Electrician Talk banner

When is a VFD cable actually "required" ?

35K views 45 replies 24 participants last post by  sledge 
#1 ·
When I was a first year I helped a journeyman pipe a mechanical room in a commercial building. We connected probably twenty 600v motors and drives and I distinctly remember using EMT, metallic liquid tight flex and 600v rated T90 conductors.

From what I learned in school and on this forum since that time, that installation would be less than ideal because the voltage spikes associated with rectifying AC would create corona discharge that pinhole the insulation. By my calculations, on a 600v system the rectified DC would be 846v.

I see that most cable manufacturers sell at least one type of VFD cable. Since VFDs are everywhere now, when is one of these special cables actually a necessity? I ask because I've never actually seen any in person. Whenever I've seen a VFD in an industrial application the motor has generally been connected using 1000v Teck 90. Is this acceptable or are you theoretically supposed to use VFD cable "every time"?
 
#3 ·
We do the same as Don. We would use FMC or LFMC for the connection to the motor.

VFD cable would only be for open wiring designs.
 
#4 ·
That's a good question that'd I'd like to know the answer to also.

I realize VFD manufacturers most likely recommend it for every install, but myself, I've really only installed it when I've also used EMI/RFI filters because of potential issues with RFID readers for cow ear tags.

Another thing I've read is to keep VFD motor leads in their own metallic conduit and away from other motor leads. I've been on one job where the VFD leads have been XHHW or THHN in their own pvc conduit, but mixed in with a bank of other conduits, some 400+ feet long, which include other VFD conduits as well.

What's surprising is everything I've read says this type of install should cause VFD issues and the THHN/XHHW insulation should break down, but this particular install is 4+ years old and hasn't had any wire or drives replaced to my knowledge for this issue.

This makes it difficult to justify the cost for me of VFD cable plus oversized conduits unless I'm missing something in the bigger picture that I don't know about yet?
 
#9 ·
I have started specifying XHHW for VFDs now, just because that is what is used in VFD cables, but it is not based on any type of failure. It is also a tougher cable and I specify it for all underground runs too.

We have also connected may older non-inverter rated motors to VFDs without any issues and we typically don't use load side reactors unless the wiring between the VFD and the motor is over 500'
 
#11 ·
This is something that you should not rely on as "Rule of Thumb" type of thing. If you're wrong and have only pulled single conductors and need to instead pull cable then your conduit may be too small. Plus you'll end up looking bad.

I try to call the VFD's manufacturer and let them tell me. Running the cable drives up the installation cost but at least you've done your due diligence.
 
#18 ·
If I called the manufacturer every time I hooked up a vfd I'd get less done than I do now. Of course drive manufacturers will tell you it's best to use their brand drive cable but why wouldn't they. I couldn't even begin to guess the amount of vfds I've hooked up in the last 5 years alone that weren't simple pipe and wire.


Sent from my iPad using Tapatalk
 
#13 ·
IIRC, the universal standard test for 600 V conductors is TWICE the labeled voltage PLUS + 1000 VAC. ( It's a NEMA thing, IIRC. )

Hence the run-of-the-mill THWN-2/ MTW/ XHHW-2 is routinely tested at 2,200 Volts before it's kicked out the door. (So they say. I've seen the occasional conductor with factory flaws -- which should've busted out into a corona at 2,200 volts.)

So, it's no surprise that such conductors can tolerate ordinary VFD harmonics.
 
#14 ·
From what I learned in school and on this forum since that time, that installation would be less than ideal because the voltage spikes associated with rectifying AC would create corona discharge that pinhole the insulation. By my calculations, on a 600v system the rectified DC would be 846v.
A regular AC sine wave from your favorite local utility at 600 Vrms will also peak at 846V, so I don't see why that would be any different on a VFD from a purely "peak voltage" standpoint.
 
#16 ·
I feel compelled to correct some slight misinterpretations here (and confirm other statements). I had a doozy of a reply all typed up on this, and my internet went down before I was able to post it... I hate when that happens. :censored:

This is part of what I do for a living, so I'm pretty sure I have posted all that someplace at one time or another and rather than re-do it now, I'm going to try to find it and link to it.
 
#27 · (Edited)
OK, I can't find anything I have already done that covers all of the issues raised, so once again, into the breech...

There are SEVERAL issues to contend with regarding the VFD to motor lead wiring; capacitive coupling, voltage spikes from reflected / standing waves, motor winding insulation damage, motor bearing damage, EMI/RFI interference and Common Mode Noise problems. They are mostly all interrelated, starting with the capacitive coupling issue and similar effects. But to avoid a long boring story, let's just say that VFD cable solves SOME problems, not all of them.

The main one it addresses is the EMI/RFI issue. The output cables of a VFD are, to over simplify it, like a powerful local FM radio transmitter. FM is "Frequency Modulation", which is exactly what a VFD is doing (radio purists, please excuse my over generalization here, it's for effect...) This is where the idea that putting VFD cable INSIDE of steel conduit is redundant. The steel conduit has the same effect as far as keeping the RF inside. But as was mentioned, if you have MULTIPLE VFD cables in a single conduit, then you MUST use shielded VFD cable, otherwise the different frequencies of the multiple outputs will cause induction between each other, as WELL as the RF bleeding from one to another.

The VFD cable also uses a more symmetric geometry of the cables, and combined with better insulation, can help reduce the cable capacitance issues you may encounter using separate conductors in conduit, which is what leads to refelcted / standing wave spikes. But if your distances are shorter, that might not be an issue anyway, so that alone is not a good enough reason to always use it.

The fact that the VFD cable always has a good over sized ground conductor, or multiples, and a shield that is grounded on both ends, also helps cut down on common mode noise creating in the cables and helps avoid transmitting CM noise created in the VFD to other nearby equipment. But again, that alone is not usually a good enough reason to use it all the time.

Lastly, there is evidence that by using a good VFD cable with XLPE insulation on the conductors, can help to reduce the surge capacitance of the cable itself. That can, again based on circumstances, be cause for concern related to capacitive charging current required from the VFD, which can "rob" your motor of available current and decrease the shaft torque, and is also a contributor to reflected wave creation.

So bottom line, the only time I tell people the MUST use VFD cable is under the following conditions:

  1. PVC conduit or cable tray installations (aluminum conduit is problematic too); in other words you are not using steel conduit.
  2. Flexible cables, not in conduit at all; do NOT use SO or other portable cord for VFD outputs!
  3. Installations where you are exceeding the maximum recommended cable distance of the VFD in question.
  4. Places where you ALREADY know you have a problem.
Some of the same issues mentioned above can be ALSO mitigated by using filters on the output of the VFD, but the only thing that CANNOT be mitigated is the EMI/RFI issue.

Re: Insulation type.
The voltage spikes that can damage the motor insulation is based on the Corona Inception Voltage (the point at which a corona discharge occurs) level of the insulation in the magnet wire. So on 600V insulation, the peak rating of older motors was 1200V. But on a 480V line, the reflected wave spikes can reach almost 1600V, so far above what the old motor could tolerate. "Inverter Spike Resistant" (ISR) magnet wire raised the peak level to 1600V or more, with a CIV of over 2400V. That is what you get when you buy an "inverter rated" motor., so that solved THAT problem. Still, if you DON'T have an Inverter Duty motor, you need to worry about this.

But the CIV issue takes place in the motor leads too, regardless of whether the motor insulation can take it or not. PVC insulation, as found in THHN / THWN cable, is typically 15mil thick and has a CIV of at least 2400V, but that can go down by as much as 50% if the wire is wet, and even further if it is nicked in pulling. In addition, over time now we have seen that because the PVC is injected in a liquid form around the wire, it can have microscopic bubbles in it, which allow the CIV to be even lower yet. So even if the wire passes muster for standard testing based on sine wave power, it might not be suitable for VFD outputs without compromising the longevity of the installation. Because cable is EXPECTED to last 25-50 years, and high speed transistor VFDs have only been around for 20 or so, the foreshortened life of THHN is only now coming to light. I have seen 3 different installations now in the last 2 years in which older THHN cables were pulled out, and you can see the burn marks that are tell tale signs of corona discharge happening.

XLPE (Cross Linked PolyEthylene) insulation, as is used in RHHW cable, is 30mil thick, heat shrink applied to the wire so there are no bubbles, and is rated for at least 1000VAC RMS with a CIV of over 4,000V. It is also less susceptible to water infiltration and has been shown to lose less than 30% of it's CIV capacity when wet, which is still FAR above the levels that can be seen on VFD outputs. Many of the VFD cables on the market will be made with XLPE insulation instead of PVC (but not all, so check). So if you ARE going to use steel conduit, I am recommending that people start using RHHW conductors now, not THHN. This only applies to the OUTPUT side of the VFD, nothing special goes on on the input side.
 
#28 ·
...

XLPE (Cross Linked PolyEthylene) insulation, as is used in RHHW cable, is 30mil thick, heat shrink applied to the wire so there are no bubbles, and is rated for at least 1000VAC RMS with a CIV of over 4,000V. It is also less susceptible to water infiltration and has been shown to lose less than 30% of it's CIV capacity when wet, which is still FAR above the levels that can be seen on VFD outputs. Many of the VFD cables on the market will be made with XLPE insulation instead of PVC (but not all, so check). So if you ARE going to use steel conduit, I am recommending that people start using RHHW conductors now, not THHN. This only applies to the OUTPUT side of the VFD, nothing special goes on on the input side.
XHHW is also XLPE and is used in some brands of VFD cables.
 
#33 ·
That's what I recommend, although the SO cord on the input would be subject to the normal restrictions, and that too can allow some EMI/RFI bleeding as well. That's why most VFD mfrs sell what they call "EMC Filter" options for the line side of their VFDs. VFDs sold here in the US have to meet basic FCC regulations with regard to EMI emissions so most people don't have issues with it, but the regulations in the rest of the world are tighter. Still, using portable cord is "poking the bear" in that regard.

I've run VFD outputs through buried PVC using THHN many times. Longest run was about 230' or so.

On longer runs I'll use some sort of a load reactor though.
Like I said, it's not a simplistic issue, there are multiple factors that go into it so it doesn't always happen, but when it does, you may not realize it.

Have you ever pulled out any of the old runs of cable and looked at them yet? That's one of the issues. I worked on a project last year that had been installed for about 9 years; parallel 2/0 THHN cables per phase in steel conduit, about 130-150' linear run to the motors from the 200HP VFDs. Multiple transistor failures on the VFDs was causing suspicion, so we pulled the cables and stretched them out on the floor. You could see the burn marks at regular distances all along the cables where the standing waves peaked and broke down the insulation, burning through to the other cables. Meggers didn't pick it up because the burn through was only when energized and creating corona discharge, and even then it was phase to phase, not phase to ground. So nobody knew until we could see it with our own eyes. But you could extrapolate the wavelength because the burn marks were fairly evenly spaced at about every 16 feet or so. Granted, this is the only one I have actually seen with my own eyes, but it's also the only time I have pulled the cables and looked.
 
#36 ·
Here in Canada in industrial environments it is pretty much always aluminum interlocked armored cable (ie TECK90) in tray.

The "VFD" variant of this style of cable (basically it adds symmetrical grounds + shield) is not significantly more expensive than standard TECK. So, we generally always use it for 50hp and greater, or 20hp and greater when the run is more than 150 feet or so.

So for us it's not so much a question of when is it needed, but given the small price difference, why not?

If the standard install is wires in conduit though I guess it's a drastic change to the install.
 
#40 ·
Here's a case of our client,

The client called and said that the incoming cable of the dynamometer equipment (an equipment similar to VFD, but it can also generate electricity and feed back to the power grid) should be shielded, and the capacity of the dynamometer is about 12x500 = 6000kVA. The initial operation consumes electricity and then generates electricity. Its power supply is required to be shielded.

After receiving the call, we asked the client to confirm with the supplier that if neutral line is needed for the power supply of the dynamometer equipment. The result is unnecessary.

So when designing its products, AVL (the world's largest dynamometer) has considered the possibility of using 3 + 3 structure cable.

Check our website for all brands of VFDs and other parts.
 
#41 ·
What are you talking about? So if the supply side must be shielded where does the shielding end? Which end it ends need the drain connected? What is the purpose? Do you require IEEE 386 elbow connectors to the transformer? Is this medium voltage? If so it makes sense. Otherwise not. As for structured cabling you can’t be serious? You are really going to push 6000 kVA across 6 CAT 6 cables?

So what cable do you actually specify/recommend? Please use standard IEC or UL terminology and specifications, not some made up AWM.

As to regeneration not being a VFD may I suggest you do your homework? Have you ever heard of an active front end (AFE). Sure the common pump/fan drive with a diode front end can’t regenerate but regeneration in general has been around for decades. It just requires slightly more expensive VFDs that are often used for centrifuges, conveyors, cranes, and various machine tools. In medium voltage drives due to the already complicated multi pulse architecture AFEs are rare but they certainly exist. It’s just that most MV drives are for process equipment that doesn’t need regenerative capabilities.

Some type of power conversion is always at the heart of dynos. They are for torque testing and are effectively large brakes that can perform sustained braking. Even the most primitive just convert it to heat and exhaust it while the most fancy regen and back feed the recovered power back into the device under test.
 
#42 ·
Unless your working above 6000 feet above sea level and your greater than 2500v. Corona is not much of a problem. I had a job located at 7500 asl and working voltages were 12.47kv. We had to wrap the exposed bus. The breakers had arc shoots and that contained the arc when opening or closing.
Also had to plastic bag all of the end of the conductors when high poting.
 
#43 ·
Unless your working above 6000 feet above sea level and your greater than 2500v. Corona is not much of a problem. I had a job located at 7500 asl and working voltages were 12.47kv. We had to wrap the exposed bus. The breakers had arc shoots and that contained the arc when opening or closing.
Also had to plastic bag all of the end of the conductors when high poting.
Repeat and simply not true. I can show you lots of photos of cables that have corona damage roughly at sea level. I haven't seen much of any of it at 2500 V (phase to phase) or even 4160 (phase to phase) but by the time you get to around 6 kV or higher, it is definitely there and gets worse as the voltage goes up. It is highly noticeable after a couple years on unshielded substation cabling especially when it goes through panels or through CT windows.

Mild cases of it but this is what it looks like:


Typically where we see it is when they jam cables into too small of a CT window for a zero sequence CT, or they drape cables through a knife-edge piece of steel instead of cutting and mounting a piece of glastic to keep it away from the grounded edge, or failing to use some kind of spacers to route cable bundles (using C-channel cut to make small spacers with cable ties works well), or just letting the cables lay against the sides/floor of grounded metal substation enclosures without a sheet of glastic or tieing it up neatly to avoid getting too close, or letting cables lay up against each other at the terminations when they are all different lengths or one is twisted up to the point where an unshielded cable lays up against a stress cone or two stress cones lay against each other. How good/bad of a job you do is the difference between having problems in a year or two or a decade or two, and voltages increase the problem exponentially. Starting at around 15 kV for instance is the point where you need to start using a rasp to clean every last rough edge off of the lugs and start wrapping them in semicon where at 4160 it's usually good enough to do a little careful routing and liberal use of glastic.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top