mbrooke
Batteries Included
- Location
- United States
- Occupation
- Technician
What advantages does Tan Delta testing have over a simple Megger? Isn't R what we are really after? ...............................
Tan Delta vs Megger
- Thread starter mbrooke
- Start date
- Status
Devin Hanes
Member
- Location
- United States
DC Hi-Pot is usually only recommended(I believe the IEEE standards and manufacturers like Okonite) on insulation less than 5 years old, Tan Delta can be used on aged. DC Hi-pot in my experience is only used for pass/fail, Tan Delta you can use the findings to watch the aging process and for predictive maintenance.
mbrooke
Batteries Included
- Location
- United States
- Occupation
- Technician
Had not though about that, good point!
What is the voltage on tan delta? How well does tan delta pick up treeing vs general conditions capable of partial discharge?
mbrooke
Batteries Included
- Location
- United States
- Occupation
- Technician
So what do you typically megger cables at?
Cables are typically tested per NETA MTS or roughly 1.4 times rated voltage. So for instance a 600 V cable gets 1000 V. Most medium voltage cables though just get 5 kV because few companies have a higher voltage Megger. The primary use is motor testing and 5 kV is the top of the chart for those.
The motor insulation test uses lower voltages. All motors up to 1 kV get 500 V. All motors 1 kV up to 5 kV get 1000 V. Then it goes up from there but I have to look at the chart since the higher voltage motors usually get automated testing. Usually motor technicians test motor feeders at the same voltages.
So a 600 V motor feeder cable would get 500 V.
Don’t get hung up on the voltages. The issue with voltage is that the “system” is a big capacitor. The reason to use a Megger is to saturate the capacitor so that you can see the leakage...well technically leakage+polarization current, which is a strong indication of insulation damage.
At higher voltages still you can activate electrical trees, aka partial discharge. So tests have been devised to raise voltage high enough to initiate corona arcing then lower it back down before it does excessive damage. That’s the principal behind VLF...do LESS damage. Surge testing motors (with corona detection) is the same idea. If the corona discharge were to occur at normal working voltages the cable would fail quickly. There is one company heavily promoting this idea for cables but all their published information is circular references and their main “research” guy is basically a fraud. You would not believe the crap he is peddling.
The second type of corona testing in my opinion is pure genius. They just measure corona discharges online. Switching naturally generates surges which is what corona testing needs. If there are no switching surges large enough then no matter what the cable condition is we don’t care. So it becomes a simple diagnostic test kind of like vibration, thermal, and UV.
I’ve seen lots of praise by people promoting tan delta but the issue is I haven’t seen anything as far as predictive or diagnostic info. What I mean is that the basic thing with a Megger test is I can take a motor and “ohm it out” with a multimeter. The motor checks good. Then I fire it up and it trips the breaker in seconds. With a Megger it clearly fails dead shorted. If I were to do one and only one test for motor diagnostics that is it. There are motors that are still bad that “pass” but it screens the majority in one test. The problem is it doesn’t indicate “why” (contamination, insulation failure, or just moisture). You can’t reliably trend the data over time. And it’s not predictive...it doesn’t tell you it will soon fail.
Tan delta is like that, too. There doesn’t seem to be a lot of support suggesting it does much more than Megger testing. I don’t mean purveyors don’t advertise but face it, it’s not one of those tests that you see printed everywhere as part of a big battery of tests. In motors for instance the popular tests once we get past Megger is the big war between the Baker tester (surge testing) and what amounts to LCR bridge testing, which is all Baker is anyway at least indirectly. You can easily convert the results to tan delta but nobody does. The biggest use I’ve seen for tan delta is that it easily finds dead or failing surge arresters and internal insulator faults. But that stuff is obvious on Megger tests, too. I haven’t seen for instance reports of marginal cases where Megger found nothing but tan delta found something and inspection verified an issue.
- Status