IEEE Std. C37.102-2006 Voltage Restrained Relay for Short, Short-Circuit Time Constant

[recs]

In IEEE Std. C37.102-2006 related with generator protection, it is recommended to use voltage-restrained back-up protection when the following condition is present (Section 4.6.1.2 Overcurrent Type Backup, page 77): Short, short-circuit time constant. The actual wording is:

"On units that have a short, short-circuit time constant, the 51V voltage-restrained overcurrent relay should be used."

Does anybody has a clear idea of why IEEE recommends this?

Thanks.

 

[dpc]

In general, using a voltage-restrained or voltage-controlled overcurrent relay allows the relay to operate at current levels below normal full load current. The generator fault current decays over time, eventually decaying to the current associated with the synchronous reactance of the generator (assuming a constant excitation). The synchronous reactance is often greater than 100%. If the fault current decays quickly enough, a normal overcurrent relay will not operate since it must be set higher than the generator rated operating current.

 

[collies99]

As an alternative one could also use impedance torque control.

51V has a large reach which can be use for backup to the transformer
overload protection while the 21 torque control can limit reach up to
the transformer bushing.

 

[rcw retired EE]

Typical impedances for a generator might be Xd= 2.00 pu, Xd"=0.15 pu.
The Xd synchronous reactance determines the steady state fault current. The Xd" sub-transient reactance determines the fault current for the first few cycles of the fault.

Using the values above, the long term short circuit = I=V/Xd = 1/2.0 = 0.5 per unit. But the fault starts at the instantaneous fault level of I= V/Xd" = 1/0.15 = 6.67 per unit.

For a solid three phase fault the current will start at 660% of full load and decay to 50% of full load. The transition between these two extreme values is determined by the generator's short circuit time constants and the response of the excitation system.

To graph this generator's short circuit current on time-current curve, put a mark at 660% current on the bottom axis and a mark at the top axis at 50%. Draw a "knee" curve between the two points similar in shape to a thermal breaker curve. If the generator has a long short circuit time constant, the knee will be up in tens of seconds range or higher. For small generators with "short" or fast short circuit time constants, the knee will be down in the seconds range, meaning the generator’s short circuit output will rapidly decay from 600%+ to less than 100%. (Most coordination programs can graph this curve given the generator data.)

The time-current tripping curve for a standard overcurrent device or relay set at 125% or higher pickup will never intersect the generator output curve unless the generator has a long time-constant with a "high" knee. An overcurrent relay set high enough to allow normal load may never see the fault as current decays to 50% of full load current.

The standard solution is to apply a 51V voltage controlled or restrained relay with a pickup of 50-80% of full load. The overcurrent element is inactive unless the voltage also drops below set point as it would during a fault.

If the generator has a long time constant, it may be possible to set a regular 51 overcurrent relay or low voltage breaker to intersect the generator curve and provide protection. Otherwise, a 51V, or 21 impedance protection is needed. 51 V is cheaper.

 

[recs]

Thank you for your responses. But I might not have clear on the purpose of my question.

I understand and agree with all what you wrote. But the questions is:

Why the IEEE determines that a Voltage-Restrained relay (not a Voltage-Controlled Relay) is advisable when the time constant of the short-circuit current is low.

Thanks.

 

[rcw retired EE]

Voltage restrained vs controlled?

I believe the restrained function acts akin to a distance or impedance relay. A remote fault will depress generator voltage less than a close-in fault. That makes the relay faster on a close-in fault and more time delayed on remote faults, possibly allowing better coordination with remote breakers.

 

[recs]

I am confused how this answer justifies a voltage-restrained relay better than a voltage-controlled relay?

 

[dpc]

In a voltage-restrained relay, the time current curve becomes more sensitive as the voltage decreases, so for very low voltages it operates much more quickly - the voltage changes the shape of the curve. A voltage-controlled relay is simply supervised by a contact from the voltage relay coil. The shape of the curve is not a function of voltage.