Standards on providing motor differential CT's ? 3

[sparky1976]

Guys,
Is there a standard for manufacturer to provide differential CT's on big motor (350HP and above) or just the custumer ask them to provide that ?
Please advise.

Thanks
Pitat

 

[gordonl]

Optional, at customers request, from my experience.

 

[busbar]

In support of gordonl's comment, ANSI/IEEE C37.96-1988, [IEEE Guide for AC Motor Protection] there doesn’t seem to be any recommendation with respect to motor size.

 

[coolfish]

I believe it is optional. NEMA standards do not require them, especially for 250Hp, which is is still 'small'. Even RTDs are not required, especially bearing RTDs.

 

[electricpete]

Sounds like good input so far. Apparently it is not a standard requirement.

Perhaps it's worth discussing briefly why the extra cost/complexity for differential protection may not be justified even for very large motors: (our plant has 8000hp motors without differential protection... uses instantaneous ground fault, instantaneous phase overcurrent, time overcurrent, current balance... all fed from the switchboard phase ct's)

- Phase to ground fault can be detected with very high sensitivity by instantaneous ground relays wired off the switchboard ct's (no extra ct's at motor required).
- Turn to turn fault cannot be detected by differential or other methods, so there's no advantage to install differential for turn-to-turn
- Phase-to-phase faults are the one type where differential protection provides a faster and more sensitive reponse than the above-mentioned protection.... but phase overcurrent will eventually protect this fault also.

Remember that ground faults are by far more common than phase-to-phase faults. And there is logical reason for this.... ground fault needs only one flaw in groundwall... phase to phase needs two flaws (one in each phase). Also phase to phase has twice the insulation but only 1.73 times the voltage.

So for the extra expense of extra set of neutral ct's and more complex relays and wiring, all you get is slightly better protection for a fairly rare fault (motor phase-to-phase).

 

[busbar]

Discussing motor-winding differential protection, C37.96-88 §2.4.7 states: “These [87] relays detect low-magnitude fault currents during normal loads and do not trip falsely during high-magnitude external faults, or during starting periods. Differential relays cannot detect turn-to-turn faults in the same winding.”

For ground-fault protection, one of the illustrations in C37.96 includes with the 6 CTs and 3 percentage-differential devices, an “87N” directional-overcurrent relay polarized by neutral current or voltage.

I think also that it’s important to note that §3.2 says: “The complete protection scheme must be chosen to achieve optimum service reliability, safety, and protection of equipment at a reasonable cost. It is essential that the operating characteristics of the chosen motor-protection system be coordinated with those of the supply and, if necessary, the process or load involved if selective operation of the protective and control devices is to be achieved.”
That seems to imply that the process loss can potentially be many multiples of the initial motor and associated protective-relay price. On the other hand, these relays must be included in the maintenance routine, so they are not “free” once installed. The driving force may be the cost of lost production for a motor outage, regardless of it’s size.

 

[BJC]

Have you considered a " self ballancing differential protection scheme? This uses three CTs at the motor. Both ends of each winding go through a CT and the "star point" is made up out side the Cts. Check the Buff Book and some of the GE literature for details.
It's cheaper and the relays are really ground fault not differentials.

 

[electricpete]

BJC. That's a good point which I overlooked when I was talking about cost/complexity. Of course it still requires the ct's at the motor (vs switchgear) to measure neutral current, and the wiring from those ct's to the relaying at the switchgear. Maybe not a bad investment to gain more sensitive protection for phase-to-phase faults.

One other factor I'd like to add on to my earlier comments: although phase-to-phase fault is less frequent, they do occur from end-turn contamination and other factors... and the motor damage can be more substantial since the source path impedance is much less than for ground fault which often must go through generator or transformer neutral grounding transformer or resistor (depending on the installation).