Contactor before and or after VFD

We are currently involved in a safety upgrade project at the plant that I work in. What this project involves is stopping drives and/or motors. I have the current NFPA 79 book and there is a bit of interpreting to do this correctly.

For an E-stop condition the code wants me to remove all power from the "actuator" i.e. motor. The code does allow me to have a "controlled" (ramped) stop before this occurs. After the controlled stopped, the code specifies redundant contactors after the drive but before the motor. Another scenario is to use one of the new style drives that has a IGBT inhibit circuit built in. This will allow for a Cat. 1 stop (controlled from a safety relay) and only require the need of one contactor. This is all to meet current E-stop standards. A thought we are considering is the use of a large (50 A) safety contactor or two and break power before all the drives. E-stops are seldom hit during normal working conditions. The advantage to this would be a faster implementation and less outlay.

AB had a pdf that I read on this and one of the downfalls mentioned was shortened drive life. They said it would be acceptable if it were infrequent. They also mentioned the use of a contactor both before and after the VFD. I suppose this would eliminate the problem with regen.

For a standard stop (not e-stop) such as opening a door or breaking a light curtain, using a drive with the built in IGBT interrupter meets code. If we went that route we would be looking at 50-70k for the lines.

One of our providers for safety equip. is on the ANSI board and has had me using safety relays inhibiting my low voltage to the drive. If the safety relay drops out the drive no longer has a run command or any control voltage (dc) to the control section of the drive. This has worked out quite well, but may not necessarily work to the code 100%. Has anyone seen this before?

The pair of us that are working on this project have had a couple vendors in and seem to get differing interpretations. If we continue as is by running control voltage through safety relays we can have this done for a reasonable amount of money. If we do have to drop power from the final actuator we are considering dropping power before the drive, or before and after the drives.

Any advise or experience on this? Thanks

Luke

If you are going to have a contactor between the VFD and motor, you will need to shut off the VFD's output before opening the contactor, otherwise there is risk of damage to the VFD.

The easiest way to achieve this is with a contactor that has a set of aux contacts that opens a split second before the contactor itself opens (early break). The aux contacts would wire to the VFD's e-stop or external fault input to instantly shut off the VFD's output prior to opening the contactor. This allows you to still use the regular ramp-to-stop function during normal operation.

What I don't know is what level of code it meets, if any.

John

Thanks John. What damage do we risk to the drives? Our plant has about 6 power outages a year. We have around 500 VFD's in house. For each outage we can almost guarantee failure of about 6 drives in one cabinet. It is closest to the MSB's and seems that we may be dumping the "ripple" or regen back top this one. If we have a one leg on a transformer drop of or a ground failure -pack your lunch -100's of failures.

The lines we are working on have never seemed to be effected from normal power outages. I'm sure these are limiting life span of the drives, but what other failures will we expect if we go this route? We have even considered putting some of the old DC drives back in to avoid future troubles. Uf-daa!


NFPA has 3 levels of stops.

Category 0 is an uncontrolled stop by immediately removing power to the machine actuators.

Category 1 is a controlled stop with power to the actuators available to achieve the stop and then remove power from the actuators.

Category 2 is a controlled stop with power left available to the actuators.

Category 0 and 1 are for an E-stop. For safety circuits there are time delayed relays that remove control power from the drive or turn of the IGBTs. This would allow the actuator (motor) to ramp for ~ 1 sec and then the contactor drops out. This meets a cat 1 rating.

We did a safety assessment awhile back to recognize risks on the machine. Implementation of these problem areas with light curtains or door safety switches has been a bit of a trouble. What I have been doing is a cat 2 stop. This is a controlled stop with power left to the motor. As we looked closer at the code (NFPA 79) we are realizing that the OEM did not meet the requirements for an E-stop by removing ALL power from the actuator (outputs, motors, etc.) OUr simple solution may be to drop everything in the event of an E-stop. If time and money were no issue we would opt for a drive with the IGBT safety feature and a contactor after the drive. Panel space is now a possible concern as well.

My co-worker and I are planning on submitting a list of options to management and from there corparate to make the final decision.

We did have a safety inspector/consultant in last
week to discuss some of these issues. He is a former OSHA inspector and is familiar with the various codes. His interpretation was that dropping the control voltage from the drive was fine for our door and light curtain safety circuit. The problem we are now faced with is updating and meeting E-stop functions. He said that even if it were an old machine, a plant is required to keep all equipment up to current regulations -no grandfathering. Knob-and-tube in my house was fine because that is residential, outdated safety circuits have an effect on work place safety and this needs to be kept current.


Any one else out there working on machine safety?


Luke

Luke -

You need to consider several additional factors when implementing a safety system. If the drive is operating a machine with significant inertia (i.e. on power loss it will continue on its merry way and smash into something or continue rotating for long enough that a person can get in contact with it) . . . you need to take measures to remove the kinetic energy from the system so that it does not pose a threat of injury.

We accomplish this as you stated above. Upon E-Stop condition the drives are programmed to aggressively brake the machine to a stop. A safety relay has a time-delayed output that then drops out the drive enable (safety rated input on Control Techniques drives that prevents voltage from going out on the motor leads). At this time, disk brakes on the machine automatically are applied as well - but in most cases the machine has already stopped - simultaneously - a safety rated contact drops out 480V 3-phase power to the drives. This has been specified to us by our customer in this fashion as determined by their safety committee. It adds considerable cost to do this - especially when you consider all the dual contact safety rated door switches with electronic solenoid locks.

One area of confusion with this is that these drives (also AB PF series drives that are equipped with this kind of output defeating safety input) are rated as CAT3 which is typically used in Europe / NFPA specifically states CAT0 or CAT1 safety systems - be aware of the differences. Using this input alone does not comply with the NFPA as far as I can tell.

Guru, glad you posted. I'll address the area of confusion at the end of your post. Yes, when looking at the PF series they do use a Cat 3 rating(this is a safety rating). Depending on how implementation is handled, it also gets a stop cat rating 0 or 1.

As we have delved into this pool of fun, I have started to read the ANSI B11-standards for machine safety (I think that's the one) and it refers to NFPA79. My NFPA79 2007 arrived just last week. The safety category looks like it comes from ISO 13849-1/EN 954-1. I don't have that one. We did find it in the Banner catalog though. A cat 3 or 4 rating has to do with frequency of exposer and specs a control reliable system. There is a lot for a maintenance guy to absorb and understand. NFPA is very clear that power needs to be completely removed from the circuit for an E-Stop.

Our interpretation is that a guard door or curtain is NOT an E-Stop, but rather a normal work function. We are fortunate that our equipment does not have much inertia, and several axis have brakes on the drive train. Even without the brake a drift of only a couple inches will occur. I watched that one vid you posted of the paper mill a few years ago -nothing even close to that. That was cool by the way.

I work for a company with a couple divisions and came across a spec that states our equipment should meet current code -like the ones mentioned above. Had this spec been passed on and the OEM held to it, I would be worrying about something else.

We are removing light curtains from a portion of the machine and installing hinged doors instead. Because of the inherent troubles with keyed safety switches we have been using a fiber optic system with great success. One reason for the doors is to get the operator to stop the process in a controlled fashion. They have been in the habit for years to just put their arm in the curtain without looking at the "full picture" of what is going on when they stop the process. We want them to now get used to the idea of using a pause switch before opening the door. I would like to put a solenoid on the door to keep it latched rather than just pulling it open. I have not been able to find anything other than Hoffman cabinet latches or the same problematic safety switches with solenoid latch. Any ideas on what may be available? I do have about a hundred short stroke Bimbas that I may be able to use as a latch. Any ideas on this? Have there been any solenoid switches that have proved not to be finicky and robust?

Thanks for the advice

"What damage do we risk to the drives?"

Opening the output of a VFD while it is still running can cause damage to the output transistors, due to the arcing across the contacts in the contactor while they are spreading apart on their way to being fully open. The voltage level of these arcs can well exceed the rating of the transistors.

"For each outage we can almost guarantee failure of about 6 drives in one cabinet."

Do the VFDs have line reactors in front of them? If not you should look into the cost of adding them vs the cost of replacing the VFDs.

John

Luke with regard to the VFD failures, what is the main components to fail? I would be guesing input rectifiers, if so then excessive dV/dT on the input would be most likely, J Lauffers suggestion of line reactors may be worth considering especially on the units close to the main switchboard. If you have line reactors then some form of transient suppressor after the line reactor may also be needed.

Re: output contactors.
The reason why some VFD manufacturers use the term "infrequently" when stating it can be done, is because the damage to the transistors is incremental. Small areas of the silicon become damaged from the contact arcing as John mentioned. When the damaged areas reach a critical volume compared to the good areas, failure occurs. But until them, you may never know how close you are; it's kind of all or nothing. So if you think about it, it is in the VFD manufacturer's best interest to allow you to believe it's OK to do it, it starts you on the path of spending more money with them later!

Re: safety issues;
It's a big muddled mess right now, as you have observed. In the IEC world, these issues are spelled out very clearly and in detail. In NFPA world, they are comparatively vague and open ended. Yet IEC rules tend to be self-regulating, whereas NFPA 79 is an enforceable code which can bring your project or machinery to a grinding halt.

I'm supposed to be becoming an expert in safety relays and systems now, but I don't yet fully understand it; it's going to take me years to wade through all of it. But one thing I am sure of, the NFPA 79 categories are NOT the same as the SIL Categories in IEC speak, so don't make the mistake of allowing vendor salesmen to lead you astray by attempting to mix them freely. Read your NFPA 79 and follow what it says to the best of your ability. That is the only thing that counts from an inspection standpoint. As I see it, those safety rated VFD disable inputs are NOT going to satisfy the NFPA 79 rules, but they do provide a good way to get there by virtue of having the VFD power output turned off before the contactor opens.

Thanks for the additional input.

In regards to the drives that typically let out the smoke -it is only two enclosures in the plant that do it. Since we have done our Arc flash assessments per NFPA 70E, these are some of the closest panels to the main switch gear. The advantage of resistance in the wiring feeding these panels is not there. They usually go out after lightning hits a pole in the area, causing us to go down. The suspicion is that since every thing is in a state of regen, that these may be the lucky ones to eat some of it. It has been awhile since I've had the pleasure of replacing them and honestly can't remember what they faulted for.

Management knows about it, but has made no serious push for a fix. We did have an instance where a transformer lost a leg causing the MSB to trip and take out just about every thing that it was feeding. Anything that had an MOV went toes up. That was a very long night. I took ~50 drives out of one line and replaced all the drives on the line that needed the production. This included an HMI or two.

Jraef, your not kidding about NFPA not being clear-cut. I don't have the '07 79 book at home, but it does allow for use of these types of drives for a cat 0 and 1 stop up to a safety cat 3. It IS NOT acceptable for an E-stop circuit. The book makes it pretty clear that for an E-stop, redundant contactors must be used (along with all the control reliability stuff that monitors), or one of these drives and one contactor.

As for my partner and I, as we move forward more questions come up than clear cut answers. There is always the -"what ifs?". Our current plan is to use guard doors with locking safety switches.

In the past if an operator wanted to stop the process, he simply moved his hand into the light curtain. Now they will have to push a "pause" button, wait for the cycle to complete before the door unlocks, and then open the door. If the door is jerked open the safety switch will fail and the area will stop. There have been some logic changes to the process that will hopefully reduce the exposure to the area.

What we will do for "phase one" is run the control voltage from the VFDs or DC drives through the safety relays. This should give us a stop category 2. Since a gaurd door opening is not an E-stop we will not have to remove power from the drive (this is an interpretation since drives don't start on there own with out a run command or an enable).

For the E-stop we will have to install contactors. Our plan is to put these on the buss feeding the drives. E-stops are not pushed very frequently on a running line -but if it's an emergency we want them to work!

NFPA does allow for a short time delay in shutting the drive down, per the example of the "safe off" drive. The safety module that I ordered for the door switches has time delayed outputs available for .5, and 1 sec duration. Most of our deccel times on our drives are around 30 ms so this should work fine for stopping the axis, and then opening the door. Any thoughts?

A side note:
I have been working upgrading these palletizers since the first of the year. Getting answers from the corporate level and even getting managers within the plant to understand and help seek answers has been a bit difficult. We had this accident occur on a portion of the machine that was not on the list for improvements. This accelerated getting a fix. My observation is that after an accident it is almost a "hands-off" approach from the managers. There was containment, lock down of the PLCs to two techs, moneys approved, but no-one supervising what will go in place and will it meet code.

I expressed my frustration to a manager yesterday and a couple hours later took my 30 lashes for saying such a thing. My drum since the first of the year is that I am not an expert and need some guidance and feedback that what is being done is correct. There are always a hundred ways to do something but when it comes to safety, it needs to be done right. I don't want to make a decision that I feel is sound, but turns into an accident a year later. Anyway I expressed this to the three bosses and they said that me and my partner are the experts. After we make a decision it will go to corporate for their sign-off. More hands-off. I'm sure liability is the issue.

If I were the manager, I would want to make sure that if my "expert" were raising his hand for some help and input I would listen. Maybe I stirred the pot, or maybe I just bought a ticket.

I do like coming out to this board at night and getting feedback from everyone here. I always feel that there is good insight. Thanks.