See my individual responses below … > On Feb 27, 2015, at 8:45 PM, Lu, Yueyun <[email protected]> wrote: > > Hi all, > > I want to study the effect of multiple line contingency by recursively > solving pf. The procedure goes as follows: start with some initial line > outage, repeatedly solve dcpf, disable the lines that exceed limit until no > more overloaded lines. I have some questions when implementing this: > --- I disable the overloaded lines by setting their status to 0. If they > disconnect the network, then island emerges. How to determine slack bus in > each island? Can I arbitrary choose one bus to be slack and will this choice > affect pf?
You can arbitrarily select any generator bus in the island to be the slack. Let me mention though that if an island does not have any generation, then all of it’s buses should be treated as isolated, since there are no gen buses to take up any slack. Your selection of slack will affect the power flow. If you do not apply any changes to generation or load and there is an imbalance in the island, then all of the imbalance will fall on the slack generator. If you do adjust generation and/or load to restore the balance, then the choice of slack bus should not make a difference in a lossless DC OPF. In an AC OPF it would be affected slightly by the amount of your error in your loss estimate. > --- If power is not balanced in an island, meaning that the generation is not > equal to demand, I’d like to apply some control to retain balance by loading > shedding or AGC. Is it fine if I directly modify PG and PD value in the case > file of that island? Yes. > --- What runpf does with the case when power is not balance? As we know, pf > is solved by P=B\theta where B is Laplacian. And the sum of all the entries > of P needs to be zero. While if not, runpf will still return some result. How > to interpret that result? That’s what the slack bus is for. The power at the slack will be set to make the total generation and load for the island to be in balance. -- Ray Zimmerman Senior Research Associate B30 Warren Hall, Cornell University, Ithaca, NY 14853 USA phone: (607) 255-9645
