I have to design a 3 phase Reactive Power Compensation System with using only reactors, no capacitor is needed. Total capacity is 505kVAr, but 475kVAr will be done with contactors, rest (30kVAr) will be done with Static VAr Compensation (SVC). This 30kVAr reactors are established using 3 monophase 10kVAr reactors (because of the SVC driver manufacturer told as like this.).

For measuring the power line, an Energy Analyzer is used. It gives the VAr power and PF, rather then a measuring with a Reactive Power Relay (since some special request from the costumer). Rest of the system is an SVC driver which is connected to 3 monophase reactors and a relay which trigers the SVC driver.

Hence, I have to design a special Reactive Power Relay, which takes the data from the EA and triggers the driver via trigger signals. My relay is able to provide any different analog output signals in addition to PWM via DACs at very high speeds and 16bit resolution for ±12V, or ±24mA outputs whichever acceptable if needed.

First problem is, I see on the driver catalogues, SVC drivers are triggered via PWM signals (as opposite to contactors, it looks like a Pulse Train). These trigger signals are shown as consecutive sharp PWM pulses, looks like there is a 120° delay between the different phases same as VAC for three different phases (like a sawtooth or just a pulse rather than a square). Hence, I think my SVC relay must provide PWMs. However, SVC relay and driver manufacturers telling me I cannot make such a relay, it is not that easy! They say that the trigger signals are not PWM or any square waves with different duty cycles. Ok, than the first question is what is the format of the trigger signal? Is a trigger signal sequence needed for 3 phase? What is the format of this sequence and how it is determined?

Second problem is, I use only reactor banks, there is no capacitor bank on the system. Second missing point occurs here; some driver manufacturers tells that the relay have to follow the zero crossing point, however some tells not! Because of this reason, some tells I have to add a 3 phase 25kVAr capacitor bank to my system. Do I really need to follow the zero crossing point for triggering only the reactors? Should I add a capacitor bank?

You might check the attached pictures for the Schematic and Trigger Signal given on SVC Driver catalogue.




I have tested my system with smaller reactors, my observations are as follow:

  1. Varying any analog value (mA or V) could not control the amount of the reactive power drived by the SVC driver.
  2. SVC driver could be triggered with different voltage or current values, i.e. 3V, 5V, 10mA etc. There is no difference at all. It totaly ON/OFF the reactors.
  3. Important thing is the Duty Cycle and Triggering Signal Frequency. I have tried triangular, square, and some other waveforms, they all open and close reactors according to reaching the threshold voltage (which means I can control SVC driver like fast switch/contactors).
  4. There is no zero-cross issue. I can drive/ON/OFF the reactors whenever I want and in any order. Hence, I do not need any Capacitor Bank.
  5. I also observed the SVC Trigger outputs of the Reactive Power Relay. It gives pulses (via PNP BJTs) about 100Hz and with ~10% duty-cycle (Please look at the screenshot I upload). I see that the SVC Relay could not operate as desired, who could try to check it yeah!

The only thing remains is the algorithm! It would be challanging, but I will also update this post for the next steps.

Reactive Power Relay, Triggering Signal Output to the SVC Driver, Oscilloscope Image


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.