Resonance - the cause of failure of MOVs ?
Posted 19 June 2006 - 04:21 AM
I have an electrical main switchboard with shunt connected MOV type surge protection devices (SPD) at the service entrance. The power supply transformer (2 MVA) is within the same building, in fact, in the next room some 50 ft away. Bulk automatic power factor correction capacitors are also installed at the switchboard - these are made up of 10, 20, 30, 40 and 50 x 6 kVAR capacitors bank (total of 10) switched in automatically by the pf controller. The cap. bank are not installed with detuning reactors.
The problem now is that it had happened twice within the last 2 years that the SPDs (of reputable make) have failed prematurely - the units burnt out.
The switchboard serves a few air-conditioner chiller plants - notably a 600 RT centrifugal chiller + 2 nos. 200 RT reciprocating chillers. I believe from the nature of the connected loads that harmonics is not an issue.
However, I do expect switching transients from the contactors in the chiller plant auto-transformer starters, and starting of chiller plant and ancillary motors.
I suspect resonance to be the culprit but :
(i) how could I prove this ? What are the instruments to use and how to go about doing a study...
(ii) if harmonics is not prevalent, then what other disturbances can cause resonance.
(iii) how does resonance cause the failure of MOVs ? What is the mechanism - Joule loading ? Transient overvoltage should not 'kill' MOVs because that is what they have specifically been designed to withstand and mitigate.
Please help. Regards.
Posted 19 June 2006 - 07:43 AM
Welcome to the forum.
The problem you are experiencing is very likely to be due to supply resonance.
Supply resonance is a gbreater problem when the load is light than when the load is high.The ower factor correction capacitors form a resonant circuit with the inductive reactance of the supply. This reactance is due to the leakage reactance of the transformers and the reactance of the supply connected to the primarly side of the transformers.
The impedance of the supply is another important factor in dampening the resonance.
A low supply impedance will have much lower resonance problems than a high supply impedance. The worst situation is a transformer that is fed by a very long overhead line.
Power factor correction is used to reduce the maximum current drawn from the supply. Therefore, the important time for correction is when the load is maximum, and at maximum load, the resonance is minimised.
I recommend that if you have a resonance problem, defeat all the correction when the load is less than say 20 - 30% of the maximum load. This way, the connected load will help to dampen the resonance and reduce the problems.
When there is a supply resonance issue, and the load is light, the supply resonance will cause a surge in supply voltage. If you connect some form of disturbance analyser on the supply and then switch one capacitor bank on and off, you will see a voltage surge triggered by the capacitor switching.
I have measured ringing voltages over 2000 volts on a 400 volt supply. I have seen these resonances last for seconds in some installations. The resulting overvoltage has two mch energy to be clamped by MOVs and so they will fail.
Try to avoid having correction connected when the supply impedance is high and he load is light.
Mark Empson | administrator
Skype Contact = markempson | phone +64 274 363 067
LMPForum | Power Factor | L M Photonics Ltd | Empson family | Advanced Motor Control Ltd | Pressure Transducers | Smart Relay | GSM Control | Mark Empson Website | Soft Starters
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users