Transcript
Introduction
Good afternoon, everybody. I’m Charlie. I’m an engineer here at Power Monitors. This paper focuses on the specific issues that voltage imbalances cause in AC poly-phase induction motors. It also mentions how to calculate it and some possible causes one can look for to help mitigate any imbalances you or a customer may be experiencing.
The Impact of Voltage Unbalance
The biggest issue with voltage unbalance is that even small voltage unbalance percentages can create large negative sequence currents in a motor. These can lead to hotter windings, less torque, nuisance trips, and a shorter life.
If we take a look at figure 1, we can see a voltage balance three-phase vector diagram. All voltages in this diagram are equal magnitudes, which in this case is 500 volts AC, and are 120 degrees apart from each other. If we were to apply a triangle using the tips of the vectors as the points of the triangle, you’d get a nice equilateral triangle. Any motors operated on this system will be very happy for the lifespan of their usage.
If we switch over to figure 2, we’ll see a voltage unbalance has been introduced. One phase, specifically channel three, is lower and the geometry of that applied triangle skews cause them to no longer be equilateral. That skew in the voltage is what the motor ends up feeling in operation. One portions of the windings aren’t pulling like the others.
Common Causes of Voltage Unbalance
Voltage unbalance inside a facility is fairly common. It’s typically caused by poor single-phase load distribution pulling the three phases down. Most times, it can be traced back to an HVAC or single-phase motor load. Some other larger causes can include:
- An open wire or open delta transformer leg
- Undersized transform banks
- High current three-phase loads
- Single-phase to ground faults
- A lack of transmission line transposition leading to the facility
Calculating Voltage Unbalance: ANSI C84 Method
There are a couple ways to calculate voltage unbalance. Personally, I prefer utilizing the ANSI C84 formula. It’s fast, practical, and easy to calculate in the field. This formula is outlined in figure 3, and it utilizes your three line-to-line voltages.
In the paper, we have an example where we see that our line one to two voltage is 480 volts, our line two to line three voltage is 465 volts, and our line three to line one voltage is 490 volts. On a 480-volt nominal system, our max deviation is 15 volts. Our average voltage for this system is 478.33 volts, which equates to a 3.13% voltage unbalance.
Now, 3.13% may not seem like a lot, but according to NEMA MG1, the standard set forth by the National Electrical Manufacturers Association, specifies that less than 1% unbalance is ideal for motors. 2% is considered unacceptable, but is within limits, and at 5%, it’s advised not to operate a motor in that system at all.
Symmetrical Components Method
The ANSI C84 method is quick and easy. However, it is flawed in that it really only works best for delta setups. If you’re trying to calculate it on a Y setup, it’ll give you different percentages. Figure 4 shows us the most accurate calculation method, which is known as the symmetrical components method. This method works for both delta and Y setups and takes into account any unbalances caused by phasing changes.
Negative Sequence Voltages and Current Unbalance
From the outside, it may seem like manufacturers require a fairly tight acceptability window for voltage unbalances in motors, which is true, but it is for good reason. In an operating motor, voltage unbalance is introduced with what are called negative sequence voltages, and these create a counter-rotating flux on the motor’s rotor. This flux in operation is actively trying to stop the motor from running, and in turn creates a current unbalance that can be anywhere from six to ten times the magnitude of the voltage unbalance.
If we take a look at figure 5, we’ll see that our example unbalance percentage from earlier of 3.13%, we can see a 37 to 40% current unbalance under no load conditions, 22 to 24% at full load, and 3 to 6% at locked rotor. This graph also reveals that even with a 1% voltage unbalance, it can produce a 6% current unbalance at full load.
Current unbalances can result in a number of issues, such as windings overheating, reduced torque output, and even nuisance trips of overload devices. One way to mitigate nuisance trips is to size your overload protection for the maximum anticipated phase current that a phase will experience versus the average in an unbalanced condition.
Temperature Rise Due to Voltage Unbalance
If we take a look at figure 6, we’ll see the paper provides a formula for calculating out your temperature rise based on your voltage unbalance you’re experiencing. And then we also have a mirror of that for reference as a plot in figure 7.
If we utilize our example unbalance of 3.13% from earlier, the expected temperature rise is a staggering 19.6%. This temperature rise percentage in a motor operating, say, at 30 degrees Celsius, would cause the motor to rise to 35.9 degrees Celsius. This 5° Celsius rise doesn’t seem like a whole lot. However, that five degrees Celsius alone results in the motor’s winding life being cut by 25%. This is due in part that every 10° Celsius temperature rise a winding experiences, the winding life is cut in half.
Derating Motors for Voltage Unbalance
You can’t get voltage unbalance perfect. It’s very, very difficult to do. One of the ways that you can help mitigate issues with your motor is to derate. In specifically figure 8, we have a derating table set forth by NEMA MG1 that gives you a reference to go by.
If we utilize a 3.13% voltage unbalance, the derating factor equates approximately 0.88. So, a 10 horsepower motor becomes an effective 8.8 horsepower. If we’re having a machine that we need all 10 horsepower out of, we’d have to size up to a motor of at least 11 and a half horsepower to be able to get a full 10 horsepower usage.
Monitoring Voltage Unbalance with PQ Recorders
Unfortunately, unbalance can be tricky to track down as it can come and go with load differences, the time of day, or even different weather conditions. The easiest method for capturing these unbalances is to utilize a high resolution PQ recorder and leave it in place to monitor remotely.
With PQ Canvass, you can look at recordings from a PQ recorder, be it streamed from a vault, Revolution, et cetera, or uploaded manually. Figure 9 shows a graph view that is built into PQ Canvass for checking voltage unbalance on recording. This graph view allows for easy immediate representation of what your system is seeing.
This one in particular, we can see we have a max voltage unbalance of 1.8%. For the majority of the time, it is able to stay under 1.2%. The 1.8% spike in the beginning correlates to an approximately 12% current unbalance spike, with the remainder of the current unbalance, for the most part, sticking around 10% when it does come back.
However, if you’re in the field and don’t have access to a PQ recorder, it does happen. You’ll have to rely on the old faithful pen and paper method by measuring out your voltages, calculating out the unbalance, marking the date, time, location, and what you measured. You can then use this data that you’ve collected to help justify the need for a PQ recorder to be stationed at this location.
The bottom line is that small voltage unbalances can create big current unbalances and even larger issues for motors, including premature failure under normal operations.
Additional Insights from Chris
Thanks, Charlie. The real big takeaway here is that motors at above 1% voltage unbalance start to run hot if they’re at full load, and that heat over time degrades insulation and can cause an early failure. But 1% is very difficult, if not impossible, for a utility to achieve. NCC 84 calls for a 3% limit for utilities. So there’s kind of a built-in disconnect between what a motor wants and what a utility can practically provide. Even 2% is not always possible for a utility to achieve.
If you have customers that have motors that are failing, they may need to derate that motor. Whatever they can do to balance the phase currents at their location to reduce their contribution to the unbalance can help. But unbalance can kind of creep up on a feeder over time as laterals get added and the feeder currents become unbalanced.
So it’s important to try to keep those feeders balanced, maybe moving laterals around to different phases so that the unbalance is as low as you can make it, because it really does make a difference for a motor lifetime.
Current Unbalance and Motor Protection
One tricky point that Charlie mentioned is the fact that the motor’s current unbalance goes up dramatically as it’s lightly loaded, in terms of a percentage. Under no-load conditions or light-loaded conditions, the current unbalance is really, really high, like 10, 20 times, even 30 times the voltage unbalance, and that can fool motor protection.
Some motors have unbalanced current protection, and like Charlie mentioned, the motor may be fooled and take the motor offline because the current unbalance is really high, even with a low voltage unbalance. The utility could actually be at 1% voltage unbalance, and if the motor is at full load, that would be fine, but if the motor is at light load, then the current unbalance ratio may be higher than that protection is set for.
Like Charlie mentioned, it’s better to set that protection not on current unbalance percentage, but on, say, maximum link current, or at least make it more aware of the load of the motor itself.
Variable Frequency Drives and Voltage Unbalance
Another thing to keep in mind is that many motors these days are driven by variable frequency drive. In that situation, the motor doesn’t see the voltage unbalance anymore. The VFD itself sees the voltage unbalance, and that’s a whole different kettle of fish. The VFD has its own sensitivities to voltage unbalance. To learn about that, we have a voltage unbalance class that you can sign up for on the website.
If you have any questions about voltage unbalance in motors, give us a call anytime at 1-800-296-4120, or send an email to support@powermonitors.com.
