Active Front End (AFE) VFD vs. 12-Pulse VFD

[elec_eng]

I have a facility with severe harmonic issues, especially for the generators. The initial harmonic analysis indicates that harmonics comes from the 6-pulse VFDs. All of VFDs in the facility are old 6-pulse VFDs and they are considering to replace them with either 12-pulse VFDs or AFE VFDs.

I have not used AFE VFDs before but based on my quick research they seem to be very much like the modern UPS with IGBT front end rectifiers. Is that correct?

What would be the pros and cons for the AFE VFDs as opposed to 12- (or 18 )pulse VFDs?

 

[wsbeih]

I have a facility with severe harmonic issues, especially for the generators. The initial harmonic analysis indicates that harmonics comes from the 6-pulse VFDs. All of VFDs in the facility are old 6-pulse VFDs and they are considering to replace them with either 12-pulse VFDs or AFE VFDs.

I have not used AFE VFDs before but based on my quick research they seem to be very much like the modern UPS with IGBT front end rectifiers. Is that correct?

What would be the pros and cons for the AFE VFDs as opposed to 12- (or 18 )pulse VFDs?

From harmonic mitigation standing, they go in this order 12 pulse, 18 pulse, then AFE. A typical Pulse Width modulated drive (including 12p & 18p) have
input rectifier, DC bus and IGBT output inverter. In AFE drives, the input rectifier is replaced with another IGBT bridge.
The cost for the AFE is so high. I assume your goal is to meet the THD (total harmonic distortion) per IEEE 519 for voltage and current. Typically a harmonic analysis is needed to determine the optimum drive type you really need.
Some times, you can get typical 6-pulse VFD's and install an passive or active filter added to the main bus you you consider as the PCC (point of common coupling), typically this would be your main Panel/MCC feeding these VFD loads. See for http://www.transcoil.com/home.htm for more information about these type of filters.

 

[Besoeker]

I have a facility with severe harmonic issues, especially for the generators. The initial harmonic analysis indicates that harmonics comes from the 6-pulse VFDs. All of VFDs in the facility are old 6-pulse VFDs and they are considering to replace them with either 12-pulse VFDs or AFE VFDs.

I have not used AFE VFDs before but based on my quick research they seem to be very much like the modern UPS with IGBT front end rectifiers. Is that correct?

What would be the pros and cons for the AFE VFDs as opposed to 12- (or 18 )pulse VFDs?

Difficult without knowing the extent or severity of the problem.
We have supplied 12, 18, and 24 pulse systems. And some AFEs.

Which is best? Depends.
The AFE takes a nearly sinusoudal current but doubles the number of active components so you double the complexity, nearly double the price and, theoretically, half the MTBF. Expensive but will mitigate the harmonics issue. Multi-pulse needs transformers to get the required phase displacement.

A couple of thoughts.
For multi-pulse system you could have one transformer and feed some VFDs from different secondaries. That would avoid the cost of replacing all the VFDs.
Or you could just convert some of the biggest VFDs to AFE.

 

[LMAO]

I have a facility with severe harmonic issues, especially for the generators. The initial harmonic analysis indicates that harmonics comes from the 6-pulse VFDs. All of VFDs in the facility are old 6-pulse VFDs and they are considering to replace them with either 12-pulse VFDs or AFE VFDs.

I have not used AFE VFDs before but based on my quick research they seem to be very much like the modern UPS with IGBT front end rectifiers. Is that correct?

What would be the pros and cons for the AFE VFDs as opposed to 12- (or 18 )pulse VFDs?

Your options are replacing the existing VFDs with AFE (active front end), 12 pulse, 18 pulse or 24 pulse drives. Also, they can use active or passive harmonic filters in parallel with their current drives and reducce the harmonics.

active harmonic filters are expensive, take more space and create some high frequency (2-10kHz) PWM on line; but they eliminate virtually all low frequency harmonics and also regenerate negative power (if any) back to the line.
12 pulse VFDs need phase shift transformers, cheaper than AFE, eliminate 5th and 7th harmonics. 18 pulse and 24 pulse are more expensive and eliminate first 3 and 4 harmonic distortion frequencies respectively.

You can also use active or passive harmonic filter with their existing drives; passive harmonic filter are less expensive but introduce leading power power factor into line. Active harmonic filter is a little more expensive and can eliminate pretty much all harmonic and correct power factor. It is essentially the front half of an AFE (IGBT bank with a huge capacitor bank) that produces 180 degrees shifted harmonic distortion back in line cancel out harmonics produced by VFDs.

 

[Jraef]

Be careful here, not all solutions are created equal. For example some AFE drives cannot operate correctly when fed from a high impedance source such as a backup generator, which in your case is exactly the issue you are looking for help with. One on particular will tell you this, but only in the installation manual, AFTER you have already bought their drive! I know of several people who have been burned by that, because who reads a manual cover to cover before installing something?

Not all of the Active Harmonic Filters are created equal either. I've been using the TCI H5 units on my projects lately (and no, I don't work for them). I've had no issues, other than my own problem in sizing them correctly in the beginning (hint: give them the one-line and let THEM size it for you). What I like about them in particular is that besides correcting harmonics, they also correct the distortion and displacement power factor, which helps a lot in generator installations.

We never use 12 pulse, they cost about the same as 18 pulse but don't perform as well on 60Hz installations. 12 pulse is more popular in the 50Hz world (you don't say where you are). The only problem I have with 18 pulse is the physical size issue. Some mfrs try to mitigate that by placing the transformers under the VFD portion, but I find that over time, the heat from those transformers shortens the life of the drive electronics. Older Robicon 18 pulse drives were famous for being smaller like that but this year, I have been replacing older Robicon drives right and left because the drive portions are failing but since being bought by Siemens, they are no longer supported. But if they put the transformer in a cabinet next to the drive, the floor space goes up dramatically. Not bad for a one-off installation, but it gets extreme when you have a pump station with multiple drives in the same room.

 

[elec_eng]

From harmonic mitigation standing, they go in this order 12 pulse, 18 pulse, then AFE. A typical Pulse Width modulated drive (including 12p & 18p) have
input rectifier, DC bus and IGBT output inverter. In AFE drives, the input rectifier is replaced with another IGBT bridge.
The cost for the AFE is so high. I assume your goal is to meet the THD (total harmonic distortion) per IEEE 519 for voltage and current. Typically a harmonic analysis is needed to determine the optimum drive type you really need.
Some times, you can get typical 6-pulse VFD's and install an passive or active filter added to the main bus you you consider as the PCC (point of common coupling), typically this would be your main Panel/MCC feeding these VFD loads. See for http://www.transcoil.com/home.htm for more information about these type of filters.

Our goal is not necessarily to meet the IEEE 519 rather fix the CB tripping issue at hand due to the harmonics. Of course, meeting the IEEE 519 would be helpful.

 

[elec_eng]

Difficult without knowing the extent or severity of the problem.
We have supplied 12, 18, and 24 pulse systems. And some AFEs.

Which is best? Depends.
The AFE takes a nearly sinusoudal current but doubles the number of active components so you double the complexity, nearly double the price and, theoretically, half the MTBF. Expensive but will mitigate the harmonics issue. Multi-pulse needs transformers to get the required phase displacement.

A couple of thoughts.
For multi-pulse system you could have one transformer and feed some VFDs from different secondaries. That would avoid the cost of replacing all the VFDs.
Or you could just convert some of the biggest VFDs to AFE.

Existing 6-pulse VFDs are old anyway so we don't necessarily have to reuse them. Of course, if they are in good shape, you might consider keeping them though. VFDs are mainly for HVAC and small pumps so they range from 15hp to 100hp. Will AFE have a price advantage in high hp?

 

[elec_eng]

Your options are replacing the existing VFDs with AFE (active front end), 12 pulse, 18 pulse or 24 pulse drives. Also, they can use active or passive harmonic filters in parallel with their current drives and reducce the harmonics.

active harmonic filters are expensive, take more space and create some high frequency (2-10kHz) PWM on line; but they eliminate virtually all low frequency harmonics and also regenerate negative power (if any) back to the line.
12 pulse VFDs need phase shift transformers, cheaper than AFE, eliminate 5th and 7th harmonics. 18 pulse and 24 pulse are more expensive and eliminate first 3 and 4 harmonic distortion frequencies respectively.

You can also use active or passive harmonic filter with their existing drives; passive harmonic filter are less expensive but introduce leading power power factor into line. Active harmonic filter is a little more expensive and can eliminate pretty much all harmonic and correct power factor. It is essentially the front half of an AFE (IGBT bank with a huge capacitor bank) that produces 180 degrees shifted harmonic distortion back in line cancel out harmonics produced by VFDs.

This facility staffs are not familiar with multi-pulse VFDs nor the harmonic filters so I am trying to minimize the new components for them to maintain. From the maintenance standpoint, I would think AFE is the best way to go?? Is AFE really expense compared to 12-pulse? How about 6-pulse with harmonic filters?

 

[elec_eng]

Be careful here, not all solutions are created equal. For example some AFE drives cannot operate correctly when fed from a high impedance source such as a backup generator, which in your case is exactly the issue you are looking for help with.

Jraef, this is an interesting point. Can you elaborate little more? They have issues with both normal and emergency cb tripping due to the harmonics but as you point out, generator side is more often than normal side.

Not all of the Active Harmonic Filters are created equal either. I've been using the TCI H5 units on my projects lately (and no, I don't work for them). I've had no issues, other than my own problem in sizing them correctly in the beginning (hint: give them the one-line and let THEM size it for you). What I like about them in particular is that besides correcting harmonics, they also correct the distortion and displacement power factor, which helps a lot in generator installations.

From your experience, are you recommending the harmonic filters over AFE in smaller hp range VFDs (15hp to 100hp)? I am not a big fan of filters for my own reasons, such as maintenance and bad engineering etc.

We never use 12 pulse, they cost about the same as 18 pulse but don't perform as well on 60Hz installations. 12 pulse is more popular in the 50Hz world (you don't say where you are).

This is interesting point as well. Can you explain why this is the case?

 

[Besoeker]

Existing 6-pulse VFDs are old anyway so we don't necessarily have to reuse them. Of course, if they are in good shape, you might consider keeping them though. VFDs are mainly for HVAC and small pumps so they range from 15hp to 100hp. Will AFE have a price advantage in high hp?

I'm not in your country but I would expect that pricing differences for the different possible solutions are comparable.
AFE is expensive. Multi-pulse requires a transformer so additional space.

Horses for courses. Sometimes it takes a bit of an off the wall approach. A short story relating to one of mine.
We were in competition for a new drive and motor for a pumping station. Harmonics were already on the edge of acceptable because of other 6-p VSDs. Others offered 12-pulse or AFE.
Expensive and there were space constraints.

I offered a different solution. Put in a non-isolating phase shifting transformer and a six-pulse inverter. It was a practical and cost effective.
Something to think about.

 

[Jraef]

I can't elaborate much more on the issue of AFE drives and generators, I just know that it's an issue. Aside from the one mfr that definitely says so in their manual, the other main player in that world will neither confirm nor deny it is a problem, but I have two customers that are having major issues with it. The drives work fine when on utility power, but when they run the backup generators (pump station), the drives just shut themselves down with no fault displayed. In one large pump station project installed last year, it has reportedly progressed to the point of calling lawyers because nobody is getting paid. I don't know what the technical issue is behind it however I do know that some new competitor's products due to be released next year are touting that they know what the issue is and have fixed it. We'll have to wait and see on that.

As to AHFs, yes, they are a better solution for multiple drives, large or small in my opinion. For AFE or 18 Pulse, you are buying a harmonic solution for EACH drive. With an AHF you are buying one unit that covers them all. I did a project last year with 4 x 400HP drives, the 6 pulse drives + an AHF cost about 30% less than the AFE drive solution. Then as it turned out, they would never run more than 2 pumps at a time, so we lowered the size of the AHF and saved them even more.

Cost wise, the issue with an AFE drive is that you are essentially buying two drives for every motor, because an AFE VFD is basically two back-to-back inverters. Yes, the front-end inverter is replacing the diode bridge, so from that standpoint the difference seems to be just the cost of transistors (and their firing circuits) vs. simple diode bridges, but there is more to it than that. Because an AFE drive is capable of putting regenerative energy BACK into the line, an AFE drive must add line loss sensing to so that it shuts down and doesn't kill a Linesman if the utility goes down. Also, there is a lot of necessary filtering that must be added because the transistors are more "delicate" when it comes to line disturbances. So the AFE drive must add line filtering, typically an LCI filter, so if you are not a fan of filters, you are adding a LOT of them. Those added protections result in the overall cost of the AFE drive being roughly 2X the cost of a standard drive.

Then from a reliability standpoint, an AFE adds a significant component count to each and every drive, so your MTBF (Mean Time Between Failures) drops significantly. I don't know what it is however, I just know it has to be lower because that's how the math works. And if the harmonic filter fails, you just don't have a harmonic filter. If an AFE drive fails, you don't have the drive.

 

[LMAO]

This facility staffs are not familiar with multi-pulse VFDs nor the harmonic filters so I am trying to minimize the new components for them to maintain. From the maintenance standpoint, I would think AFE is the best way to go?? Is AFE really expense compared to 12-pulse? How about 6-pulse with harmonic filters?

It is rather difficult for us to figure out best option because, as others have pointed out, there are a lot of factors in play. I can tell you this however: if you have several VFDs running various motors, it'd be a lot cheaper to use a single Active Harmonic Filter (AHF) in parallel with regular 6 pulse diode front end (none AFE) drives.

Or you can replace your drives with all Active Front Ends which are typically larger and more expensive because their advantage is not just lack of harmonics, it is also ability to reverse power and regulate DC voltage to run motors at high voltage, if necessary. Needless to say, reversing power back to gens is not what you want so make sure you disable that if you go with AFE!

Either way, you need to make sure your generators are OK with PWM introduced to the line by either AFE or AHF. Also, both AEF and AHF typically have capacitors in front in form of L-C-L filters.

 

[LMAO]

I can't elaborate much more on the issue of AFE drives and generators, I just know that it's an issue. ......

could it be the 2.5k to 15kHz PWM spikes on the line? After all that's the only difference between diode vs IGBT front end, other than reverse power management which I doubt anyone is dumb enough to use it on gens.

 

[Ingenieur]

Lot of good info

Couple of questions
spectrum/fft?
what % of the total load are vfd's?
how many drives? HP range
what problem is the thd causing?

imo look at iso xfmrs and passive filters first
As noted some side pf benefits
either central, local to loads, or selectively at the most thd contributing loads
or a combo, w/o a single line all speculation

 

[junkhound]

Impedance issue question:

Search for 'middlebrook criteria' for complete explanation and theory). Applies to AFE and all switching power supplies.

Quick and simple:
Control loop response of an AFE is typically in the 1-3 kHz region to control a 20Hz + PWM.
If the AFE boost circuit cannot pull the current from the line (or genset) at that frequency the AFE will be unstable (wont work). Thus, the leakage reactance of the genset becomes the limiting factor in many cases.

 

[Jraef]

Impedance issue question:

Search for 'middlebrook criteria' for complete explanation and theory). Applies to AFE and all switching power supplies.

Quick and simple:
Control loop response of an AFE is typically in the 1-3 kHz region to control a 20Hz + PWM.
If the AFE boost circuit cannot pull the current from the line (or genset) at that frequency the AFE will be unstable (wont work). Thus, the leakage reactance of the genset becomes the limiting factor in many cases.

Wow, excellent! Thanks.

 

[Jraef]

Our goal is not necessarily to meet the IEEE 519 rather fix the CB tripping issue at hand due to the harmonics. Of course, meeting the IEEE 519 would be helpful.

Oh oh, I somehow missed this issue. You are convinced that the harmonics are causing your circuit breakers to trip? How did you come to this conclusion? By eliminating ALL other possibilities? The reason I ask is that although not unheard of, that is fairly rare.

 

[elec_eng]

Oh oh, I somehow missed this issue. You are convinced that the harmonics are causing your circuit breakers to trip? How did you come to this conclusion? By eliminating ALL other possibilities? The reason I ask is that although not unheard of, that is fairly rare.

The owner hired a 3-party to analyze the system and they concluded that harmonic was the issue. I have not seen the report personally and only got involved in providing new VFDs for the facility. Thank you for all your insights.

 

[Besoeker]

For AFE or 18 Pulse, you are buying a harmonic solution for EACH drive.

Not necessarily so.
We did quite a few paoer mills with 12 and 24 pulse systems. The largest, as I recall, is a 4MVA system with 17 variable speed drives. The drives were a mixture of DC, DC 4-quad and some variable frequency inverters.
The system comprised two 2MVA 11kV/520V transformers, four six pulse plain rectifiers, and a 700Vdc busbar distribution system with all the drives hanging off that.

Lowest order harmonic is 23rd and, as you know, harmonic amplitude is roughly proportional to the reciprocal of the harmonic number. The arrangement has benefits on system efficiency but that's another story for another day.

 

[gray.one]

Our goal is not necessarily to meet the IEEE 519 rather fix the CB tripping issue at hand due to the harmonics. Of course, meeting the IEEE 519 would be helpful.

CB tripping may be a result of the GF protection picking up. Remember HF signal sees capacitance to ground as low impedance.