, this is a paper mill that uses massive quanitities of water from 6 wells. I have examined by analyses all wells. The only marginal problem chemistry wise is the hydrogen sulfide and chlorine. Hydrogen sulfide can reduce the chlorine and produce sulfuric acid. Last week I discontinued the chlorine for a trial. Currently I am monitoring carbon steel and copper corrosion rates with a Corrador and test corrosion coupons.

I will have to take your questions and recommendations to the plant next Wed. Regarding any changes in the plant. Maintenance records are being searched for changes.

Zinc anodes are only effective in the general area of the anodes and not the entire system. I have an entire system effect here.

Thanx again. H20dr

 

[waross]

Hello H20dr;
My expertise is electrical. I am happy to suggest electrical tests and remedies, however the problem here may be chemical.
The normal mechanism for electrical related corrosion is a flow of current from or to the material being corroded.
I tend to discount the 4160 volt motors as a source of corrosion current.
There are several reasons for this;
One reason is the ground path. The expected current path is from the motor windings to the motor frame, from the motor frame to the mounting skid, and from the skid through the ground cables back to the supply. This should be a low impedance (resistance) path. If there was a situation where the ground connection from the skid to the current supply was compromised, and the equipment at the cooling towers was well grounded with a ground impedance to the source that was significant in relation to the ground impedance at the compressor skid, you could get currents passing into and out of the water. You would expect corrosion at both locations. That is, if the current was passing from the condenser to the water tower, you may experience corrosion at the water tower also.
Ground current from the 4160 V motors implies an internal fault in one or more of the motors. Such a fault could be expected to progress to total failure fairly quickly.

I would estimate your 4160 motor current to be about 60 amps. That's about the same current as a 50 hp. motor at 480 volts. Don't let the size and voltage of the big motors over-impress you. All motors are suspect.

Things to look for;
On both your 4160 volt system and your 480 volt system, there should be one connection from the transformer neutral to ground. This will be the place to check for ground leakage.
CAUTION OPENING A GROUND CONNECTION WITH A PARTIAL GROUND FAULT ON THE SYSTEM IS EXTREMELY HAZARDOUS. AS IN "WE FOUND THE PROBLEM BUT THE CREW IS DEAD."
You may measure the current in the grounding connection with a clamp meter, or carefully measure voltage across a grounding resistor to infer the passage of ground current. Be aware that there may be a potentially lethal voltage across the resistor in the event of ground currents.
The same precautions should be used as would be observed checking the full system voltage.
Measuring the voltage across the grounding resistor is hazardous but can be done with adequate precautions.
Opening the ground connection must only be done with the current off and the area flagged of and identified as a lethal hazard area while testing is being done with the ground connection compromised.

Some other things to check. Pipe connections;
Some types of couplings can not be depended upon to maintain electrical continuity and must be bypassed with a jumper if electrical continuity is important. Some connections are designed to interupt the electrical continuity of the piping.
I suggest checking to see if any bonding jumpers have inadvertantly been removed from any couplings, and more importantly, have any couplings or connections that are intended to be insulated been inadvertantly jumpered. With steel piping and copper heat exchanger tubing there may be electrically insulated piping connections and/or heat exchanger supports that have been compromised.
respectfully