Effect of VVVF drives & PF Capacitors on Power Quality

[normannicklette]

Mark

 

I work for a large Iron Ore Company and we run a Bulk PFC/Power Monitoring system. The system is generally well designed and work well. It uses Genius Power Track blocks for monitoring/control of the Cap Banks(total of around 20MVARS). We have two banks which were installed during a recent upgrade and they cause problems. This area has quite a few VVVF drives and the problem was that they would cause the DC link fuses to blow when the banks switched in. The system has time delays for switching in as well as delays so as to ensure the demand has not reduced since the close request was initiated. Some testing was done and someone came up with an idea that the banks should be locked out when the VVVF Drives were in operation and so the PLC was edited to delay closing and opening during operation. This did not solve the problem. We engaged a consultant to do monitoring but when the job was done the current monitoring did not operate so we do not have a result from this option.

 

Would you be able to give me a step by step explanation of what happens when the banks switch in or out and also the effect of the VVVF Drives being either in operation or on satndby. We have a total of 5 X 15 MVA transformers(Auto tapchanging) fed off the grid and the area where these cap banks is fed from is fed from one of these five 15 MVA transformers. Our incoming voltage is 66 kv and we reticulate 11 kv. The Cap Banks are connected to the 11 kv network of which we run three separate networks(because of the auto tapchangers). The PFC is used to correct the incoming Power Factor and is sized to match the line area loads so we generally end up with near enough to a power factor of 1 in line areas. The two cap banks in question are1.67 and 2.5 MVARS other banks in the syatem are a max value of 4.175 mvar and a minimum of 2.5 mavr. Hopefully I have supplied enough info.

 

Thanks in advance,

 

Norm

[marke]

Hello normannicklette

 

This area has quite a few VVVF drives and the problem was that they would cause the DC link fuses to blow when the banks switched in

This is not an uncommon problem and is caused by the voltage transients generated on the supply when the capacitors are switched in.

The capaciotrs and the supply inductance form a resonant circuit, and at switch on, there is a big inrush of current into the capacitors. If the supply was purely resistive, this would result in a downward voltage deflection, but because of the inductance of the supply, you will get some ringing. The ringing magnitude is inversly proportional to the supply impedance and load. i.e. high supply impedance and light load results in high transients.

The transient voltage applied to the input of the VVVFs will cause an inrush of current into the DC link capacitors and that can cause the fuses to fail, or worse still, can cause the rectifiers to fail.

One solution, is to disable the correction when the drives are operating, (the harmonics generated by the drive inputs will cause additional current to flow in the correction capacitors and premature failure) but this can result in an undesirable increase of KVARs. Sometimes, minimising the size of the steps will minimise the size of the transients and will reduce the problem to an acceptable level. i.e. more smaller steps.

You could try adding detuning reactors in series with the capacitors. This will not eliminate the resonance, but will transfer alot of the ringing voltage across the reactors instead of across the load. It will also increase the high frequency impedance of the correction network, reducing the inrush current into the rectifiers and DC link capacitors. Detuning reactors will also reduce the effect of the drive created harmonics on the capaciotrs themselves.

 

Best regards,

[theDOG]

You could also try using a line reactor of 3%z connected as close as possible to the input of each of the drives. The reactor will affectively oppose any sudden change in current, and protect the drive. Also, does the drive have MOV protection for the input bridge? If not, I would suggest adding this, as the PFCC would also be creating quite large voltage transients.

You could also look at the PFCC switching method as Marke suggested.