VFD starting

[HamidEle]

I am designing a pipeline project, where the system is very weak, so that we have to use VFD to start the biggest motor(5000KW).

Utility company informed the client that even VFD starting would cause voltage drop on the consumer bus and on the utility bus, which would exceed the flicker limits.

But I did load flow study. with all the loads running, the volatge drop is only 3%. There is no voltage drop on utility side.

Since while VFD starting, the power factor is 0.95 and current is limited only around 100%FLA. Theoretically, the voltage drop should be the same as under normal running condition.

However, ETAP gave me the different results. Whenever I started motor by 100% or higher, i always got much higher voltage dip on the system, which didn't make sense to me.

I presumed that ETAP cann't model VFD starting properly.

Any inputs would be appreciated.

 

[dpc]

My guess is that AFD starting will not cause much more voltage drop than running the motor at full load since you should be able to limit the current. What is the load - a centrifugal pump?

But I'm a little suspicious when you say that "there is no voltage drop on the utility side". There is always voltage drop when you have current flowing. Are you sure you are not looking at the swing bus? Its voltage will be artificially constant due to the limitation of the power flow reduction method.

If you are buying a 5000 kW AFD, the potential vendors will be happy to work with you to estimate the line current demand during starting and probably do the power flow calcs for you. Their estimates will be more accurate than the generic ETAP data you are using.

 

[ozmosis]

I am also not sure how else you will be able to start your motors if the VFD is going to cause a problem with volt drop on the supply (I agree with dpc, it shouldn't). Sounds suspiciously like the utility want you to upgrade their network at your cost...

 

[aolalde]

If you start at very low frequency and then with gradual frequency increase up to the operating speed, the current could be controlled to remain under 100% of FLC. The voltage drop should be the same as that for full load condition. The current on the power supply is that of the VFD input, not the current driven by the motor which anyway remains low too.

 

[HamidEle]

Thanks for your responses.

I remodeled the system again and found out that the voltage dip while starting motor (by VFD) is the same as under normal running condition, which will exceed the utility flicker limits.

This means, the system is too weak to handle such a big load, unless the utility build another transmission line.
DPC. It is a centrifugal pump.

 

[dpc]

The centrifugal pump and fan are among the easiest loads to accelerate. So, as has been said, if you can't start it with a VFD without excessive voltage dip, you probably can't run it either.

 

[HamidEle]

To my surprise, load flow based on ETAP is not reliable. ETAP considers Utility bus as infinite as it does load flow calcualtion.
Actually it is not the case, when the system is weak, there is always a high impedance on Utility side, which will cause high voltage drop in the system even under normal running condition.

what else we can do to solve this problem?

 

[dpc]

You just need to add another bus upstream of your "utility" bus, connect it with a T-line and make it the swing bus.

This is not a question of being "reliable" - it just the way the load flow calculation works. There has to be at least one bus that is designated to have a constant voltage and capable of providing/absorbing all necessary watts and vars.

Virtually all power flow programs work this way.

Are you talking about the ETAP problem or the actual problem?

 

[HamidEle]

DPC,
Thanks a lot. Could you do me a favor to verify the load flow calculation result? This is extremely important to us.

Our system is configued like this,

Utility supply, 25KV, 80MVA S-C
Transformer 15MVA: Z=%7 25V/6.9kV
Running load: 13MVA

It is a radial system. Any utility said they couldn't supply a strong power supply.

The voltage drop on 6.9KV bus I got is 20% under normal running condition.

 

[powerjunx]

HamidEle, i dont see any problem upon using VFD on utility side unless the configuration is fairly refined for its application.
Like dpc, with such large size machine OEM's technical support unwaveringly helps you out on the mess. They'll provide you with what you are doing this time, accuratetly rather than ETAP.
The only problem that it remains unsolved is your load flow analysis.

 

[dpc]

Well, the voltage drop is going to depend a lot of the power factor of the load, but you seem to be off by quite a bit with your 20% drop.

I ran it in EasyPower and got about 2.5% drop at a load pf of 0.95. As the power factor decreases, the voltage drop increases.

Are you sure you've got the transformer data entered correctly - like maybe MVA instead of kVA?

You can deal with the voltage drop by means of transformer taps, including possibly a load tap changer if necessary.

But your transformer may be a bit small if your load includes this big AFD.

Maybe it's time to call in some grey-haired folks who are used to dealing with this type of situation.

 

[Marke]

On thing to be careful of is that the displacement power facor of the VFD will be better than 0.95, but the distortion power factor of the VFD may be in the order of 0.7 - 0.8 depending on the input circuitry.
If the VFD causes a 2.5% voltage drop while running, there will be significant voltage waveform distortion and this may cause other problems.

The starting current of a pump on a VFD should not exceed the running current except where there is extremely fast acceleration, so the issue is how the supply performs with the VFD under full load conditions.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[HamidEle]

DPC,

What I assuemd is the infinite swing bus(voltage constant)at the utility power supply side, not on the primary of the transformer. If you add another impedance after this utility bus, you will get a higher voltage dip because of S-C impedance from the utility.
We are in the process of investigating how weak the power supply is because our loading to see what the utility can do for us.
Tap changer cann't help for the primary side of the transformer.

 

[dpc]

I can't model what I don't know about. The impedance you add has to have some relevance to your system and requires adjusting the utility impedance to compensate.

There's no way you are going to have 20% voltage drop for a normal running condition.

With all due respect, I think you need to get some direct assistance from a EE with some experience doing power system studies.

 

[HamidEle]

DPC,

Thanks for your suggestion.

The actual situation is, the power supply is so weak that we will see a higher voltage dip at the Tie-in side.

We will talk to the utility guys and get them to look at the possibilty of boosting their supply voltage level.

 

[motorspert]

What about a synchronous motor started by an AFD? The AFD will give a voltage dip during starting, but if the starting does not oiccure too often some utilities will allow a bigger drop

Once synchronised the synchronous motor will stiffen the system, you can specifiy the motor to have a leading power factor and maybe sell them some vars if the system is that weak. It will cost more than an induction machine but will be much cheaper than any improvements to their system.

If they wont allow the AFD, you could use a pony motor and a fluid coupling with a synchronous motor.

 

[HamidEle]

Motorspert,

In our case, we don't have any synchronous motor.
We won't have any motor starting problem with VFD. The only problem is the voltage dip during normal running. We are trying to convince the utility company to elevate their supply volatge to suit the cutomer's need.

 

[ScottyUK]

Sounds like raising the voltage is a bodge to overcome an undersized utility source. What happens when the load comes off and the voltage to the remaining loads rises? Will these loads operate normally at the higher voltage?


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[electricuwe]

You should consider that there a different types of VFDs, behaving different during start up. Especially in the power range under consideration VFD with current source topology and thyristor frontend have been quite common. These topology might create problems during start up (because of low fundamental power factor at start up). With a voltage source VFD there should be no problem provided you may start up very slowly if necessary.

 

[avante]

VFDs can produce up to 150% of the motor FLA during start up. This is dependent on the amount of torque needed at start up. What kind of VFD are you using?