Motor Diagnostics Part 5: DC Motor Analysis by Dr. Howard Penrose, ALL-TEST Pro, A division of BJM Corp.
ALL-TEST Pro Scientists and Engineers continue to lead the industry in advanced data interpretation research and development. In the fifth of twelve presentations for 2004, Dr Penrose presents the latest secrets of evaluating the condition of DC motors and drives using motor diagnostics. You will learn how to troubleshoot and trend in minutes with motor circuit analysis and electrical signature analysis, including learning the limits of current signature analysis only versus voltage and current FFT analysis.
Once you have viewed the presentation, email Dr Penrose for your copy of the excerpt of the Electrical Signature Analysis Pattern Recognition Manual, the complete version only available to onsite and offsite ALL-TEST Pro Motor Diagnostics Level 1 trainees.
This tutorial runs for 10 minutes. Windows or REAL Media Player required for narration playback.
http://www.rcm-1.com/forms/bjm_reg.htm
Originally Published at Ask an Expert
Question: Large Motor Terminal Connection Surveys
Our station has just recently purchased an I.R. camera,and would like an expert's opinion on our first questionable image.
This involves the "peckerhead" connection on a 6.9.Kv motor. Our first survey of this motor (and an identical one at the same amperage), indicated a 5 degree C overall temperature difference within the connection box, and also the area of the box where the heat was concentrated (neutral leads). This was shot with the cover in place.
The second survey was done with the cover off. The neutral leads are 9 degrees C hotter than the surrounding areas within the box. The other motor has leads that are within 1 degree of the equivalent areas.
Question - is 9 degrees C significant enough to have the neutral leads checked for looseness?
Answer:
Thank you for your recent question posted at Reliabilityweb.com.
Although we have not seen your imagery, we would like to offer the following comments for your consideration.
For accurate infrared inspections, electrical covers must be removed to afford a clear line of sight to the subject equipment. This is required by published standards and represent current industry practice. Failure to open or remove covers can allow seriously deteriorated components to go completely undetected and potentially run to failure.
For your motor connections, the temperature differentials you are measuring are likely to be understated since you are not measuring the connection at the origin of the hotspot. Other issues which may cause a further understatement of the temperature rise would include, but not be limited to, spot measurement size of your camera, reflected temperature and emittance.
The larger issue for you is that an anomaly exists at all. Because there is no way to accurately predict time to failure based upon temperature rise, any inexplicable heating localized to electrical connections should be investigated as soon as possible. Any delay in correcting the cause of the anomaly merely puts you closer to an actual failure.
Infraspection Institute's Level II Certified Infrared Thermographer Courses teaches thermographers to take accurate non-contact temperature measurements and addresses the issue of utilizing scientifically based criteria for evaluating electrical and mechanical components.
We would invite you to join us for an upcoming course where we can show you how to increase your temperature measurement accuracy, reference applicable temperature standards, and get maximum benefit from your imager for your maintenance program.
We hope this information is helpful. Please keep us updated on your findings.
We hope this information is helpful and wish you the best of luck with your maintenance program.
Very truly yours,
R. James Seffrin, Director
Infraspection Institute
425 Ellis Street
Burlington, NJ 08016
Tel: 609-239-4788 ext. 13
Fax: 609-239-4766
Originally Published at Ask an Expert
Question: Induction Motor Formulas
I have seen formulas for motors running on full load, but have never seen a formula for no load, just rule of thumb. Question! is there such a formula ?
Answer:
The no-load current for an AC induction motor is actually complicated to calculate accurately, which is why most will use a rule of thumb ranging from 25 - 35% of full load current for design B motors. Single phase motors generally are close to full load current. I had also asked the EASA technical group, at a local meeting (we are a member of the MidAtlantic Region of EASA), whether the new software (coming in January) will calculate no-load amps and, unfortunately, it will not. There are no easy ways to determine no-load current without knowing the losses and total motor impedance at no-load slip. One method of estimating the no-load current is to get a copy of MotorMaster+ via www.oit.doe.gov/bestpractices then go to software tools. It is about a 10MB download. In the ‘List’ portion of the website you can search for a similar motor then look in the details section. For many manufacturers, the manufacturer’s no load and full load current draw values are available. If you are interested in the more dramatic details of calculating no-load current, please let me know. Sincerely,
Dr. Howard W. Penrose, Ph.D.
General Manager
ALL-TEST Pro
Originally Published at Ask an Epxert
Question: Phase To Phase
We have a compressor motor at 480v 217fla.
L1=135A
L2=126A
L3=144A
Is it a problem having this much difference phase to phase on L2/L3?
Answer:
The likely mechanisms of a stator winding fault are either a turn-to-turn or phase-to-phase short, or an insulation to ground fault. A turn-to-turn short is identified as a shorting of one or more windings in a coil. This can develop into a very low impedance loop of wire, which acts as a shorted secondary of a current transformer. This results in excessive current flow through this shorted loop, creating intense heat and possible insulation damage. Due to the nature of the low voltage random wound design a shorted turn could occur with much higher impedance, allowing the motor to run for extended periods of time before eventually burning up the coil with the high currents. As a result it is not unexpected to find low voltage motors still running with bad stator windings. Form wound coils however, do not exhibit high turn impedances and will therefore burn up quickly following the presence of a turn-to-turn short. A phase-to-phase short is identified as a shorting of one or more phases to another phase. This fault can be quite damaging due to the possibility of very large voltage potential existing between phases at the location of the short. In your case we have an inductive imbalance of 6.67% which is not considered abnormal for a 480v motor. This imbalance may be caused by the rotor by design or as a result of a rotor anomaly. Performing a Rotor Influence Check will allow a quick root cause analysis of the 6.67% imbalance identifying either the rotor or stator as the cause. Once you have identified the source you can better focus your analysis and trending.
Noah Bethel,
Product Development
PdMA Corp.
Originally Published at Ask an Expert
Question: Motor Service Factors
How does the service factor, i.e., 1.15, relate to the amperage rating of a motor in continuous service?
Is it acceptable to run at the top of the range e.g. 1.15 X 172 amps for an extended period and what, if any are the consequences?
Answer:
The service factor for a motor can be considered only when the motor is operating at a balanced voltage and at the correct voltage rating of the motor. So, in a motor that has 460 Vac on the nameplate and the system voltage is actually 490 Vac, the 1.15 Service Factor will most likely not be acceptable. Operating a motor above full load will also put the motor into an area of lower operating efficiency, making it more expensive for the motor and will reduce its operating life.
Dr. Howard W. Penrose, Ph.D.
General Manager
ALL-TEST Pro A Division of BJM Corp
123 Spencer Plains Rd
Old Saybrook, CT 06475
Ph: 860 399-5937
Fax: 860 399-7784
All Test Pro
Concepts in Calculating Failure Estimation
Howard W Penrose, Ph.D.
General Manager
ALL-TEST Pro, A Divison of BJM Corp
hpenrose@alltestpro.com
1. Definition
Predictive maintenance is a misnomer, as anyone worth their salt in the reliability industry understands. In reality, you take a series of tests with the intention of detecting problems then make a ‘best guess’ estimation of how long the machine will survive. This means that ‘predictive’ maintenance is actually ‘estimation’ maintenance. The more information, the better the estimation. However, as with all things, there is a point of diminishing returns.
2. Consideration
“Oh, so how long can we run it before it fails?” The most common, and feared, question in the reliability industry. The answers tend to come in one, or more, of the following forms:
a) “Well, it depends on…”
b) “I don’t know.” (the ‘honest’ answer)
c) “It will require more study…”
d) “I will get back to you.”
e) The dumb look, the ‘reliability shuffle’ (staring at feet, or elsewhere, while moving uncomfortably), or the shrug.
After the question and answer period, which may actually take some time, the reliability or maintenance professional wonders why their warnings are not taken seriously. The only way to be taken seriously is actually to provide a figure or estimation.
If you go to a manufacturer/vendor, engineer or other, you will also go through the above answers. Either that, or you will be given a LONG list of reasons why it cannot be done, including all the different requirements and information necessary. By the time all is said and done, either management/production has forgotten, you have moved on to other things, or the machinery has failed. The result of all of this is that the program will be brought into question and you, and your position, may be next for ‘cost cutting’ consideration.
Should you be surprised? No. In today’s competitive economy anything non-essential or broken must be removed. By not being able to provide answers, right or wrong, your program is broken. Many present programs are actually held in place, not because of success, but because it is becoming the ‘in thing’ to do, much like using words such as ‘paradigm’ in management.
What can you do?
3. The KISS Principle
The reasons why the answers are difficult for us are simple: We tend to be conservative; We are technical and exacting (used to dealing with exact measurements or values); and/or, We are concerned about our position or reputation. The middle point is of greatest concern and one of the primary causes of reliability program failure.
The communication gap that develops will break the program. Management wants to know about how long the equipment will last and at what risk. The engineer or technician thinks that they would have to provide an exact the answer (“The motor will fail next Tuesday at 2:59 PM and 30 seconds”). Is this possible? No (well, actually possible, but not probable or effective). And, as a result, the engineer or tech ‘freezes.’ In effect, by trying to stay credible, you have the chance to lose your present credibility.
The best way, in my humble opinion, is the KISS (Keep It Simple Stupid) principle. For this, just do these simple things:
a) Remember that you cannot get all of the information to provide an exact answer.
b) All things become less reliable with time.
c) It’s an estimation, not a prediction!
Use the information you have available in tests, the equipment and operation. Remember, the manager wants to know risk, not really an exact time. For instance, a good conversation would look like:
Tech: “The product pump motor on Line 1 is showing signs of insulation degradation.”
Manager (who probably may not even know what ‘insulation degradation’ is): “So, how long do we have?”
Tech: “I estimate that the motor will last 2-3 weeks, maybe less, if we do not perform corrective action.” (having infomraiton and trending data).
Manager: “What is the chance it will last to 3 weeks?”
Tech: “If we change the motor now, and commission a new one, the rest of the line looks good and should not have a problem. However, if we need to keep production running, I would give the motor a (estimation) percent chance. We can improve the chances by…”
You have now provided the manager with options and risks. This is normally enough for the manager as s/he now has the information needed to make a decision. Also, put the options in writing and stress ‘estimation.’ Otherwise, you may end up with the opposite problem of “You said it would last three weeks.” By presenting the risk and providing ‘educated estimates’ (don’t just throw out numbers), you will begin to gain a positive reputation.
Just as in health care, when a physician provides a preventive exam (check up), if s/he spots an issue, there may be the requirement for more tests. The better doctors will end up providing options to correct the problem and may even provide chances for success.
4. How to KISS a Motor
Ask any electric motor or insulation system engineer about how to predict insulation or winding failure in a motor, “It is not possible.” However, by looking at a few simple things, you can begin to estimate the chance that the motor will operate for a given time.
In the Time to Failure paper posted in ReliabilityWeb in May, 2004, I introduced the concept behind motor circuit analysis time to failure estimation. In Part 2, to be published in June, 2004, I discuss the impact of winding contamination, load and cycling on the estimation. Part 3, to be published at a later date, will view other issues such as power quality and load condition. All of these are guided towards providing a ballpark time before catestrophic failure and require only a reasonable amount of information. We have built the estimation tool into our EMCAT motor management software, to be released as Version 1.4.0 in June, 2004.
5. Conclusion
While most systems do not lend themselves well to predicting time to failure, the technician can provide a ballpark estimate on time to to failure. This, combined with options, will provide managers with the infomration required to make decisions. You will look decisive and will assist in advancing your reliability or maintenance program.
Originally Published at Ask an Expert
Question: Surge Test
What is a surge test and how do I interpret the results?
The particular application is a 5 kV motor, approximately 1200 hp that was tested in a shop and I am wanting to understand the results data and the pass/fail evaluation.
Answer:
Very basically: Surge Testing is a procedure for detecting and evaluating faults in electrical windings, such as: Turn-to-turn and phase-to-phase shorts, opens, grounds, and weak or damaged insulation. Motor rewind shops have been using surge testing equipment for years. It is a good test for new and/or rewound motors. Surge testing is not recommended as a predictive maintenance tool. Almost all electrical windings are comprised of several identical coils or phases. Surge testers evaluate the integrity of these windings to detect faults. The surge tester applies a pulse to motor windings. This produces waveforms on the instrument display. If there is no current leakage between turns or phases or to ground, the winding is good. The waveforms will be stable and identical. If there is a turn-to-turn or phase-to-phase weakness, the wave patterns will separate and appear unstable. If there is weak insulation, the wave pattern will indicate the weakness. To accurately analyze the waveform produced from the pulsating voltage application takes a skilled trained individual. There are charts available from the manufacture of surge testers which show basic pattern recognition examples.
Don Shaw.
Application Engineer
PdMA Corp.
Originally Published at Ask an Expert
Question: Motor Amperage Imbalance
There is much information on the effects of voltage imbalance on 3 phase electric motors but I can not find specifics on the affects of amperage imbalance. We have three motors with amperage imbalances in the range of 20 and 30 percent on one leg of the 3 phase motors. The other two phases are very close. The imbalance is not on the same phase and switching the leads tot he pumps only moves the imbalance with it. These are two speed single winding motors. What affects does this have on the efficiency and longevity of the motors?
Answer:
A current imbalance of 20-30% is rather significant. The effect of the higher current is higher temperatures in that phase. The higher temperatures will adversely affect the insulation. The insulation will deteriorate and cause coil-to-coil shorts to develop resulting in catastrophic failure of the motor. A current imbalance is normally caused by a high resistance somewhere in the circuit downstream from where the current readings were taken. I would recommend opening the motor junction box and record the currents in all three phases. Assuming that the currents are balanced at the motor junction box, start looking for the cause of the high current between the motor and the Motor Control Center (MCC). If the currents are not balanced at the motor junction box the problem is in the motor and the motor needs to be replaced. Efficiency with an imbalance like this would not be what would be expected form a balanced current situation.
Don Shaw,
Application Engineer
PdMA Corp.
Originally Published at Ask an Expert
Question: Maintaining Redundant Motors
Our maintenance program is being assessed and we are looking at better alternatives to maintaining our redundant motors (100 - 200 HP). These non-critical motors are not going through any extensive motor testing like MCA but we'd like to at least do an insulation check on them every year.
Most of them are quite old (>20 years) and exposed to a moist environment. Currently the electricians have been logging a "quick" resistance check value and I'm wondering if a Polarization Index would be more helpful. What are the differences and advantages between a Polarization Index evaluation, insulation resistance check and a dielectric-absorption test (ratio)? I'm looking for some justification that would qualify doing a P.I. check since this type of check is found to be a dubious amount of information and time involved. Yes, it's a 10 min reading but with phase-to-phase and phase-to-ground checks this could result in going from .5 hrs to about 1.5 hours and we have a numerous amount of motors in our water distribution systems. Help anyone?
Answer:
Part 1. A "quick" resistance check can be a fair comparison or trend of the motors insulation resistance to ground value . Temperature correction is essential to maximize the effectiveness of this type of basic trend. Humidity changes can create an additional variable in the resistance value which can make the analysis a little more difficult and require more historical data to make a proper assessment. A Dielectric Absorption (DA) test is a better method of trending or comparing insulation integrity. The ratio of the 60 second resistance divided by the 30 second resistance should be around 1.5. The advantage is that rather then trending a single reading effected by variables, the DA ratio removes those variables and provides a more trendable value. The Polarization Index (PI) is the best of the 3 tests in that the 10 minute time frame not only removes the variables like the DA test but it also provides time for the absorption current to completely decay leaving only the leakage current for evaluation. Technology offering a DA or PI profile (graphing capability) of the resistance values throughout the DA or PI tests further increase the effectiveness of these tests. The profile of the changing resistance measurement is often a more effective tool in identifying insulation integrity problems. Part 2. With so many companies being faced with limited manpower and production demands it is understandable that thought processes of reducing the amount of testing to a minimum on non-critical redundant motors. However, there are numerous fault zones involved with an electric motor that can cause the motor to fail. As seen recently in the testing tips at Reliabilityweb.com the Power Circuit, Power Quality, Insulation, Stator, Rotor and Air Gap are each fault zones that can cause your motor to fail and remove the redundancy. A PI, DA or Megger will not allow you to evaluate each of these fault zones. At minimum I would recommend a baseline of all fault zones followed by a reduced approach of insulation integrity measurements of a DA or preferably a PI. If the PI test indicates degradation in the insulation system then a full reevaluation would be recommended to ensure the motor is reliable for backup. Best Regards,
Noah P. Bethel
Product Development
PdMA Corp.
Originally Published at Ask an Expert
Question: DC Hi-Pot Testing of High Voltage AC Induction Motor
Do you recommend the DC high voltage testing of AC high voltage induction motor say for a two year old 6.6 kV AC DOL start 3.8 MW induction motor at 14 KV DC ? If yes , then describe the precautions & procedures .
If no , then Pl. say why not and with the quotation from widely accepted standards or guides.
Answer:
As stated in Electrical Apparatus Service Association, EASA, Principles of Large AC Motors. Version 600CL100-700, Chapter 9, Test and Inspection, page 9-19, High Potential, Paragraph 5, NOTE: The test can be destructive and deterioration effects are cumulative. Repeated application of test is not recommended. Refer to NEMA MG 1 - 1998 for proper precautions when performing this test. In general high potential testing can be used to assure increased probability of continued operation for finite periods of time by elimination of weak units through failure. Insulation testing, Polarization Index, resistance measurements are a preferred method for evaluating the insulation integrity over the life of the motor. The high potential test is not recommended as a predictive tool. Data generated is either pass or fail. High potential testing is a good means for testing insulation integrity of new and rewound motors. If used on motors in service 60% of the recommended test voltage should be applied. Be careful you may damage the motor.
Don Shaw
Application Engineer
PdMA Corp.
Originally Published at Ask an Expert
Question: PI vs. HiPot
I am having a debate with a fellow worker on the whether to hi-pot motors already in service or to do only a PI test. I have told him that a PI test is a non destructive test and gives a very good snap shot of insulation condition. He feels that we should HiPot so that he knows that the insulation is good. Please give me some feedback on this. I have shown him various articles but would like an answer here also. Thank you.
Answer:
This debate has been around for a long time and is currently at a peak for discussion. We are regularly asked which application should be used, if Hi-pot testing is destructive, and whether or not a PI will be able to identify weakened insulation. From our view, the most widely accepted approach to insulation testing is to reserve the potentially destructive testing for refurbishment or repair and trend the insulation health to maximize the life of the insulation. There should be no debate that a Hi-pot is a potentially destructive test even with today's technology which allows for quick shutdown in the presence of excess current flow. The Hi-pot/surge testing is an extremely critical tool however, in ensuring that a motor was properly manufactured or rebuilt. Analogy I spent some time on a submarine. I was comfortable knowing that a deep dive quality assurance test for hull structure integrity was performed on the submarine after commissioning. Also, after each repair or refurbishment of the hull a deep dive was performed to verify the integrity of the hull. I would not be comfortable if every time we took the submarine out on a patrol that we would be stressing the hull to that level. I would much prefer saving the stress tests for an actual situation that we needed a deep dive. Conclusion It is recommended to trend the condition of the insulation system using non-destructive testing such as a Polarization Index "Profile" and Step Voltage tests. This will give you a trendable indication of the surface contamination and insulation integrity. If the contamination is excessive or the trend of insulation integrity is indicating a significant reduction then the motor should be visually inspected and sent in for a refurbishment as necessary. Following a refurbishment or rewind a stress test should be applied for dielectric strength integrity before returning it to your facility. If it fails it can be repaired or rewound as necessary. If it passes, it should be able to withstand the stresses of normal operation and trending again will give you a safe and early detection of a future insulation integrity problem.
Best Regards,
Noah Bethel
PdMA Corp
Originally Published at Ask an Expert
Question: Electric Motor Temperatures
We are trying to justify the costs of TPM. As we clean, inspect motors, and install filters, we are reducing the motors' temperatures. Is there any information on cost savings related to a reduction in motor temperature.
Answer:
Yes, there are very definite ways of determining cost savings related to a reduction in motor temperature. As a matter of fact, in the PG&E market transformation study that I designed, and a US Department of Energy project I was involved in, the savings can be related to energy, production and reliability costs (cost avoidance). I outline these in my book (Motor Circuit Analysis: Theory, Application and Energy Analysis). By the way, general rule of thumb for motor life: reduction in insulation life by half for every 10 degrees C increase, double for every decrease by 10 degrees C. Hope this is of assistance.
Sincerely,
Dr. Howard W. Penrose, Ph.D.
General Manager
ALL-TEST Division BJM Corp
123 Spencer Plains Rd
Old Saybrook, CT 06475
Ph: 860 399-5937
Fax: 860 399-7784
ALL-TEST Pro
Originally Published at Ask an Expert
Question: Slip Frequency
Can I expect to find a slip frequency when operating a AC squirrel cage motor through a variable frequency drive? At 60Hz through a variable frequency drive will the slip be the same as across the line operation?
Answer:
Yes, a slip frequency in the form of a pole pass frequency (Fp) sideband can be seen in both across the line systems and the output to a motor from a VFD. It is important to remember that the math changes significantly with the different line frequency being issued by the VFD. It is also important to remember that the noise levels on the output of the VFD will be higher and may mask the amplitude of the Fp sideband until the severity of the fault producing the problem increases. If you were to increase the frequency of the drive to 60 Hz you should see the same Fp sideband on the drive motor as you would an across the line motor. However, the noise levels should be much higher. A good software program associated with your current analysis technology should provide the ability to identify the line frequency first then automatically adjust for the correct identification of the Fp sideband. Also, high quality current probes will reduce the noise level offering earlier detection of rotor defects even with the elevated noise of a VFD.
Best Regards,
Noah Bethel
PdMA Corp