Unexplained SCR Failures (Soft Starter)

[GGOSS]

Hello all,

I have encountered 2 problem sites in which the SCR's within a soft starter have failed. A post mortem of the failed devices strongly suggests overvoltage damage however one would expect other sensitive electronic equipment also connected to the same supply bus would also have suffered, but this has not the case.

As the soft starter is used with line and bypass contactors, the failures could only have occurred in the short period of time when the motor is starting. This is an approximate 8 second window.

Any guidance would be greatly appreciated.

Regards,
GGOSS

 

[jbartos]

Suggestion: There may be several causes possible. E.g. overvoltage, possibly spikes, voltage swell, interaction with other nonlinear load (e.g. harmonics), etc. It may be better to contact the soft starter manufacturer tech support for a right solution or troubleshooting

 

[busbar]

A couple of basic items on the checklist might be verifying that snubbing/transient-suppression components are doing their job, and that correct fuses [if used] are installed.

Insulation checks on load devices and wiring should not be overlooked.

 

[VicP]

Are there power factor correction capacitors connected to the line or load side of the starter?
Capacitors should never be connected directly to a motor which is started with a soft start. I have also heard of problems when the capacitors are connected directly to the line side of a soft start. It is best to have a little impedance between the soft start and the capacitors. Usually 50 feet of wire is enough. If the capacitors must be connected at the starter, use a contactor for the capacitors which energizes when the motor is up to speed.

 

[GGOSS]

Hello all,

Many thanks for your considered responses. Application details are as follows.

Motor: 375kW/605A/415VAC 3-phase TEFC

Load: Direct couple Screw Compressor (Unloaded start)

Connection: 6 Wire (Inside delta) with main and bypass contactors as previously advised. The main contactor is positioned in the line circuit, thereby providing total off-state isolation.

Starters: Utilise closed loop controlled current technology.

Duty cycle: The starters are being operated well within their designed/rated operational duty cycle.

Cables: Cable run to the motor would be in the order of 15 - 20 meters. Cable type & run back to the transformer will be advised. Can advise however that the cables are not screened.

PFC Equipment: Certainly none connected to the output of the starter. Understanding is that there is none connected to the input side either but will re-check.

A total of 3 such failures have occured at 2 seperate sites. In all cases the starters/motors are supplied from their very own transformer with negligable volts drop detected during startings.

In 2 of the above cases, there have been no repeat failures since replacing the power assemblies ie more than 3 months ago. As no other components (PT's, CT's, PCB's) were replaced, one can only assume that there are no issues with the snubber circuits and that the SCR's are seeing sufficient gate current. We are still working through the third case.

Note also, there has been 1 motor failure. Motor manufacture advised transient over-voltage damage thus supporting our evaluations of at least one of the failed SCR's.

Any ideas?

Regards,
GGOSS

 

[jbartos]

Suggestion: The soft starter should have some voltage transients/spikes protection in terms of TVSS.
Visit

http://www.thomasregister.com

and type Suppressors: Transient Voltage Surge under Product or Service, which will return 21 companies to approach to

 

[jOmega]

GGOSS:

Remembering back to the days of SCR DC drives when the highest PRV was 1000 volts .... and the mains voltage was on the high side of 480... (like around 490) .... there was an installation at Coors in Golden, Colorado, that had several 15 HP 480 vac 500 vdc drives all hung on a wall. Along would come a voltage transient and ONE drive would lose its diodes and/or SCRs (3-phase half-wave bridge)...

Similar to your question ...
".....would expect other sensitive electronic equipment also connected to the same supply bus would also have suffered, but this has not the case."
.....I too asked why one drive was damaged and the others not.

The answer was quite simple..... Whichever drive absorbed the transient energy, it protected the other drives from being subjected to it. In otherwords.... one drive in taking the hit... acted like a sink for the transient (or surge suppressor) and saved the other drives.

Kind regards,
jO

 

[HarryDampers]

GGOSS

I have always been wary of inside delta on large machines . Some years ago I had a few spectacular thyristor failures and flashovers similar to those you describe....funnily enough , at least two of the starters were on compressors.

One thing to remember is that the thyristor 'sees'line Voltage when in the Delta loop and Phase Voltage in the inline connection.

This means that if (as can be normal practice)the Starter manufacturer has used the same voltage rated devices for inline as inside delta .......then his safety margin is greatly reduced .

Of course he will argue that the device rating is ok for both modes .......but I have had my fingers burned more than once.

 

[jOmega]

HarryDampers:

The compressors associated with the starter failures you mention...... would they by chance be of the Reciprocating type ?

and did the failure occur during initial starting ?

jO

 

[Marke]

Hello HarryDampers

Yes the SCRs during their OFF state are subjected to the phase to phase voltage on an inside delta installation and during that period, they are more susceptible to line transients than SCRs in a 3 wire installation, however during start and run, the voltage stress on both devices is not to dissimilar. The normal rule for devices connected to the three pahse supply is to apply devices rated at 3 x the rated line voltage and that is generally appropriate. It is a statistical situation. On any supply there are transients and the chances of a transient being high enough to damage the thyristors reduces as the voltage rating of the devices is increased. If the rating of 4 times the line voltage was applied, there would be a lower possibility of failure, but at a much higher cost.
There are some manufacturers who appliy the 2.5 times rule and a few go even lower. This is quickly accelerating the rate of failure due to transients.
From my experience, the 3 time rule is appropriate and generally one only gets problems in regions where there is a high supply impedance, there is power factor correction, and the starter is installed without a line contactor. This is true for both 3 wire and 6 wire installations. In these resonant cases, the voltages can be very high and you would need to raise the voltage rating of the device by a factor of 3 - 5 times to get any security. Any overvoltage protection device across the SCR simply serves to excite the resonance and make it worse.
I have experienced situations where the supply voltage has jumped when the motor is switched OFF and the magnitude of the surge has been sufficient to damage SCRs or cause them to overhead trigger. I recall one installation where the motor gave a kick and thump at stop. Enabling the soft stop feature with a one second ramp overcame this problem as would the use of higher voltage SCRs.
I do not believe that there is anything fundementally wrong with the 6 wire or inside delta connection provided it is done correctly. I have been associated with successful installations using this format for over twenty years, and I certainly would not discourage such an installation based on my experience.
Best regards,

Mark Empson

http://www.lmphotonics.com

 

[jOmega]

Mark,

Have you ever looked at the current when starting a reciprocating compressor having back pressure on the discharge port ? .... and likewise, have you observed current when stopping such a compressor where there is a high pressure on the dischage port ?

The reason I ask is that I've seen such situations where in the starting case, the high pressure on the discharge port causes the compressor/motor to initially kick backwards.

Then consider the stopping situation where the high discharge pressure over-drives the piston(s) hence the motor.

In the starting case, I believe that a high current results from mechanically forcing the motor to turn backward while the applied voltage is trying to rotate it forward.

In the stopping case, a regen situation occurs ... can subject the devices to a large voltage stress of significant magnitude to cause their PRV to be exceed with consequent failure.

Been a long, long time since I first looked at this failure mode... but I recollect that the line voltage and the regen voltage become additive ... and pop goes the device...

Comments ? ? ?

 

[Marke]

Hello JOmega

From my expereince, the start situation just requires a higher start current to develop enough torque to get the beast running, and I have not experienced a problem related to regenerated voltage causing SCR failures. What I have seen though, particularly in three wire situations, is the premature commutaion of the SCRs causing conduction overlap problems and essentially a burst of DC through one SCR causing a sever back emf at the end of the DC and SCR failure due to that. Essentially, lost control of the balance in the SCRs. Maybe if the voltage rating was on the low side, this could be a problem, i.e. using 1000PIV SCRs on 480 volts. I would never go below the 3 x rule for PIV rating.
Best regards,

Mark Empson

http://www.lmphotonics.com

 

[GGOSS]

Hello all,

Your input here is very much appreciated.

HarryDampers, there is no doubt that there are many soft starters on the market that do not perform well when connected in 6 wire mode, however this is certainly not the case with the products in question. Although I cannot divulge information re the technology used, I can advise that we have had tremendous success with 6 wire applications, so much so that almost all starters issued for motor above 250kW are connected in this manner......and continue to provide troublefree operation.

What concerns me about the three failures mentioned here is that they are extremely abnormal occurences based on our expreience. That is, with sales exceeding 3000 soft starters per annum (18,000 SCR's), 20 plus years in the game and only 3 such SCR failures to speak of, I'm sure you can appreciate why these have me scratching my head! Also, we have never experienced the "spectacular thyristor failures and flshovers" as described in your post of September 2, 2003, which I guess says something for the quality and/or robustness of our products.

For those who have raised the very important point of SCR's voltage rating and/or provided comment to this, I can confirm the devices used here are in fact 1600V rated ie 4 times supply voltage.

Marke made the comment "the SCRs during their OFF state are subjected to the phase to phase voltage on an inside delta installation and during that period, they are more susceptible to line transients than SCRs in a 3 wire installation" however in these installations the starters are used with isolation contactors, installed in the line circuit and therefore the SCR's are NOT connected to supply/exposed to transients when they are in their OFF-state.

jOmega raised the issue of back-spin on reciprocating compressors. This is also possible in some Screw compressor installations and is addressed either mechanically or by setting a very short soft stop time as proposed by Marke.

I'm not sure we've uncovered the fault cause yet, but hopefully through more discussion here, will do so.

Thanks again and regards,
GGOSS

 

[HarryDampers]

GGOSS , I was not in any way trying to knock anyones product , nor was I trying to suggest that Inside Delta should never be used in any circumstance .

I was expressing a personal opinion based on some unexplained occurences . The Softstarters I refer to were also robust and well designed but the supply Voltage was 575V.

Regarding the thyristor Voltage , I dont think that it can be argued that unless you use a higher rated device for inside delta applications(most manufacturers do not), the thyristors safety margin is very much decreased as opposed to inline use.

So at the end of the day , both you and I have had similar problems using this connection on a few applications, therefore,I think that until some explanation is forthcoming , I will continue to advise caution when using large drives in open delta.

PS.

Have you solved the problem or is it ongoing ?

 

[HarryDampers]

Just another point regarding Markes comments , surely the thyristor is "off" at the end of every half cycle . Of course the Contactor will isolate the starter from mains borne transients when the starter is non operational but at other times , it is exposed.

 

[GGOSS]

Hello HarryDampers,

Your points are well taken and I do understand that you were not trying to knock anyones products.

Most recent issue is still on-going. Just waiting for the return of failed SCR for post mortem analysis. On that subject there is a paper that may be of interest to you and others at

http://www.lmphotonics.com/forum/files.php?action=category&id=4

Thanks angain for your input.

Regards,
GGOSS

 

[GGOSS]

Hello all,

The post mortem analysis conducted on the most recent SCR failure revealed clear evidence of transient overvoltage damage. Refer Fig. 11-19 in the B G Meyer document referenced in my post of Sep 2, 2003.

I have now confirmed that there is no power factor correction equipment installed at this site and am quickly running out of ideas.

I plan to evaluate the performance of the supply transformer further as I suspect automatic tap changing with load may be the source of the transients but have not ben able to confirm to date.

The only other clues available to me at present include a series of phase/current imbalance trips (approximately 2 in every 100 starts) and a single CPU error. Both of these strongly suggest supply disturbance but do not assist me in determining whether those were limited to the LV side of the supply transformer or if they were reflected from the HV side.

Any further thoughts/advice you may have would be greatly appreciated.

Regards,
GGOSS

 

[jbartos]

Suggestion: When it comes to thyristor applications, there are about five possibilities of the thyristor failure:
1. Thyristor is of poor quality
2. Thyristor controls is poor quality
3. Thyristor application is poor
4. Thyristored soft starter has improper input
5. Thyristored soft starter has improper load
6. Etc.

 

[GGOSS]

Hello jbartos,

Re your points 1 & 2; Thyristor is of poor quality & Thyristor controls is poor quality. I'm sure you would agree 3 SCR failures in a total of 18,000 SCR's distributed implies outstanding Thyristor & Controller quality. Also you will note from my previous posts in this thread, post mortem analysis of the failed devices strongly suggests transient over-voltage damage.

Re you points 3, 4, 5 & 6; needless to say these are still under consideration.

Thanks again for your efforts.
Regards,
GGOSS