HARMONIC FILTERS & PF IMPROVEMENT CAPACITORS 1

[deeveeyes]

I have a Power Control Centre (415V), from which a feeder feeds around 500 kw of load of Variable speed drives (6-pulse). Another feeder from this PCC feeds a Motor Control centre having 300 KW load. To improve the PF of motors, capacitors are connected across the motors directly.

We are planning to instal a filter(4.7th harmonic tuned) on PCC to absorb harmonics from non-linear loads. However the filter supplier insists that we disconnect the capacitors connected across the motors as otherwise he cannot guarantee performance as per IEEE standard for Harmonics. What should I do ?

 

[jbartos]

Suggestion: Follow the supplier advice and have them the tuned filter to be guaranteed, and the product and load safe functions be part of the quaranty too. The tuned harmonic filters are using capacitors. The existing capacitors might potentially stay where they are; however, they may cause the design of the tuned filter more difficult; especially, if the supplier is using some "tuned harmonic filter design" software.

 

[jack6238]

The problem the filter designer has is that if you leave the capacitors installed on the motors, then every time a motor is switched on/off line the resonant point of the system changes. The resonant multiple of the system is calculated as ( MVAsc/mvar caps)^.5. I agree with the manufacturer, eliminate the local capacitors and I wonder why they were installed in the first place. Location of capacitors at the motor terminals may be best for the system but it is by far the most expensive solution to correcting the system power factor.
Remember that the purpose of the filter bank is to trap the system harmonics in the series connection of the tuning reactors and the power factor correction capacitors. Those capacitors installed throughout the site will cause numerous problems. A word of warning. Make sure that the capacitors used in the filter design have the capability to handle the increase in voltage due to the reactors. Failure to account for this will doom this installation to early failures from capacitor overvoltages. Also the capacitors must be able to handle the harmonic currents. I'm sure the designer is aware of these watch outs.
There are numerous references on this topic but one of the best is the McGraw-Edison-now Cooper Power Systems- book on Capacitors. See also the information in Chapter 10 of the book Power Capacitor Handbook (1984) available from Butterworths

 

[rhatcher]

Jack6238's post is very good with respect to the posted question (he got a star from me). The only thing that I can add is with respect to the individual motor overload protection. Right now the motor magnetization current is being supplied by the capacitors attached across the motor terminals and the overloads should be sized for the remaining load current (a value less than the motor full load nameplate amps). When the capacitors are removed, this current must be supplied from the line. The result is that the overloads must be increased in size to match the full nameplate rating of the motor. As well, you must be certain that the leads from the motor control center to the motor are rated for the full motor current.

 

[deeveeyes]

Thanks everyone for lucid explanation.

The filter manufacturer has again reverted to us telling that as long as the total quantum of capacitors being switched in along with the motors is below 8 - 10% of Proposed Filter Bank capacity, he does not foresee any problem in meeting IEEE limits of THD(I) AND THD(V). Alternatively he proposes that the motor feeder is isolated from the main Power control centre by means of an interposing transformer or there is sufficient inductance in the cable connecting PCC and Motor control Centre.

 

[jbartos]

A small encore from Reference:
1. Beeman, D. "Industrial Power Systems Handbook," First Edition, McGraw-Hill Book Company, Inc., 1955
Reference 1 Section "Selection and Application of Motors and Capacitors" on page 474 indicates "A recent development is the practice of connecting the capacitors directly at the motor terminals in order to permit switching the capacitors and motor as a unit. A typical installation is shown in Fig. 8.19. This unit arrangement is desirable because of the capacitors are always on when the motor is in operation. The power factor of an induction motor is quite good at full load, usually between 80 and 90%, depending on the motor speed and type of motor. At light loads, the power factor drops rapidly, as illustrated in Fig. 8.20. Generally, induction motors do not operate at full load, resulting in a low operating power factor. Even though the power factor of an induction motor varies materially from no load to full load, the motor kvars are essentially constant. This characteristic makes the induction motor a particularly attractive capacitor application; with a properly selected capacitor the operating power factor is excellent over the entire load range of the motor, as shown in Fig. 8.20, generally 95 to 98% at full load and higher at partial loads. The reason the power-factor curve with capacitor is so flat over the entire motor-load range is that the net kvar is low and varies only a little with load. The addition of capacitors for power-factor improvement does not change the motor performance characteristics, as the operating speed and shaft output depend upon motor load and applied voltage." Further down is stated that capacitor location at the motor terminal is advantageous since the capacitor and motor work as a unit, the capacitor is on according to motor needs, a separate switching device for capacitor is eliminated, etc. Apparently, the resonance issue is not a problem by a proper selection of the capacitor. Jack6238 post on Mar 24, 2001 somewhat overlooks that the total kvar of the paralleled capacitor with the motor inductance is smaller, therefore, this poses smaller problem than the motor inductance and the larger motor-inductive kvar.

 

[ICEMAN]

I have been in drives and harmonic mitigation for some time, the latter both passive and active. I would NOT connect standard passive filter to your Power Control Centre as this could cause resonance ih future. In addition, the use of standard passive filter vary rarely comply with IEEE 519 (1992).

It is also better NOT to have individual power factor correction on each fixed speed motor. Much better to have group correction with a detuning reactor to prevent damage to the capacitor bank due to harmonics from the VSD load.

A better solution for the VSD load would be to look at active harmonic filters (which are expensive) or look at using a Mirus Lineator patented passive specifically for 6 pulse AC diode front drives.

If you want any additional information re drives and harmonics please use my email address ; iancevans@vsdtechnicalservices.fsbusiness.co.uk

I hope the above helps.

 

[jbartos]

Comment to the previous posting: Please, would be able to compare efficiency of central power factor correction with respect to local power factor corrections at individual motors, considering that each motor will draw higher currents for the central power factor correction scheme; therefore, RI**2 will be higher with all ramifications of RI**2.

 

[ICEMAN]

You are right, but the difference is marginal. If the motors all had individual pfc there is significantly more chance of reasonace and other harmonics problems due to the 500kW VSD load than if we had only group pcf with a detuning reactor, designed for the 4.7H, for protection.

The efficiency of the power factor here is not an issue. The effects of the harmonics, direct and indirect is.

Harmonics and capacitive pfc do not go well together.

 

[jbartos]

Suggestion: For more accurate statements about harmonics, one would have to know more about VSD principle, and its accessories (harmonic filters, reactors, etc.).