Medium voltage soft starter - commissioning

[ters]

The SSM series 6kV ABB soft starter is used to drive a 3.5MW asynchronous motor, which drives a large fan. The vendor seems to be unable to provide commissioning assistance in a timely manner, so the Client may have to commission this using its own resources.

The primary soft starter configuration starts with a disconnect switch, fuses, contactor and thyristor modules. There is also a bypass contactor.

On the secondary side, the equipment uses an ABB controller and an optional Multilin 469 motor protection. Optional – since the soft starter electronics seems to have a number of protective functions already built in.

We have an idea how to set the protections, but the controller also needs all kind of settings to be applied. The process is relatively steady and the maximum number of starts is probably less than three per day.

Could anyone provide some general guidance as how to set things like start mode (there is several to choose from – dual ramp, custom curve, jog voltage, etc), starting torque, acceleration (ramp time), current limits (min and max), initial voltage, etc. Thank you.

 

[Handlin]

There are a number of reasons to use a Reduced Voltage Solid State (RVSS) starter. The reason you chose an RVSS for the application will have a direct impact on the selection of starting parameters. For example if you are using an RVSS to reduce line voltage drop during starting, you probably want limit the maximum starting current.

For a fan application, you will not need the dual ramp (used for different starting parameters for different load conditions - chipper, crusher, etc.) Most RVSS fan applications I have started use a current ramp with three parameter settings, initial current, maximum current and ramp time.

Several manufacturers having RVSS simulation software to predict the performance of a set of starting parameters for a given application. However several required system, motor and load characteristics required for these programs is not easily obtainable.

My recommendation would be if you can not get an ABB serice tech to commission the unit, find a qualified service engineer or commissioning agent to test and program the RVSS. Typically the start-up should include checking SCR gate resistance, mechanical inspection, and programming the parameters. Also decide which motor protection (RVSS ofr Multilin 469) you want to provide the primary protection. Most RVSS onboard motor protection is roughly equivalent to the Multilin 269.Since you have invested in the 469, I would use that as the primary protection

 

[DanDel]

You may be able to get some more guidance from the manufacturer of the driven load (fan).

If not, I agree with Handlin about getting a consultant or qualified power system testing company to help. You need assistance with providing (and implementing) the settings, and testing the system once it's set up.

 

[jraef]

The ABB MV soft starters are actually built by Motortronics in Clearwater Florida. ABB is, by their brand label contract, supposed to provide their own people to do commissioning. But last I knew of ABB had very few people qualified to do it, so that may be what is behind the delay. You could contact Motortronics directly, but they are going to want to be paid since their discounted brand-label price to ABB does not include field support services.

And if you are willing to pay for someone qualified so that you can get going, you can also contact Siemens Industrial Services. Siemens has brand labeled the Motortronics soft starter as well (although they recently stopped doing so), and they have a lot of trained technicians in the field.


"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

 

[ters]

Thank you very much Handlin, DanDel and jraef for very useful comments.

We will use Multilin as a main motor protection and use the built-in one as some soft of backup. There are actually plenty of protections... In addition to those two, there are also MV fusses in the soft starter, while the upstream protection, which is on a switchgear feeding the soft starter, is just 10 meters away and is a feeder relay REF5542plus, but it could even work as 87.

So we are unsure how to coordinate all this, what protective functions should be used at each particular device, and which functions should actually trip what – 87 and overcurrent protections will obviously be set to trip the switchgear breaker, while some other protective functions are set to trip the soft starter contactor only.

We knew that ABB makes no its own soft starter and the sell relabeled Motortrponics equipment, but we were not aware that they could not find anyone to assist with the commissioning… The equipment was supplied by ABB USA and after looking all around the Globe (and possibly wider...) they finally indentified someone who knows something about their soft starters in Egypt, however this project is in Europe, so there are now visa problems involved (plus the rate is the highest commissioning rate I have ever heard of...)

Thanks for the hint jraef, will try Siemens also. The additional difficulty here is that the likelihood of finding someone who is familiar enough (and equipped enough) to commission both, the mysterious SSM electronics and the optional Multilin 469 (which is stand alone for the most part)is probably next to none, so two testing agencies may have to be involved...

 

[jraef]

In general, when I commissioned them I would recommend to people that if they have a Multilin and their technicians are familiar with programming the Multilin, then allow that to be the only protective device. Troubleshooting after an incident can be extremely difficult when you have multiple protective devices that have different error tolerances. Besides, although the printed specs on the Motortronics / ABB RVSS micro shows a lot of functions that appear to duplicate what is in the Multilin, it is not as sophisticated as a 469 relay (I would know, I wrote the specs for the literature they both use). Just let the RVSS micro handle the ramp control and sequencing of the contactors.

Just be sure to make allowances in the Multilin programming for the fact that the RVSS will extend the acceleration. And if you use Deceleration, the Multilin can get "confused" because the current profile is a mirror image of the Accel profile. You usually have to watch it a fews times and make adjustments.


"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

 

[slavag]

Hi.
Im not familiar with this softstarter, but if it is include protection functions..from my point of view you don't need additional protective relay.
In additional, you have line protection with REF542plus.
Best Regards.
Slava

 

[oldfieldguy]

I agree with jraef. The Multilin has better fault reporting. Set the protective elements on the soft start as backup to the Multilin. Also, the Multilin is a better device to communicate with for your SCADA.

old field guy

 

[ters]

Here is an update on this, hoping that some of you may have put more light on the problems we are experiencing.

We have gone through various stages of the commissioning the soft starter, had an ABB specialist, also had some other consultants involved but still have problems. The soft starter does not behave as we expected.

The motor: TECO-Westinghouse 3.5MW, In = 385Amp, 6kV 50Hz, 1485 rmp, locked rotor current 2300Amp, driving a big fan with a large rotational inertia. The fan has no load at the present (no air, dampers closed). To drive the fan with no load (air),the motor need to develop some 500kW.

The soft starter was set to starting voltage to 50% or around. Ramp time varied anywhere between 10 and 30 seconds (various trials). The current limited to 330% In (limit by some upstream protection).

The motor starts moving smoothly, but it never reaches the full speed before one of several protections trips on an overcurrent or overload (1500 Amp and up). The speed may be reaching about 30% and then it just stays there(no more acceleration)until the protection trips. During that time (running at flat low speed) the motor runs rough, strange sound, starts even to de-accelerate slightly and something finally trips.

We installed a test PT to monitor the motor voltage during the start, and it appeared that the motor (soft starter output) voltage never recovers to 6kV from some 4kV, which it reaches several seconds after the start. This is the voltage after SCR’s. The voltage before SCRs is OK, dips from initial 6.3kV to about 5.8kV.

To develop enough torque and reach the full speed, looks like the soft starter would need to deliver about 5 x motor nominal current for a few seconds, which would not be a particularly soft start.

Our expectations were that the soft starter would accelerate the motor (and the fan) at about 3 In and a prolonged time, which may be in the range of 40 sec. And then once at the full speed and the normal operating current, a bypass contactor would close and bring the motor directly online. If the fan is not moving any air, the current would be in the range of 200 Amp or less.

Since bringing the motor to the full speed and then closing the bypass contactor proved to be impossible so far, a decision was made to try to close the bypass contactor much earlier, when the motor is still far away from the nominal speed and when the current is still 3 x In or higher. This may not be a particularly healthy condition for the contactor since it would closing on a current which is much higher than its nominal rating.

Any comments would be greately appreciated. Thank you.

 

[oldfieldguy]

Compare your actual time-voltage-current with the manufacturer's provided curve. Then compare them with the parameters for soft start by your facility. This may include limitations to the facility by the supplying utility company.

You may find that with your rotational load, you cannot do a successful start within the parameters allowed. More likely, though, you can tighten up your overload and acceleration time protection to squeeze a start in.

I had the same problem a few years back at a facility here. The client had NO parameters from the utility other than it had to be a 'soft start'. We ended up with a very short, fast ramp-up to full voltage, keeping an eye on the voltage on the supplying bus to make sure we didn't drop out other loads during startup.

old field guy

 

[rbulsara]

The motor starts moving smoothly, but it never reaches the full speed before one of several protections trips on an overcurrent or overload (1500 Amp and up). The speed may be reaching about 30% and then it just stays there(no more acceleration)until the protection trips. During that time (running at flat low speed) the motor runs rough, strange sound, starts even to de-accelerate slightly and something finally trips.

Indicative of insufficient torque and not building up of the volage. Torques is the function of voltage. Try higher voltage and/or faster ramping of voltage from starting voltage to full voltage.

 

[ters]

oldfiled guy and rbulsara thank you very much for replying. Looks like we have several problems here.

Seems that we exhausted most of the combinations of starting voltages, currents, times, etc. with no success. Yes, we are limited with the configuration of an upstream substation and that seems to be a good part of the problem.

I finally talked to someone from Motortronics who knows something about this soft starter and he indicated that with the fan we have (The Impeller weight is 15,000 lbs, and the Impeller moment of inertia (WR^2) is 66,000 lb-ft^2.) a likely minimum starting current may need to be about 450% of the nominal motor current.

Well, this is not far off from the LRC which is 600%, so use of a soft starter does not seem to provide a big advantage. Except, there is still one. With the DOL start, 600% In is instant while with the soft starter, it will start with a much lower current (voltage) but needs to develop some 450% In to overcome to torque hump and finally bring the motor to the speed.

So if a high starting current cannot be avoided, then we seem to have more problems. The initial system design was based on the assumption that the motor + fan can be started at about 300-350% In with a prolonged start time. Now, to get 450% In from the upstream switchgear and the transformer seem to be problematic, both have limitations. The transformer is only 20MVA and is a high impedance one, so getting 4.5 In while there are other loads on the same transformer will probably cause a large voltage drop. The switchgear is also problematic since both, the feeder breaker and protections seem to be inadequate for 450% motor current.

Any ideas how we can make our life less miserable?

 

[rbulsara]

I suspected weak source as well. If that turns out to be your real problem...well it is a problem. I am not sure how many starting parameter combinations have been tried so far.

Even if the starting current goes up but drops off quickly, they system may like it better.

The fact that it does start at 50%V, I think ramping up voltage little faster would help. There is a sweet point somewhere between how fast you ramp up the voltage and how long you want to keep accelerating (trying to limit the current). One thing you cannot avoid is providing adequate torque to acceplerate, which is keeping up the voltage.

 

[ters]

We tried various acceleration times, from say 25s to 40s. We also tried various voltage rump up periods from 10s to 20s. All with limiting the current in the range of 300-330%.

We then made the upstream 110/6kV transformer fully available to feed only this load, but that does not seem to help either. Maybe we still have not hit that magic combination of numbers to bring us over the hump, although we have obvious problem not being able to pull more than 330% of the FLA due to upstream breaker and protection limitations. With this current limit, and the voltage rump time of 10 sec, the line voltage (before SCRs) during the start drops for about 10%.

With the starting current of 400% FLA, which we have no luxury of using yet since it trips the upstream breaker as the associated protection cannot be set that high (but we tried it anyway), the line voltage (before SCRs) drop is substantial. It drops from some 6.4kV to about 5.6kV. This is too much of a drop since the transformer is feeding only this load. The transformer rating is about 13MVA ONAN, 20 MVA ONAF, impedance 11%.

One time we closed the bypass after some 25s, but then the voltage drop was even more harsh, from 6.4kV to 5kV, but the condition was cleared quickly the overcurrent protection which was set at about 2000Amp.

We are unsure what to do next. The motor accelerates smoothly up to about 250 rmp, which is some 15% of the speed and then the speed stalls and even slightly de-accelerates before something trips or we shut down it intentionally after about 30 sec realizing that it will not accelerate any further.

Various parties involved in this suspect various causes of problem. Some believe that the week source may be a the most contributing problem, some speculate that there could be some mechanical problems making the high turning resistance at higher speeds, while some still believe that the soft starter is defective since, when the speed stalls, the motor behaves like it lost one phase (vibrations, abnormal noise and slight de-acceleration).

Tomorrow we will be connecting 3 PTs on the motor side to monitor the starting voltage and the wave forms, hoping that we can make some things more conclusive and eliminate some of the potential causes.

 

[rbulsara]

There is something wrong with the picture here.

Are you saying a 13 MW transformer cannot start a 3.5MW motor? Even a transformer close to motor kVA rating should be able to start a motor with acceptable voltage dip.

4XFLA of the motor is still close to 1XFLA of the transformer!. Why would anything trip at or near full load rating? What is the upstream protection?

5.6 kV from 6.4 is still only 12.5% VD, why is that an issue? Perhaps you are sitting there too long. Key is to recover from the dip quickly and only a higher voltage would do that. You are just not allowing motor to accelerate that will help it drop the current.

 

[ters]

Yes, I'm saying that a 13MVA transformer has problems starting a 3.5MW motor. I do not know why yet, but as I mentioned above, at 400% FLA starting current, the voltage dip was big, at least 15%. At 300% FLA, the voltage drop is about 10%.

A contributing factor is surely an unusually high transformer impendence, but there is likely more to this. Either the soft starter problems, or motor problems, or fan/bearing problems. Or maybe a mismatch between the fan characteristics and/or motor and/or soft starter in some ways.

Why anything would trip... Heh... Because the soft starter is fed from a new switchgear where the protections can be set as high as necessary and we do not have problem there.

However, that new switchgear is fed from an old switchgear where that old feeder breaker is only 800 Amp, and an old protection cannot be set any higher than 1440Amp and 5sec, that is the max, cannot be changed.

During the design of the system, it was assumed that the soft starter will limit the current to about 300-350%. That is now quite questionable, and is we don’t find some magic soft starter settings, this all may lead to necessity to change that 800Amp breaker and the associated CTs and protection before we can apply 400% or even 500% FLA as the soft starter manufacturer is now recommending. 500% is not much different than LRC which is 600%...

And even if we change the upstream protection and breaker, that is not going to solve the problem of the voltage drop, for which we still don’t have a sound explanation...

 

[ters]

And yes, forgot to add, that is true that we are sitting there too long.

But we don’t seem to be able to avoid it. The voltage dip I was referring to is before SCRs, not on the motor. The motor voltage appears to be about 1kV lover than the dipped line voltage. So when we adjust the voltage ramp time to 10 sec, and start the motor, here is what happens:

1. The current quickly (3-4 sec) develops to the set limit, in our case about 1300Amp, and stays there.

2. The line voltage (before SCRs) for about 10-12% and never recovers before something trips off or we turn the soft starter off after 30 seconds.

3. The motor voltage (after SCRs) builds to about 4.2kV and never moves up.

4. The speed reaches 15% and stays there.

So there seems to be a consensus that we need much more current in order to build the motor voltage to a much higher value, close to the line voltage, to go over the torque hump.

However, we cannot get more current for now, until we deal with the upstream limitations...

 

[rbulsara]

I understand your frustration.. I hope some other people would chime in here if I am beating the dead horse or the wrong horse..

So there seems to be a consensus that we need much more current in order to build the motor voltage to a much higher value,

Above statement has fundamental issues..current does not build the voltage! To raise the speed from 15% you need higher 'Voltage' which may increase some current but will also drop off..

If you can sit for 30s with 1300A, you sure should be able to take say 1600A for 5 sec. What does the TCC of the over current protection looks like? At 1300A for 30s, motor will trip out on thermal protection even.

You may not keep it that way, but you still need to find out if this motor-load combination even works at higher or full voltage with shorter ramp time. Then worry about voltage drops.

Who knows the motor is mismatched for the load, but that seem less likely with the info you are providing.

 

[ters]

I knew that the sentence "so there seems to be a consensus that we need much more current in order to build the motor voltage to a much higher value" may be confusing or maybe conflicting some basic principles, but when I asked the soft starter people why the motor voltage never seems to get higher than some 4.5kV at 300% FLA, the answer was that the SCRs are basically limiting the current by controlling (chopping) the voltage, so unless we let more current through, the motor won't see much higher voltage and will not develop sufficient torque.

The newest input from them, and looks like I'm talking to someone who knows his stuff is: let 400-500% FLA and rump the voltage very quickly within 2-3 sec.

 

[rbulsara]

That sounds much more promising..to try anyway.