Negative and Zero Sequence Protection Applications

[111R]

For power transformer protection of a standard Delta-wye radially-fed transformer, is it normal to utilize negative sequence protection to gain sensitivity? What is commonly recommended for settings sensitivity to avoid inadvertent trips? As in, should it be set to pickup at 50% of expected maximum load current or something similar?

Where else is it common to see negative sequence protection and where is it a good idea to avoid? I don't see this utilized very often.

For zero-sequence protection upstream from a radially connected transformer, I've seen very sensitive settings applied. The Delta-wye transformer should have no zero sequence current flow for low-side faults, so greater sensitivity can be achieved for high side faults. What are some things to consider before settings zero sequence elements too sensitive? Non-transposed lines, CT saturation for low-side faults, ??

Thanks

 

[ScottyUK]

Never seen dedicated NPS relay on a transformer application - although I may learn something in this thread - but they are commonly used on synchronous machines because the negative sequence currents circulate in the rotor forging at double line frequency causing severe heating of the non-laminated rotor core.

 

[HamburgerHelper]

I know you can do it but I have never seen it. You have to coordinate it with both phase and ground overcurrent curves. Here is a SEL paper that talks about it.

https://cdn.selinc.com/assets/Liter...4_NegSeqOvercurrent_Web.pdf?v=20151204-094653

The ground overcurrent on the high side shouldn't be set too sensitively. You don't want to dump the transformer due to CT error or saturation.

 

[prc]

In Transformer differential relays negative sequence current is used to detect sensitive turn to turn faults. Zero sequence voltage in unearthed systems (tertiary delta loading auxiliaries) is used to detect line to earth faults.

 

[veritas]

Many years ago I've used nps oc on radial feeders to assist with the detection of downed conductors/open circuits (conductor theft was rife!). No trip - only alarm to operator. Its success depends on the amount of load current flowing at the time of the open-circuit. Higher the load current, the higher the amount of nps current. NPS is less effective with interconnected networks. Open-circuits on these lead to more zps than nps currents. Think of hot-connections with the open-circuit being the ultimate hot-connection.

Great caution would need to be exercised if nps is to be used for tripping as co-ordination is required with the OC relay for unbalanced phase faults as well as with the EF relay for earthfaults.

I recently used nps diff for the first time in the SEL-487. As it is load insensitive it can provide improved sensitivity for unbalanced faults in the diff zone. Will see how that goes! Was quite a head-scratching exercise to test the nps restraint characteristic but we succeeded.

When it comes to zps on the HV side of a trfr I normally use a two stage definite time characteristic. Stage one set to around 10% of CTR with a 350ms time delay. Stage two around 5x Stage 1 pickup with no intentional time delay. Has never come back to bite me yet! Of course, this only applies to a delta winding or star unearthed.

As Scotty pointed out nps is more commonly used on generator or motors. I've applied them to both numerous times and have had them operate legitimately on a few occasions due to single phase conditions. Note that these are usually thermal nps functions and not just plain nps OC.

Hope this helps.

 

[HamburgerHelper]

If you set your alarm on a ratio of I2/I1, you can get away from the load dependency that you have with just looking at I2 to detect open phase.

 

[dpc]

Some discussion of directional I2 overcurrent protection for transformers here:

https://cdn.selinc.com/assets/Liter...r_AG-NF-CL_20090129_Web.pdf?v=20150812-083827