Revision history [back]

It seem your problem is a UVLS study in disguise. Check the post "How to calculate the best load shedding bus?"

In a dynamic simulation, add UVLS relays to all loads in the study region (model LVSHAR for area x, use a single stage) with generic settings for the voltage level to trip, the time delay before tripping and the fraction of load to be shed. (0.9 V pu, 3 sec, 0.05).

Let the UVLS relay model find those buses with low voltages during the simulation. Then load will be shed (5% of total bus load) and the simulation continues. At the end, several buses would have load shed. Verify with a voltage scan (vchk activity) that the lowest voltage is acceptable. If not, repeat the run with a higher fraction of load to be shed, say 10%.

Local buses would have different voltage thresholds or time delays tolerances to activate the relays. A more realistic UVLS study would require to vary all three parameters (v, t, f) in 3 to 5 stages model customized to the bus load requirements, all these tested under several extreme contingencies and network conditions.