If you are running an OPF, and it is converging successfully, then the initial values of mpc.gen(:, VG) or mpc.bus(:, VM) are not critical. The important thing is which constraints are binding at the solution. Are there binding reactive power limits?
-- Ray Zimmerman Senior Research Associate B30 Warren Hall, Cornell University, Ithaca, NY 14853 phone: (607) 255-9645 On Feb 11, 2014, at 11:38 AM, angelina sirri <[email protected]> wrote: > Yes, I am using dispatchable loads and as far as the costs, grid infeed is of > free cost, distributed generation is more expensive and the energy not > supplied generators are of the highest cost. When there is energy not > supplied the voltage in that nodes reaches the maximum upper limit (1.06) . > When I change the voltage limit, the dispatch is again the same and the > problematic nodes reach the new maximum upper limit. I think that the problem > is the initial voltage of the gens in mpc.gen(:,6). What values should I put? > The system originally had only load buses( with voltage less than 1 ). Now > that I have transformed those buses to gen buses I should predetermined the > voltages of the gens. I tried to put the same voltage as the mpc.bus (:, 8). > But these values seem to be not so accurate in the end. How should I evaluate > what values to put in the mpc.gen(:,6)? > > From: [email protected] > Subject: Re: reluctant energy not supplied in a distribution system with opf > Date: Tue, 11 Feb 2014 11:14:31 -0500 > To: [email protected] > > I also suspect it is a reactive power or voltage problem. Is there a voltage > limit binding at the buses where energy is not supplied (I assume you are > using dispatchable loads)? If so, you might get some insight from doing some > perturbation analysis. Simply change that voltage limit by a very small > amount and see how the dispatch changes. My guess is that the only way to > maintain the voltage is by using the expensive distributed generators. > Shifting generation to the grid infeed would likely result in more load > shedding to keep voltages in line. Presumably load shedding is even more > costly than the distributed generation? > Kind Regards, > Angelina > > -- > Ray Zimmerman > Senior Research Associate > B30 Warren Hall, Cornell University, Ithaca, NY 14853 > phone: (607) 255-9645 > > > > On Feb 11, 2014, at 5:48 AM, angelina sirri <[email protected]> wrote: > > Dear Dr Zimmerman, > I am using an 11kV distribution system, which is consisted of two radial > circuits connected by a normally open point. The network has one grid infeed > (Vg=1) whereas all the other nodes where occupied only by loads (with Vm<1). > Now I increase the loading of the circuits and have added distributed > generators in all the load buses, which will only operate in case there is a > line fault in the system(otherwise there is no need for them to operate > because the grid infeed can give really high power, comparing to the load and > the lines rating is also really big) . For the voltage problems, I have > added reactive power in each node. > The thing that confuses me is , that when there is a line fault (the worst > case scenario line loss), there is always energy not supplied , although a > path from the main supply exists and also the rating of the lines allows > that power to pass through the line. This happens mainly for the most distant > nodes. In that case, all the available distributed generators generate at > their Pmax , but the grid infeed value is really small. > Why the opf prefers to use the really expensive generators of the distant > nodes (used for the energy not supplied) instead of absorbing that power from > the free of cost grid infeed? I can guess that is a reactive problem or a > voltage problem but I cannot find a suitable solution. > Any help would be of high importance, > King Regards, > Angelina
