Can someone please answer why Thyristors in Soft starters burn?
We are using a 400 KW AC3 (690V 60Hz) rated soft starter to start a 300KW (690V 60Hz) motor. Power is supplied by a 750KVA Generator set. The soft starter is configured according the motor datasheet. The soft starter manufacturer is saying that we have to use power form the grid; the generator set is not big enough for this application. I don't quite understand the reason though. The motor starts with no load. Soft starter is bypassed using a bypass contactor after top of ramp is reached. So far four soft starters are burnt during testing with seperate motors.

Hello Quazi Haque

Welcome to the forum.

Some soft starters do not perform well on the out put of a generator, especially if it is marginally sized.
The normal cause of failure on the SCRs used in soft starters, is either a severe overcurrent, or overvoltage. Both failures result in a shorted SCR. A suitably trained/skilled person can determine the cause (over voltage or over current) by opening up the SCR and studying the actual die and failure mode.

If the generator is marginally sized, some soft starters can go unstable and this can result in current and voltage surges which can do some damage.
If the alternator is self excited and fitted with a half wave, single phase peak reading AVR, the output regulation will not be good under the starting conditions of the soft starter. In effect, the soft starter and AVR can "chase" each other.

If the generator you are using has a PMG excitation and a three phase averaging AVR, then it should not be a problem with that motor size as long as the soft starter is suitable for operation on a generator.

Best regards

What brand of soft starter is it out of curiosity?

Hi Mark,

Thanks a lot for the info. I am impressed by the quick response. I already added this site to my favourites list as I am sure I will be using this forum much more often from now on.

I am checking with the generator supplier if the generator has a PMG excitation and a three phase averaging AVR. We are using ABB PST210 690 soft starter (answer to jraef's question). It has built-in over-current protection according to its datasheet. So I will be surprised if the thyristors are burnt out due to over current. By the way, we connected the soft starter inside delta of the motor winding, but it can still measure the phase currents.


Best regards,
Quazi

Hello Quazi

The overcurrent protection will not protect the SCRs. That is for motor protection only. The problem is that once an SCR is turned ON, you can not turn it OFF. It will self commutate when the current passes through zero.

If voltage wave form distortion is causing SCR miss triggering, it is possible to get DC in one phase and the current can go high enough to damage the SCR. Similarly, the voltage spikes can be very high.

I have no experience with the ABB units on generators, so I can not comment on that one.

Best regards,

Hi Quazi,

Just interested to know if the the motor failed together with the SCRs? You might want to do a meg test on the phases of the windings.

There has been a couple of inside delta failures of late with compressors applications on my end...



Marke and Jraef,

Just like to know your views on if a soft starter is wired inside delta with settings on parameters on soft starter - Line, would that caused SCR to fail and in turn caused the motor to fail? There is no bypass contactor in the system.

I do understand that the triggering timing will be different but what kind of effect would it caused?

We currently have a failure as per above on screw compressor application and soft starter supplier is saying a overvoltage that caused a blow to the motor windings and the soft starter would not have caused that even with above settings.

Application was running with the wrong parameters for 3-4 months before changing over before 1st motor and soft starter failed.


I can't name companies here... So please take soft starter as a general big in the market soft starter.

This case above is just one off the I have gone thru with.... I still got others that are unsolved which I am still waiting for more information.

Thanks
SCIM

Hello SCIM

Welcome to the forum

The answer is yes and no, it depends. How is that for being evasive??

Actually, it depends on the actual starter and the way is works internally.

At one extreme, we have soft starters that will handle that with no problems. You do not need to set them up for three wire or six wire (inside delta) connection. At the other extreme there are soft starters that can be installed in inside delta connection provided that you have the correct phase sequence and wiring of each winding from the output of a phase to the input of the preceding phase, and then there are the starters that can not be wired inside delta at all.
There are issues with firing overhang, initial phase angles etc etc that must be set correctly depending on the connection.

If the starter is not able to auto configure for inside delta connection, then there is a good possibility that there will be some mistriggering that could cause DC to flow in one or more windings and that could cause the winding to fail and the SCR to fail. This is totally dependent on the starter being used.

Similarly, if the starter is connected inside delta and there is no line contactor, then any line transient is blocked by the SCRs and in the inside delta connection, the OFF state transients applied to the SCRs are equal to the line to line transients, whereas in three wire connection, the transients applied to the SCRs are equal to the line to neutral transients and so are significantly lower.
If a high transient occurs, then the SCR could be damaged and this would in turn cause the motor winding to fail.

Analysis of the failed SCR and the failed motor winding can give clues as to the cause.
If one whole winding is badly burnt, that would suggest to me that the motor failure could be due to an SCR failure. A skilled technician can give a good indication of the mode of SCR failure from the damaged die.

Do you have an indication of when the failure occured? was it while the starter was OFF, while it was starting or while it was running?
Was the motor damaged on one phase or all phases?
Was the winding burnt, or was there a whole blown in the winding?

Best regards,

I personally am not a fan of using soft starters inside the delta, and I have written two papers on the subject for my previous employer, who did not offer it as an option. To me, the risks far outweigh the advantages. There is only one advantage; lower cost of the soft starter power components. There are however multiple risks above and beyond what is normal for all soft starters. Here is my (not so) short list:

1. High risk of a single shorted SCR causing motor damage. In a normal connection if one SCR fails it isn't a big deal. The motor will not ramp properly and a shorted SCR detection scheme may lock you out from re-starting next time, but there is no inherent danger of damaging the motor. You would need for there to by multiple SCRs shorted in at least 2 legs for the risk to present itself (the risk being unimpeded current flow though one set of windings in the motor). Once you connect inside the delta however, even one shorted SCR presents an immediate opportunity for that unrestricted current flow through the winding, because 1/2 of the current path is already there. So the only way to ensure it doesn't happen is to use an isolation contactor (called a Fault Contactor by some manufacturers). Ironically though, the cost of this extra contactor is usually MORE than whatever you saved by using inside the delta anyway, so hardly anyone puts it in. That means they are at a much much greater risk of losing the motor.
2. Higher risk of SCR failure as well. SCRs should technically have a higher PIV (peak inverse voltage) rating when used inside the delta, yet nobody does that, which means SCRs are more likely to short on faults. When SCRs are picked for a soft starter application, a major consideration is that the PIV rating should be 3X the line peak voltage. So on a 480V Y system, the maximum voltage to ground is 277 so the maximum peak voltage across the SCR in a fault can be 277x1.41=390V. Therefore the PIV needs to be at least 1170V, which is rounded up to 1200V. Because a lot of IEC countries use 415V, 1200PIV has become the minimum rating in use all over the world (it never hurts to have a little extra). So that rule ends up equating to the PIV being based on 2.5x the nominal line RMS voltage (1200/480=2.5) and over time, that is how many people have come to interpret the rule and have even come to believe that IS the rule. Most of the time it works out fine to do it that way, exceptwhen you put the SCRs inside the delta! If used inside the delta, the peak fault voltage potential to ground that the SCR can see becomes theoretically the line voltage. So on a 480V system, the peak voltage is 480x1.41=677V, which means if you want to select the SCRs correctly, they SHOULD have a PIV of 679x3=2030V! But nobody does that because the SCRs would be extremely expensive. The values would of course be lower for 415V Y systems (1755PIV) and that's why you see inside the delta used more freely in IEC countries, but there is a significant extra risk for users in the US. Even with the 1755PIV value in IEC countries, 1800PIV SCRs are readily available and there have been arguments in the soft starter manufacturing community that this should be the minimum rating when used inside the delta. But as soon as one manufacturer ignores it and uses 1200 or 1600PIV, his pricing becomes incrementally better and everyone else has to follow suit. The loser in this argument however is the end user. At the very least, users with 415V maximum systems should select the 690V class soft starters, because they will have (should have) 1800PIV rated SCRs. But salesmen are not educated on this issue so they sell the 500V class products and put the users at risk.
3. Risk of having a technician in the future not understand the unique nature of the connections, and do something catastrophically wrong. Having to connect 6 leads in the correct pattern is not too bad when the equipment is first installed and the manual is right there. But 5 years down the road, the reality is that the manual will be lost and the technician who has to work on it may have never seen this odd-ball arrangement before. One mistake and all the savings will go up in smoke.
4. Can't use low firing angles, so initial torque may always be higher than you want, which may end up negating one of the best benefits of soft starting anyway. I'm not really sure about this one, it was related to me anecdotally, I have never seen it.

Now I'll admit that I have not been as intimately involved in the ground up engineering of soft starters from scratch as Marke has, so he may have some corrections to what I have come to believe, but in general, inside the delta will always have no attraction for me just based upon point number one.

Thanks for the replies guys,

Marke,


This soft starter that failed does not have auto configuration and when it failed, manufacturers were 1stly quick to point out the there is no effect and could not have contributed to a motor failure. (If they see this, they will most likily know who I am)

After questioning them that the triggering timing would be different (that's why the parameter settings) and questioing them as to if that would lead to more harmonics which in turn caused the motor to fail, they agree that it might caused more harmonics but only at start and they do not know what kind of effects it would caused on the motor. They are also not very willing to supply the PIV rating of their SCR.

I have just received a letter from them stating that the Soft Starter would not have caused the current situation we are seeing in the winding failure of this motor.

This motor failed on start. There is no discolouration in windings and the only blow is just between 2 phase. It's along 2 smaller leads at 11O'clock covered with fiberglass between two phases on a leased up point and the blow isn't a big blowout. Most who had seen the failure commented that it looks more like a high spike in voltage without any high current or it would have blown alot of copper out. I am still quite young in analysing winding failures and been in the industry for 4years but am not convice that the wrong parrameter settings with a correct inside delta connection who have no effect to winding insulation.

Could you please explain on how mistriggering can caused DC to flow and inturn caused winding failures?
Do you mean DC voltage? or Direct current? Is there a way to prove so? The SCR will be taken apart soon and we are waiting for the soft starter manufacturer to organise for a time.

We currently have on the same site, another motor that failed last week with failure being like an explosion of windings with copper everywhere. The motor was opened today and I have yet to see the photos but will get them tommorrow. Base on the explaination on the phone with someone who saw it, my believe is that it will be totally different from the winding failure of the 1st motor.

As said earlier, there have been other failures elsewhere on other sites. Others I am led to believe that they have the correct settings and correct connection of Inside Delta with fingers cross.
All failues(6-7 so far till date) are after commissioning of about 4-6months. My thinking is, could inside delta connection actually be determital to the insulation system or breaking it down slowly. There have been cases whereby SCR failed but motor was still fine and after SCR was replaced, motor failed together with SCR. Does this mean that for Inside Delta Connection, motor to be supplied should have the best insulation system possible? e.g ISR Wires only?

I only happen to chance upon that SCR were failings together with motor as well only after the 4th failure!


Jraef,

Yes, I have read alot or almost all of your writings with regards to Inside Delta Connection (Of coz, Marke's as well and GSOSSS) and I agree with your saying that for Inside Delta Connection, there is a higher chance of failure rate than line connection. I have actually been talking to many soft starter manufactuers and reliease that most of them do not know what they are selling only for a rare few in my current place of location. I am impressed by the rare few breed that actually knows what they are selling and marketing.

I do believe that I will make a better salesman than some if I will to join the soft starter industry here.


1:High risk of a single shorted SCR causing motor damage

Could you please tell me what kind of fault isolator contactor are you taking about? Are you saying the contactor between 3phase source and soft starter?


2: Higher risk of SCR failure as well.

It's 415V 50Hz here.

I believe industry here is telling their customers that Inside Delta is good on their 415V drives with increase in amps by 1.73 by some.
Till date I have yet to seen any that they will recommend their 690Volt system for inside delta connection with 415V supply.

The rating of PIV is a very well guarded information here, in fact some don't even know what the PIV stands for.

There is one case that I know of whereby a customer has complain about knowing of these failures and the soft starter manufacturer has changed all soft starters to suit Inside Delta to Line. From what I hear, it's FREE OF CHARGE! It only leads me to think that manufactuers themselves are unsure of what might happen and can't guranttee their products for Inside Delta.

3. Risk of having a technician in the future not understand the unique nature of the connections

It has happen here, correct connection but the parrameter settings wrong on the soft starter.



I am actually quite worn out by this current situation but it's affecting alot of companies in terms of dollars of damage done. I have even been threaten by a big company saying they will consider sueing me for not recommending inside delta connection with their products!!! I said it's due to the number of failures that it was said..


It's a bloodly crazy world today....

Hello SCIM

The details in the SCR control and effects of incorrect triggering angles and overhang angles gets quite involved and is beyond what can/should be discussed here, but if the triggering angles are set up incorrectly, it is posible to get 1 SCR turning on only in one phase. This results in a burst of DC current. This could cause a high back emf on that winding, or for a high voltage to be induced in an other winding.
I have seen some pretty nasty currents and voltages occur with incorrect triggering. - if it was not a problem, they would not have the settings!!

Your description of the motor failure indicates that the SCR did not fail first. The initial problem was the punch through on the insulation of the motor and this fault will have damaged the SCR.
As the fault occurred at start, I would suggest that the possibility of the failure being a misfire problem can not be discounted. A line transient problem in an inside delta installation would probably damage an SCR before damaging a winding and would be most likely to occur when the starter is powered up but not running.

Other comments:
1) I believe that ALL soft starters wired inside delta must have an isolation contactor that disconnects the incoming supply when the motor is not running. If you do not have the isolation contactor the chances of an SCR being damaged by high voltage line transients is much higher than in the standard three wire connection. More important however, if you install the soft starter inside delta, and just rely on the soft starter to start and stop the motor, then you are leaving full voltage on one end of the motor tails when the motor is not running. In the event of a fault, you can not disconnect the motor from the supply. - It is very much like using a two contactor star delta starter.
If you do have a high voltage transient on one phase, it could easily cause one SCR to fail. This puts full voltage on one winding only and this will result in a high current flow on that winding until something breaks. - usually the smoke coming out of that winding.

2) The same PIV rating is used for both thre wire and inside delta connection and this is appropriate, however in the OFF state, the SCRs are blocking 415 volts rather than 230 volts in a three wire connection. - less margin for transients.
As far as the current is concerned, in a situation where there is a continuous sinewave (Full voltage) applied to the motor windings, the current inside the delta loop is 1/rt(3) of the current in the line. However, when you have a discontinuous sinewave, (during start) the thermal effect is very dependent on the wave form and the conduction angles. In inside delta connection, the conduction angles are very different from the conduction angles in standard three wire connection. The correct re rating of the unit for thermal reasons, is to increase the motor size by about 50%. If you increase by 73% as most suggest, you are actually heating up the junction of the SCRs by a much higher value during start than you would in three wire connection.

3) Inside Delta connection is OK and is not a major problem with correctly designed starters. I have done many, but I believe that you must rate correctly, and you must use a line contactor that opens when the motor is not running. It is definitely preferable to use an auto setup starter so that there can be no problems with incorrect configuration or wiring.
I have experienced problems with earlier systems that were not intelligent where the installers were also not intelligent and got the installation and configuration out of synch.

Best regards,

Dear Quazi,

Your generator 750 kva is definitely undersized for the application even if you use a softstarter of appropriate size.

I am using a 1250 kva generator for with Reduced voltage starter for a 450 hp motor.

The starting current would be high causing a severe voltage drop and hence the softstarter would be under lot of stress.

Check the voltage at the input of softstarter, after you give the start command.

Resize your generator.

regards,

Chaterpilar

Hello SCIM


I do not believe so, provided that the SCRs are correctly controlled. This is supported by many installations that have been running successfully for up to and beyond 20 years.
If there is mis firing of the SCRs such that there results a large DC component for a short period, then yes, at the expiration of the DC current, there will be a very high back emf generated and this will cause a high voltage transient that could damage SCRs and/or windings.

To be more specific on the risks, I would need to know the topology used for controlling the SCRs. I have found that it is important to use a different algorithm for inside delta connection than for line connection in order to ensure the correct triggering of the SCRs.

Best regards,

Hi Marke,

I forgot to add with regards to 2nd failure on site is that the soft starter was on wrong settings for 3 months before Soft starter's technical engineer went to change to the correct settings and increasing the starting current to 580%. The reason they claim for not being able to start the compressor was due to it being set at 380% for limitation of start current. Therefore, this motor failed with right settings after about 3 months on site with right settings and 3months with the wrong settings and only had running hours of less than 50Hours.
The first motor failed after 3 months on site with wrong settings.


Soft starter company is currently claiming with regards to 1st motor that failed with incorrect parameter settings:

" The soft starter can not cause an over voltage to the motor. The initial incorrect settings could have lead to additional motor heating during starting, but the physical nature of the soft starter will not significantly increase the voltage levels "


The electrical company that did the installation for both motors has also make mention that when they tried to do a test run with "Inside Delta parameters" , the motors were rocking and couldn't seems to get to speed. I am unsure of what their starting voltage settings are.


1.
With regards the "correct" triggering of the SCR on inside delta or line connection, does that mean that there is actually no way of determining so if the manufacturer has the right programming for timing of the triggering.

I am led to believe so because manufacturers involved in the situations that I am involved in could not provide a reply as to where went wrong other than stating that the winding insulation failed first that caused the SCR to fail.

Another problem I been thinking is, if it's transist voltage to be in the system that's causing these problem, it would be the same for DOL or Star/Delta system and it would still take the motor windings out if insulation system is unrated for it.

Question being now is that why are we seeing these failures only with Soft Starters with Inside Delta? Due to the source or due to Others like installation?



2.
With regards to the rating of soft starter being rated Line Connection x 1.732 or x 1.5 for inside delta connection:
When I raised the questions to soft starter manufacturers on why the difference in ratio for inside delta connected,

One of the Manufacturer Product Manager (MARKETING) response ,
(Nothing mention of the PIV rating other than with regards to thermal rating)

" ...However, there are many variables which can be taken into account during a design phase which allows for someone to take full advantage of the connection. A couple of those variables revolve around SCR thermal capabilities and there ability to handle peak surge currents. One of the reasons we maximize our ratings is because of the way we have tested the device. Most of our thermal tests and models were done with motor/loads set up in a