VFD wiring methods question

[Aleman]

I have always been under the impression that it is a bad practice to run VFD line and load conductors in the same conduit.
My understanding is that the loads will put a lot of noise into the line conductors. Anyone care to school me on this one? Some of
you guys know this stuff much better than I do so I figured I would ask.

Thanks

 

[Cow]

Jraef has made a few posts about this in the past explaining the downsides. Maybe he will chime in again.

I know I've personally seen a lot of installations with multiple VFD's load side conductors in the same conduit. Or sharing the conduit with 120v and 480v circuit conductors. I can't say I've really seen an issue to speak of, but I know I haven't seen it all. With Jraef preaching the negatives on the forums and I know his experience is far more vast than mine, I've started running load side VFD conductors in their own conduits as well as using a lot more VFD cable when I can get away with it.

 

[MattS87]

I have always been under the impression that it is a bad practice to run VFD line and load conductors in the same conduit.
My understanding is that the loads will put a lot of noise into the line conductors. Anyone care to school me on this one? Some of
you guys know this stuff much better than I do so I figured I would ask.

Thanks

I can't speak for all manufacturers but Toshiba does state it their manual that you can not run both in the same conduit. I have also talked with a tech of theirs to see if it is related to lead length or drive size and he said no.

 

[Jraef]

Whenever you run multiple 3 phase conductors in the same raceway, there is a potential for the expansion and contraction of one conductor's magnetic fields to induce a current on an adjacent conductor. But when all three phases come from the same source, even all the way back to the utility generator, these effects will cancel each other out.

But the output conductors of a VFD are being created from a NEW source, the DC bus and VFD transistors. So even if the VFD is putting out 60Hz, the sine waves of the outputs are not in sync with the sine waves of the input, so they no longer have the same cancelling effects. The same is true for multiple drive output cables, because if you have drives, you not only have outputs that are not in sync, but are at different frequencies altogether. The result is generally unpredictable, it CAN be devastating in that very high voltages get induced onto the cables and this can GREATLY exacerbate any reflected wave issues as well, which leads to insulation failure in the motors and/or the cables themselves. I witnessed one installation on a 250HP inverter duty motor that failed, it turned out that they doubled back on the cable tray putting the output cables with the input cables for just 25ft of the run, and the voltage spikes at the motor exceeded 2500V so even though the motor was made with 1600V insulation, it failed anyway. Or the cables can coexist fine for years with no noticeable effects. There is not way to predict, so the best policy is to avoid it all the time.

In cable tray if you can't avoid it, you bundle input cables together and output cables together (trefoil them) then separate the bundles by distances according to the current / field strength, upward of 3 feet, but most people using cable tray do so to AVOID having to mess with cable routing. Using shielded VFD cable helps with not having to worry about it as well, which is one reason people do that too. Whenever someone tells me that shielded VFD cable is too expensive, I respond with "Compared to what?" because doing NOTHING is expensive too, usually MORE expensive when you count down time costs.

 

[Aleman]

Thanks guys for confirming that this is a bad practice. We have an installation that was done this way. I was a little pissed when
I saw the job but it was another guy and it's done. Oh well. Hopefully we beat the odds and it causes no problems. Jraef you are
a walking encyclopedia of knowledge. I always enjoy reading your posts and have learned a lot from you.

 

[Jraef]

My wifey calls me a "fountain of useless information", but thanks.

When people ask me how I came to know all this stuff, I always say "One mistake at a time, a few of them more than once." :slaphead::slaphead::slaphead:

On this particular subject, the very first time I installed a cabinet with multiple VFDs (8), I ran one one 4" RGC out of the box to the gravel pile feeders I was working on and at each feeder, I put in a J-box and split off the one motor connection for that unit, then continued on the run to the next. So the longest run ended up about 1800' from the control tower.

Out of 8 drives, I lost 12 within the first month, meaning I replaced and RE-killed 4 of them, before asking for help. I was lucky enough to get connected to a drive guru at that time (1986 I think) who explained it to me. Never made that mistake again.

 

[don_resqcapt19]

We never mixed line and load or multiple motors, but we would pull the 120 volt motor control circuit in with motor conductors....until we had issues with the motors not stopping when the controls told them too. The induced current in the control circuit conductors would keep the motor running when the PLC was telling it to stop. Now we pull the motor conductors in a raceway by themselves.

 

[Jraef]

We never mixed line and load or multiple motors, but we would pull the 120 volt motor control circuit in with motor conductors....until we had issues with the motors not stopping when the controls told them too. The induced current in the control circuit conductors would keep the motor running when the PLC was telling it to stop. Now we pull the motor conductors in a raceway by themselves.

Yep, another painful lesson for me too. Induced voltage was not enough to make a PLC output go high, but it was enough to KEEP it high when it was supposed to go low. I had to run a new conduit which meant digging a trench in a newly sodded baseball field because of that...

 

[sameguy]

Three pipe system.
Had an enormous fight about this years ago, I ended up calling the shop to have the "know it all" removed from my job to one he was better suited to; lighting retrofit if I remember correctly! )

 

[don_resqcapt19]

Yep, another painful lesson for me too. Induced voltage was not enough to make a PLC output go high, but it was enough to KEEP it high when it was supposed to go low. I had to run a new conduit which meant digging a trench in a newly sodded baseball field because of that...

The first time it happened we were done with all of the wiring so we added transformer pilot lights on the control circuit as additional load and solved the problem that way. New projects after that were run with a power conduit and a control conduit.

 

[RumRunner]

Whenever you run multiple 3 phase conductors in the same raceway, there is a potential for the expansion and contraction of one conductor's magnetic fields to induce a current on an adjacent conductor. But when all three phases come from the same source, even all the way back to the utility generator, these effects will cancel each other out.

But the output conductors of a VFD are being created from a NEW source, the DC bus and VFD transistors. So even if the VFD is putting out 60Hz, the sine waves of the outputs are not in sync with the sine waves of the input, so they no longer have the same cancelling effects. The same is true for multiple drive output cables, because if you have drives, you not only have outputs that are not in sync, but are at different frequencies altogether. The result is generally unpredictable, it CAN be devastating in that very high voltages get induced onto the cables and this can GREATLY exacerbate any reflected wave issues as well, which leads to insulation failure in the motors and/or the cables themselves. I witnessed one installation on a 250HP inverter duty motor that failed, it turned out that they doubled back on the cable tray putting the output cables with the input cables for just 25ft of the run, and the voltage spikes at the motor exceeded 2500V so even though the motor was made with 1600V insulation, it failed anyway. Or the cables can coexist fine for years with no noticeable effects. There is not way to predict, so the best policy is to avoid it all the time.

In cable tray if you can't avoid it, you bundle input cables together and output cables together (trefoil them) then separate the bundles by distances according to the current / field strength, upward of 3 feet, but most people using cable tray do so to AVOID having to mess with cable routing. Using shielded VFD cable helps with not having to worry about it as well, which is one reason people do that too. Whenever someone tells me that shielded VFD cable is too expensive, I respond with "Compared to what?" because doing NOTHING is expensive too, usually MORE expensive when you count down time costs.

In the early years of VFD (PWM) this has been a concern. However, VFDs at the time, were mostly limited to 5HP. Power distribution engineers in heavy industrial application had always made it a point to provide feeders (load side of magnetic starters) dedicated raceways, trays and wireways for motors.
I had spent some time in design of power distribution for industrial machinery that involved diverse sizes of motors. I have never (as well as others) spec'd running both load and line side in same conduit.
I agree that varying output of VFDs caused by induced voltage is a possibility. Since I've never designed a system with load and line in one conduit, I've not seen burned motors caused by this.

Even if this were a possibility, voltage spikes on the load are monitored by the control logic in the VFD controller. It may help to understand how modern VFD (PWM) work. . . it basically adjust the resultant wave form based on what the demand requires by switching to line voltage when output is low and then switches to the modified wave form when it sees a high (voltage output).

If the controller sees a voltage drop or unacceptable rise on the output it triggers the logic- control- switching feature . This rapid switching happens in nano seconds.

This design criteria makes for a more efficient system compared to always-on older design and less heat build up.