Need for shunt trip breaker with soft starter? 5

[bentov]

I notice specs for soft start combination controllers usually call for shunt trip breakers to isolate the motor in case of a soft start failure. How is that soft start failure (shorted SCR's, most likely) any different than, say, welded contacts on a regular starter caused by fault current? Why is an extra level of Short Circuit Protection needed?

So, if we retrofit a failed ARV or wye-delta controller with a soft starter, is there some reason we need to incur the extra cost of a shunt trip or fault contactor?

 

[jraef]

I notice specs for soft start combination controllers usually call for shunt trip breakers to isolate the motor in case of a soft start failure. How is that soft start failure (shorted SCR's, most likely) any different than, say, welded contacts on a regular starter caused by fault current?

Good observation. It really isn't any different as far as the effect goes, but the likelihood is slightly higher especially if you are in an area prone to lightning hits. SCRs can "self commutate" meaning turn on by themselves, however breifly, when there is a high enough dv/dt (delta voltage / delta time) event on the line feeding them. For that reason, soft starters usually have "snubber networks" on each pole, either an RC circuit or an MOV, or both. That will take care of 99% of the dv/dt events the starter will see (although MOVs tend to fail after the first one). But a lightning strike somewhere down the line is sometimes enough to overwhelm that protection and cause self commutation. If, during that brief conduction, there is another insulation failure downstream (which is often the case), the resulting high current can then damage or "short" the SCRs and then they cannot be turned off. That is the point where it becomes the same as a welded contact.

Why is an extra level of Short Circuit Protection needed?

Needed? That's a matter of opinion. High speed "semiconductor fuses" can typically act faster than regular fuses, so in that high current (di/dt) event, a fuse that can interrupt the current flow fast enough may be able to prevent that SCR damage. Fuses are cheaper than SCRs. But certainly there are plenty of people who feel that the fuses are superfluous and go with just circuit breakers.

So, if we retrofit a failed ARV or wye-delta controller with a soft starter, is there some reason we need to incur the extra cost of a shunt trip or fault contactor?

Need to? No. Prudent? Yes. You don't "need to" buy medical insurance either, but the first time you needed it and didn't have it, you are going to wish you did. If your installation is attended and someone is there to observe a motor single phasing because of multiple shorted SCRsw, AND they have the sense to open the circuit breaker, then you don't "need" the shunt trip. But if you have a remote pump station where nobody ios going to be around for weeks, it might be cheap insurance. Compare the cost of that shunt trip to that of a motor rewind.


That said, I have installed a LOT of soft starters, mostly in the Western US where lightning is not that big of a problem. I would rarely recommend isolation contactors or shunt trips and I rarely saw any problems. I had co-workers in Central Florida however who had lightning related failures almost every day. They recommended them every time.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[Marke]

The SCRs used are typically rated at 2.5 - 3 times the line voltage.
There are many high voltage transients on the supply, typically very short, but high in amplitude.
If a high voltage appears across the SCR, it can cause the SCR to begin to turn ON. This is typically known as overhead triggering and under this condition, only a portion of the die is turned on initially and it takes time for the charge to spread across the junction. As a consequence, the rate of rise of current must be limited relative to the rated di/dt when the gate is driven. If the overhead triggering results in a high rate of rise in current, a portion of the die will be overloaded and the SCR will fail as a short circuit.
Alternatively, the voltage can cause tracking around the outer edge of the die.

The bottom line is that the SCRs are far more vulnerable to transients than a mechanical set of contacts, and a breakdown is permanent.

Where a supply is commonly subject to high voltage transients, the probability of failure is greatly reduced by adding mechanical isolation when current is not flowing so that the SCRs spend little time blocking line voltage plus transients. Alternatively, the probability of failure can be reduced by increasing the voltage rating of the SCRs at great cost. The majority of supplies are adequately covered with the normal SCR selection.

Best regards,

Mark Empson
L M Photonics Ltd

 

[bentov]

Thanks for your replies. The (2 phase with 2 pole internal bypass contactor) soft starters we use say they provide 3 pole Class 10 overload protection, which I presume means the typical phase current unbalance response. I guess the difference is what constitutes "off" then, right? If one (of two) SCR fails, controller responds correctly by opening the remaining one, but L3 is a pass through, I guess the winding is history unless the SCP provides protection.

Our applications are ag pumps in remote areas, definitely exposed to power transients. Retrofitting an existing breaker (or more common fusible disconnect) is problematic, so maybe the most cost effective extra level of protection would be a fault contactor, rely on the soft starter fault relay to operate it - must it be full sized though, since it's purpose is to open fault current?

I considered the semiconductor fuses, also a problem with retrofitting though. If the most likely cause of failure is transient damage to the SCR's, maybe I should focus on that? I presume that can happen any time the power is present, whether or not the motor is running. Maybe a line reactor? Heavy and bulky though, and a waste of energy . . . could use a TVSS with fault relay to disable the soft starter, wouldn't save the motor though if the transient occurred while it was running . . .

 

[Marke]

Hi bentov

There are three major causes of SCR failure, one is excess current while the unit is running and that can be due to the magnitude of the current of the rate of rise of the current, and another is OFF state over voltage which once again can be related to an overvoltage (surge) or a rate of change of voltage (transient) and the last is thermal where the junction temperature has been allowed to operate above 125 degrees C for too long.

Failure during RUN is invariably a current problem which can be due to a voltage surge with high current going to the motor.

Failure during the OFF state is because the SCRs are blocking an excess voltage and this can be transient or surge in nature.

If the starter is bypassed, it is almost immune to current related problems during run and the current will bypass the SCRs and flow through the contactor.

So your major problems are in the OFF state and thermal.
The thermal situation is related to SCR (model) sizing and enclosure ventilation, so provided that both of these are correctly addressed, there should be no issues.

In the off state with a three pole device connected in series with the lines to the motor, the transient or surge must be high enough to cause a breakdown of two SCRs in series. For example where 1600V SCRs are used, the voltage transient must be over 3000V assuming reasonable balance in the voltage blocking of the SCRs.
In the case of a two pole device, the voltage transient must be above 1600V to cause a breakdown so these devices are much more prone to damage from transients than three pole devices.
If the motor protection operates and turns OFF the starter, one phase is left connected to the motor and if there is a leakage to ground, it will continue to draw current until fuses fail.

With a two pole device, you should use a contactor to disconnect the starter when not operating.

The regulations in some territories require that you disconnect all three live conductors from the motor.

Best regards,

Mark Empson
L M Photonics Ltd

 

[jraef]

bentov,
With 2-phase soft starters, having an isolation contactor is a good idea. Again, not absolutely necessary, but prudent nonetheless.

When you have SCRs in all 3 phases and you get one shorted SCR, it is a problem when ramping but is not in and of itself an immediate danger to your motor. But if you have 2 shorted SCRs in opposing phases, now you have an unrestricted power path through the motor winding; essentially like having 2 welded contacts. When you have a 2-phase soft starter it's like having one set of SCRs shorted all the time, so all you need is for one more to short and there's your path. An isolation contactor eliminates that possibility. I prefer that method over the shunt trip in the breaker, because if you are going to go to the trouble, you also get the line isolation AGAINST damaging transients when you use a contactor.

PS:
Got your email, go ahead and contact me at the work email address I provided for you.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[LionelHutz]

We ship I'd guess around 25,000 SCR's a year. We might replace maybe a dozen a year under warranty and maybe a 2-3 dozen a year total. A good soft-starter with good self-protection features does a good job of protecting itself.

In your case, I'd expect the most likely cause of SCR failure will be the starter not protecting the SCR's against things like overload, overcurrent, shorts, too many starts, etc. When we changed from the dumb open loop voltage starters to the intelligent closed loop starters the number of SCR failures we saw dropped very significantly.

Are you actually having failures of these 2-phase units or are you just looking to prevent failures? Are these little motors, say less than 10hp? The only 2-phase controllers I can think of at the moment are the small, cheaply build units for use with little motors.

Have you tried a 3-phase starter? The difference in how smoothly a 6-SCR starter works is very apparent compared to a 4-SCR unit. I'd never use 4-SCR, 2-phase control on one of my motors, it just does not work right.

I would not say that a 3-phase unit will break down at a higher voltage compared to a 2-phase unit. When dealing with transients, the motor cable capacitance and motor capacitance very much come into effect. Most likely, this capacitance will keep the motor side of any starter near 0V during a transient event.

In the end though, if you must stick with these 2-phase controllers then an isolation contactor is required. It must be HP sized to the motor.

 

[vinnygalbo]

The top 2 soft starts that I've used are ABB & Saftronics. I think Saftronics changed their name to Emerson/Control Techniques now.

http://www.clrwtr.com/ABB-PST-SoftStarters.htm

http://www.clrwtr.com/Saftronics-Soft-Starters-SoftStart.htm

I think they all have a built-in bypass contactor now.

Motortronics is a small company that specializes in soft starts. I don't have much experience with them. But somebody else may have had good experience with them.

http://www.ctiautomation.net/Motortronics.htm

 

[bentov]

Hello everyone,

Thanks again for your replies, much to think about here. The 2 phase starters are available up to 300hp (@480V), seem to function smoothly on centrifugal pump loads, certainly the least expensive way to achieve the utility required reduced current starting. We have a fair number out operating, have yet to see an SCR failure except on a 3 phase unit (400hp).

The question of risk to motor windings is of course the main thing. We are a rewind shop, have seen a fair number of failed windings connected to soft starters, typically groundwall insulation failures in otherwise OK looking windings (i.e. no evidence of phase loss, overheating, contamination, etc.) of new-from-the-factory motors. So far none have coincided with a soft starter failure. At the one 400hp site, at least two windings failed (though we didn't get to examine them, not sure of failure mode). Some months later that particular soft starter failed ("switching element failure", found one shorted SCR out of the set of three). When the soft starter was replaced, the motor was still OK - few months later that (factory new) motor winding failed to ground as described above. The soft starter survived, accurately recorded the ground fault in it's event log, working fine now with a replacement motor.

We have to wonder if something's going on here, but hesitate to point the finger at the soft starters because we haven't seen the failure mode we would expect if they were the culprit (failed soft starter with one or more shorted SCRs plus a burned winding with the classic telltale phase loss pattern). We also haven't had a failure with one of our rewinds, but we do use "inverter duty, spike resistant" magnet wire only so that could be a factor. All the motors were from a manufacturer that acknowledged "problems", wouldn't specify except to say they were resolved, always honored the warranties. We clearly indicated use of soft starters in our warranty report, got no reply or comment about that, could be they just had a run of faulty insulation in their manufacturing process.

In any case, now I'm thinking about SCR "overhead triggering" per Marke and "self-commutation" per jraef - is it possible that such events (when starter is in the OFF state, connected to the motor, not isolated from mains supply by a contactor) are letting through transients with enough energy to breach motor winding groundwall insulation, even though no damage (thus no fault or event log data) to the SCR's occurs? Indeed, users typically report that everything was "fine" for last use, motor "just blew" on startup at next demand, days or weeks later. Likewise, the 400hp soft starter was said to be OK at last use, failed on next attempt ("just wouldn't do anything", fault code on screen when I arrived, no reset possible). We always have to take user reports with a grain of salt though, as no one (especially a humble irrigator) wants to risk being blamed for anything, so our conclusion that things are happening in the OFF state may be a bit tenuous. It's possible a rash of defective motors simply wore down the soft starter, shunt trip breaker might've saved it though not the winding . . .

If there is OFF state leakage damage though, a shunt trip breaker won't help, need the isolation contactor instead (no small expense at 400hp, whether "full hp" or bypass sizing is used). So I'm thinking again about the Transient Voltage Surge Suppressor route (not to trigger a contactor, just to absorb transients), would appreciate any comments or suggestions about experience with TVSS's in this kind of application, and also of course about the whole scenario . . .

 

[Marke]

Hello bentov

Yes, I think that you on to a something. The increased risk of operating the motor without an isolation contactor certainly puts the SCRs under a much increased risk of failure and this has been proven in the field.
We have seen installations that have SCR failures every few months without a contactor, but 5 - 10 years once the contactor is fitted.
The transient that is potentially damaging the SCRs is also potentially damaging the motor insulation.
Effectively, one could consider that the SCRs block the steady state voltages (assuming a three pole device), but not necessarily the transients which would be blocked by the contactor.
Improved insulation ratings will overcome some of these failures.
Perhaps, the insulation rating of the motor should be higher where soft starters are employed without line contactors.

Best regards,

Mark Empson
L M Photonics Ltd

 

[bentov]

Hello Marke,

I can almost visualize those nasty transients buzzing around, eroding our groundwall insulation during OFF state periods. We have quite a few fairly high HP VFD installations in the same neighborhood with no trouble, but always use line reactors on those. I can see how with the soft starts, when the motor's running, the large inductive "sump" absorbs those transients, greatest vulnerability may in fact be during OFF times via leakage through the SCR's.

Here is a spec sheet for one of the three thyristors in that 400hp starter:

http://pdf1.alldatasheet.com/datasheet-pdf/view/217372/SEMIKRON/SKKQ850.html

A typical thyristor in one of the smaller two phase units:

http://pdf1.alldatasheet.com/datasheet-pdf/view/217668/SEMIKRON/SKKQ800.html

Looking at specs for another manufacturers' combination enclosed soft starter (350hp 480V) I notice in addition to the shunt trip breaker they specify a "Secondary surge arrestor". The part number isn't given, but I suspect it is this one:

http://ecatalog.squared.com/pubs/El...dary Surge Arresters/Hardwired/1311HO0101.pdf

Since neither manufacturer appears to recommend an isolation contactor with their combination package, I'm in a peculiar spot if I choose to insist that one is necessary (given the expense). I'd be very interested in your opinion of how that particular surge arrestor, connected in parallel to the line supply of those particular thyristors (without isolation contactors), might behave in our application, given the dv/dt specs. Those can be had for a little over $100, easy to install, considering them as a form of "cheap insurance". I'm guessing they are the multi-parallel MOV design, so while "consumeable" by nature they might nevertheless have some reasonable service life, depending on circumstances . . .

 

[waross]

Hi friends. I have been following this thread with a little confusion.
I know a shunt trip as an added option to a molded case circuit breaker. (Yes, I know the meaning is a little different for large power breakers.)
Normally, under North American codes, I expect some form of disconnect ahead of any motor starter, be it soft start, VFD or DOL.
I have seen Soft Starts with a shunt trip option added to a manual bypass breaker so that the soft start could trip the bypass breaker if it detected a motor overload.
Is it common in some places to install a soft start without a supply breaker?
Thanks for your patience.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

Honestly, I would guess that these motors are purchased because they are the cheapest possible motor available (much like it appears the bottom line is driving the choice of soft-starters) and they are failing because they are newly manufactured junk. The fact that your rewinds stand-up fine backs this up. I run into the exact failure you describe most often with so called "VFD Rated" motors failing on a VFD.

If the SCR's were being exposed to high voltage transients then you would be seeing SCR failures.

The surge arrestor is a good idea, especially if the use of arrestors upstream is not known. You would likely see them fail every now and then if there is a transient voltage problem.

Bill - There is always a breaker, it's just a question of installing a shunt trip coil in the breaker or not.

 

[waross]

Thanks Lionel.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Marke]

The reason for the shunt trip breaker, is so that the motor can be automatically disconnected if the protection operates.
In the case of a two phase controller, opening the SCRs only opens two phases and will not interrupt an earth fault.
If the SCRs are damaged, then you need an electromechanical switch that can disconnect the motor under fault.

My preference is for a line contactor to be fitted in all soft starter installations, but most are installed without contactors or shunt trip breakers.
If the SCRs fail, the motor fails also.

Best regards,

Mark Empson
L M Photonics Ltd

 

[bentov]

Out of many installations, we have seen only one shorted SCR (on the 3 pole 400hp unit), and that failure did NOT coincide with a motor winding failure. We have seen quite a few motor winding failures though, so my questions remain:

Is it possible that insulation damaging transients can leak across the semiconductor junction of typical soft starter thyristors in the OFF state?

If so, would a surge suppressor provide some protection, and how would one select it?

The one I linked to above, for example, lists clamping voltages from 1525 to 3000, depending on lead length and surge current. The thyristor in the soft start says (among many specs) "VRSM 1500" and "VRRM-VDRM 1400". Does that mean transients above those voltage levels will "leak", though maybe not damage the semiconductor if current and duration are low? That's certainly high enough to stress winding insulation (especially if old and dirty on a wet day), maybe make tiny little pinholes that become blast craters when normal startup rolls around.

Should I look for surge suppressors with lower clamping voltages to go with these particular thyristors, make a point in the future of always investigating soft start specs and applying coordinated suppressors?

 

[Marke]

Is it possible that insulation damaging transients can leak across the semiconductor junction of typical soft starter thyristors in the OFF state?

Yes, there is an RC snubber across the SCRs and this will allow some of the transients straight through, and if the transients are sufficiently high in magnitude, they will either cause the SCR to overhead trigger (turn ON) and present the transient fully across the motor winding, or fail and once again subject the motor winding to excess stress.

I am not convinced on the surge suppressor option. It is a shun option and one needs to be very careful of high frequency impedances as it is very possible for the high frequency component of the transient to walk right past the transient suppressor and cause the SCR related issues before operating. The physical connection would be very critical.
I have repaired units that have MOV varistors across the incoming phases and still suffered voltage problems.

we have seen only one shorted SCR (on the 3 pole 400hp unit)

It is common to lose two phases on a three phase controllers.
Over 50% of the units that I see that have failed, lose two phases.

Best regards,

Mark Empson
L M Photonics Ltd

 

[LionelHutz]

Why don't you put a power quality analyzer onto one of the sites and see if there really are transients?

Is it possible that insulation damaging transients can leak across the semiconductor junction of typical soft starter thyristors in the OFF state?

Yes it is possible. Is it the likely cause of these motor failures - I don't think so. I believe you would have SCR failures on a regular basis if high voltage transients were occurring on a regular basis.

Surge type protection is difficult to apply effectively. It's hard to get it to respond quickly enough and to clamp at a low enough voltage. You may also want to investigate surge capacitors installed on the line side of the starter.