A Word About Transformer Oil Color

A Word About Color

I need to add a few words about color here. Color is pretty much just an observation. It becomes more relevant when you compare it from previous years. Oil that has been heated during a reclamation process will usually show up at about a 2.0. The color will be just a bit off from the disk on the colorimeter because it’s been “roasted”. The acid and ift numbers will show up fine on these. If over a few years it starts to drop rapidly then you know that the oil reclamation process did not do a good job of cleaning out the sludge. Probably a result of too little exposure to the filtering media and hot oil. It could also be due to an incomplete draining during an oil change out. Again future tests will reveal the process.
I have seen some oils in that were new and they were at a 4.0 to 4.5 on the colorimeter. I remember the first time I ran across these and was mystified as to why the oil test results were showing up as good. It was explained to me that that was an oil produced for Westinghouse corporation and it was used in a lot of units. You can identify it first by the test results and second by the bluish cast that it has when held up in sunlight. Anytime you get an oil with a 4.0 or better color you still had better sniff it to assure no arcing or overheating.

For those of you that are visiting my blog I want to thank you and make an offer that you can’t refuse. I am currently working on a program that I can post on the blog site that you can use to evaluate your own oil screen test results oil screen consists of Acid, IFT, Color, Dielectric, Specific Gravity, Visual, and Sniff tests). If you send me your results at my e-mail address I will look at them for you for free and send you back a short evaluation sheet. Just send them to: transformerbob@gmail.com. Allow a little time for me to evaluate because I don’t open up my email every day.
Thanks and Best Regards,

Transformer Bob

“Good breeding consists in concealing how much we think of ourselves and how little we think of the other person.”

MARK TWAIN, Mark Twain's Notebook

Transformer Oil Acidity and Interfacial Tension

Transformer Oil Acidity and Interfacial Tension

Continuing on we have to look at the acidity level and interfacial tension of the oil. These readings should always be compared together. As oil oxidizes over time it begins to create acids within the solution. This process can be greatly accelerated by heat. In normal conditions it takes many years for the acid levels to reach a point where servicing is needed. Most transformer oil has an inhibitor added to slow down the oxidation process thereby extending the life of the equipment. Chemically what happens is that the oil molecules have spaces reserved in them that oxygen atoms will fit into perfectly. The inhibitor molecules fill up those spaces (although not as perfectly as an oxygen atom) until they are forced out at a later date by oxygen atoms. When this happens the acids formed begin to attack the tensile strength of the cellulose materials (wood and paper) in the transformer making them brittle and subject to cracking due to constant vibration. As a byproduct the formation of acids produces a tarry like substance in the oil that we call sludge. The best way to visualize sludge is to think of it as asphalt. It will deposit on the horizontal surfaces of the transformer and harden causing a furthering of the brittleness. So sludge in a transformer is a direct result of the acid formation in oil.
Acidity levels are measured on KOH per gm. KOH is potassium hydroxide and is used to neutralize the acid. The test technician will put enough drops of KOH into a measured mixture of oil and isopropyl alcohol to turn the sample pink and hold. Then he/she calculating the amount of KOH used in a set formula to render a neutralization number. This is a pretty straightforward test. There is some variability in that the technicians’ judgment is employed in determining the correct shade of pink but it’s so simple that most technicians can come up with reliable results easily and there aren’t many things that can throw off the test result. An extremely dark color can also make testing difficult.
The interfacial tension test is a test that must be done by an experienced technician to get valid results. It takes quite a bit of experience to generate reproducible IFT tests. If it’s not something that the tech does on a regular basis you would want them to run it at least three times. Interfacial tension is reported in units of force or dynes per centimeter. The technician must be knowledgeable, have steady hands, use clean equipment and have a clean sample. To better understand how delicate the test is the IFT of clean distilled water should be somewhere around 75 to 80 dynes per centimeter. That is the force that it takes to penetrate the surface of the water. That’s why some insects can “float” on top of the water without getting wet. They aren’t heavy enough to penetrate the surface.
In performing the test a sample of uncontaminated distilled water is poured into a beaker sitting on a platform mounted on the IFT instrument. The water needs to cover a platinum ring suspended on the apparatus that protruded into the beaker. The technician must then carefully pour some of the oil into the beaker slowly so that it sits upon the water. Normally about ½ to ¾ inch of water and about ¼ inch of oil. The platform is then loosened and the ring is carefully placed at the top of the water level just at the interface of the water and oil. The ring is then released from a locking device and it should float motionless in the mixture. The platform under beaker is then slowly lowered using a screw knob by the technician’s left hand. At the same time the technician exerts upward force on the ring by turning another screw knob in his right hand. Care must be taken to assure even pulling in each direction. What will happen is that the ring will get dragged down and resist breaking the surface of the water into the oil. When the ring does break the surface it will jump into the oil. At that point the technician stops and makes note of the reading on the dial of the instrument. This test requires coordination and a very delicate hand. If the readings don’t make sense then it should be repeated for accuracy. There is a plethora of reasons why an IFT test can go wrong.  The ring and beaker must be clean and free of any residue of soap. Soap will break down the interfacial tension of the water to prove this carefully place a needle on a sample of distilled water on a clean beaker. It sits on top of the water and will not sink if carefully placed. Add a drop of soapy water and it will immediately sink to the bottom. The soap has broken down the interfacial tension of the water. The ring must be cleaned and all residue removed before the test. Normally a dip in alcohol will clean it adequately.
Most of the time comparing the two will confirm a correctly performed test.  
If the neutralization number is between 0.00 and 0.10 the IFT should be between 30.0 and 45.0
If the NN is between 0.05 and 0.10 the IFT should be between 27.1 and 29.9
If the NN is between 0.11 and 0.15 the IFT should be between 24.0 and 27.0
If the NN is between 0.16 and 0.40 the IFT should be between 18.0 and 23.9
If the NN is between 0.41 and 0.65 the IFT should be between 14.0 and 17.9
If the NN is between 0.66 and 1.50 the IFT should be between 9.0 and 13.9
If the NN is greater than 1.51 its safe to say you have a problem.

What do you do if they are not in the above ranges? Same as above, we go back to the history of the transformer. If the NN is low and the IFT is low was the oil changed in the transformer? Look at the color was there a lightening from the last test?  If the color hasn’t changed was the oil filtered in some way? Possibly a “cold” oil reclamation process? Did you repeat the tests? How do they compare from the previous data if any?
What if the NN number is too high for the IFT? Check for a marked color change. Has the unit been overheating? If it is overheating is the oil at the proper level or is it low? If it’s low why? Is there a leak?
Make sure the sampling apparatus is clean, uncontaminated and that the technician knows what he is doing when he draws the sample. (Sampling will be discussed in a future post)
Finally confirm that you have the right sample. It’s easy to mix them up.

Specific Gravity is not really a test. It’s merely recording the number and comparing it to past observances. The specific gravity device used to measure most oil samples is not a precise instrument and the number that you get from it is a ballpark. The specific Gravity of water is 1.0. I have seen uncontaminated transformer oil anywhere between 0.865 to 0.890. If the oil is grossly contaminated with PCB (well over 1.0, usually about 1.5) it will be above the 0.890 mark. In many cases you will be able to detect PCB contamination just by sniffing a sample. A few manufacturers dipped the windings in PCB for testing (the test facility was indoors) back in earlier days. Those transformers have high specific gravities accompanied by a strong PCB smell.
Visual test is an observation looking for particulate matter like filter media residue or degrading cellulose material. If the sample has any free water it can be seen at the bottom of the sample. As mentioned in the other tests above if you see something in a sample that is not supposed to be there you have to begin an investigation of 1. What is it? 2. How did it get there. Again most times a second sample should be taken. If it looks like cellulose further more extensive testing should be done. In the case of water if the second sample shows it Karl Fischer tests should be performed to determine how much is in the oil and windings. GE made a great nomograph precisely for that purpose. You can find it in many places on the internet.
The sniff test is just that. The technician has to sniff the sample and determine if there is any unusual odor present. A good nose can detect arcing, overheating, PCB’s or other problems. As with the above tests always put things together and compare all of the other tests when you get an unusual result. It will help lead you to potential problems. Sometimes it will force you to perform other more expensive tests but if you saved the equipment it’s mostly worth it.

I’m working on a spreadsheet that can analyze oil test results for you  to post on my website Transformer Oil for Dummys.  Remember if you have questions email them to me at transformerbob@gmail.com. 

Regards,

Bob

“You can always count on Americans to do the right thing—after they’ve tried everything else.” – Winston Churchill

Field Testing of Oil Filled Transformers Part 1

Field Testing of Oil Filled Transformers

Field tests performed on transformer oil consist of:

Dielectric Test

Specific Gravity*

Color*

Visual*

Acidity

Sniff test*

Interfacial Tension

*Note Specific Gravity, visual, sniff, and color are not really tests, only observations for later comparison.

Dielectric  testing is performed with a hypot. A sample of oil is poured into a cup that has two electrodes set at 1/10 of an inch apart.  A controlled gradually rising current (from 0 to 50Kv) is applied to the electrodes until an arc jumps across the gap. The meter then stops and the breakdown voltage is recorded. The unit is reset for another test.  The technician performs three such tests on a sample then averages the three to come up with a reading.  For most observers any average that comes out to 27Kv or better is sufficient. Any readings lower than that can indicate a problem in the insulating value of the oil. 

When an unacceptable dielectric test occurs the testing technician first has to look at several factors. First, what does a visual observation of the sample reveal? Is there dirt carbon or other contamination suspended in the sample? Look at the bottom of the sample is there any free water present?  Was the sample collected in a clean uncontaminated container? Was the drain valve properly flushed? Was care taken to make sure that the sample was not contaminated by the technician’s hands? Is the hypot working properly? Is the electrode gap set correctly? Was there any contamination left in the cup from a previous sample?  A lot of things can make a dielectric test go south so it’s always best to confirm the readings with another properly drawn sample.

If the second sample renders an unacceptable reading as well then go back to the previous paragraph and recheck all of the items listed.  If the visual shows particles either at the bottom or in suspension then you must try to identify what the particles are.  Carbon in a sample will settle over time and create a black covering over the bottom of the sample container.  When the oil stirs carbon will wisp up looking like smoke in the liquid.  By the way, in case you haven’t guessed your sample should be taken in a dry uncontaminated container that you can see through such as a clear plastic or glass bottle.  If the sample has particles in it try to identify what it is. Anytime you have oil with particles in it I would say it’s not good.  It can be deteriorated paper or wood from the inside of the transformer or it could be media material from a filtration or reclamation process. Most of the time paper or wood from the inside of the unit will not cause a breakdown in the dielectric properties of the oil. They are good insulators, that’s why they are used inside the transformer. Experts may disagree somewhat and speculate that the deterioration of the cellulose material in the oil created a byproduct of moisture in the oil so further testing could be required.

Free water in transformer oil is the easiest to spot. Oil and water don’t readily mix and if you look closely free water will show up as tiny beads in the bottom of the sample container.

Now it’s time for a little research.  First look at the unit. Is it a sealed tank or is it open to the air? If it’s sealed does it have a positive pressure on it or a vacuum? If it has a positive pressure find out if they put a nitrogen blanket on it or something else? If something else then what? If the tank has a vacuum or the pressure/vacuum gauge reads zero check to see if there’s any water sitting on the horizontal surfaces of the tank.  If yes has there been significant temperature fluctuation lately? Could the unit have heated up in the daytime and when it cooled down at night sucked in some water sitting on it? Open or closed has there been any precipitation lately? Has the unit been serviced in some way? What was it serviced for? How was it serviced? What equipment did they use? Was the oil filtered? Was it changed out? Sometimes filter paper can absorb enough moisture from the air to add it to the oil. Thorough research can, in most cases, uncover the reasons for an unacceptable test without having to perform further tests. However keep in mind that further testing needs to be done to confirm what you may have narrowed down as the source of the problem.  A Karl Fischer Test looking for parts per million of moisture should be done. If the Karl Fischer test shows the moisture to be within limits then there could be something in the oil like fibers from the filtering media or some other difficult to spot contamination in the oil. As I said above, a little research can usually dig out a problem like this.

A color comparison may give some clues as well. Transformer oil darkens very gradually over many years.  If the oil is one digit or ½ digit different from the last observation it could just be the opinion of the tester.  It may have been slightly lighter or darker than what the colorimeter slide was showing and was recorded according to the judgement of the testing technician.  If you can go back further and find earlier test results a tracking of the progression would be helpful.  If the oil is significantly lighter (by more than 1 digit) it probably indicates that the oil has been changed out. Significantly darker could indicate an overheating problem,  the oil may have been reclaimed with an oil reclamation unit, could have had some darker oil added to it, or there could be something going seriously wrong in the transformer.  First give the oil a sniff. What does it smell like? Could there be any foreign contaminants in it? Something soluble that would mix with the oil? Does it have a burnt smell? Check the temperature gauge and feel the tank with your hand to confirm it. Most temperature gauges have a recording needle that gets pushed to the maximum temperature reached that doesn’t recede when the current temperature recedes. What does that maximum figure read? Is it unusually high?  If so try to find out why. Was there an unusually high load placed on the unit for a period of time? Was the air restricted in some way so as to not be able to flow over the cooling fins? Feel the tank and the cooling fins to make sure that the oil is circulating through both.  Along with all of the observations a color change mandates a dissolved gas analysis to rule out several potential problems. 

I’ll continue on my next post and discuss acidity. Remember if you have questions email them to me at transformerbob@gmail.com. 

Looking forward to writing again soon.

PS Check out the website I’m setting up at Transformer Oil for Dummys.

Thanks,

Bob

The holy passion of friendship is so sweet and steady and loyal and enduring in nature that it will last through a whole lifetime, if not asked to lend money.

MARK TWAIN,