Surge Arrester Rating for a 138 kV system?? 1

[sky2020]

For 138 kV system, solidly grounded, we choose an arrestor with MCOV > 138/sqrt(3)*1.05 = 83 kV.
For ungrounded system, we choose an arrestor with MCOV > 138 *1.05 = 145 kV.
What about for LRG or HRG systems? Do we choose the arrestor as an ungrounded system? i.e. MCOV > 145 kV?

 

[davidbeach]

Yep.

 

[Kiribanda]

1)For an HRG system yes. MCOV=145kV, because the system is not effectively groundded. Therefore COG=1.
2)For a LRG system since a considerable ground fault current is allowed to flow during a ground fault,
the system is effectively grounded (COG=80%). Therefore, MCOV=83kV.

 

[davidbeach]

Even with a LRG system you can get a nearly full neutral shift and nearly full phase-phase voltage to ground on the unfaulted phases. I wouldn't use a 83kV MCOV arrestor on a LRG 138kV system, but that's me.

 

[bacon4life]

The duration of the voltage shift matters. Even if the LRG and HRG systems have the same magnitude of overvoltage, a system that trips in cycles will stress the arrestors much less than a system that allows a fault to persist for longer periods (seconds, minutes, hours?). For unusual grounding conditions, you can review both the magnitude of the overvoltage and the duration as compared to the arrestor TOV curve. It is possible you may end up with an arrestor somewhere in between typical recommendations.

Some arrestor guides offer multiple rating options for each voltage. For example

https://www.nemaarresters.org/understanding-arrester-datasheet/

suggests 84 kV or 98 kV for solidly grounded 138 kV systems and either 132 kV or 144 kV for delta 138 kV systems. Both 84 kV and 144 kV seem like extreme choices for LRG systems.

 

[EE123]

Not to hijack this tread, but I have a similar question with different voltage. We have a 12kV campus loop and the question came up regarding sizing of the surge arrester at a 12kV-480V step down transformer. There was debate on whether to size the arrester at 9kV ((12kV/1.732) * 1.3) or 12kV (phase-phase). Most articles talks about sizing the arrester in a delta distribution at phase-phase voltage. I believe the rational is that if there is a ground fault, the the phase voltage will be phase-phase voltage. However, with the relay protection of the feeders in a campus loop the overvoltage setting is at most 130% and if there is a fault the relay will operate before the phase voltage get anywhere close to the phase-phase voltage. Would it not make sense to have the arrester at lower rating (9kV) to provide better protection than to have to have it at 12kV? Any feed back and comment is much appreciated.

 

[davidbeach]

So, I've seen the results when 8kV (ish) arrestors were applied on a 13kV system with low impedance grounding. SLG fault on an adjacent feeder (same substation transformer) and the neutral shift did in the arrestors faster than normal clearing time for the fault. Two sets of arrestors failed. The failure of the first one went 3LG, shorting out the original cable fault; those overcurrents dropped out. After the feeder with the first set of failed arrestors cleared, the SLG fault returned and the whole process repeated itself on the second set of arrestors. Made a huge mess; wouldn't take my chances on anything less than phase-phase rated.

 

[bacon4life]

EE123-Even if the campus loop relays are set to trip at 130%, this does not limit the voltage to 130%. When a ground fault occurs on a delta system, the voltage immediately jumps to the phase-to-phase value, and stays there for multiple cycles until the circuit breaker has operated.

The analysis required for using something less than the LL rating requires extensive evaluation of the possible overvoltage conditions. I assumed sky2020 was asking from a utility perspective where they have control over all aspects of voltage coordination. Customer loads generally do not enough control over the upstream equipment to allow the same sort of analysis.