Hello, thank you for your interest! I modeled the transformers as simple branch, referring the impedance to te primary coil. Looking at model of branch tap ratio=1 should be ok because it means that there is no variation of the voltage amplitude. I marked all the the generators as PQ busses because the are photovoltaic generators and I’m considering the voltage rise that these kind of generators introduce in the network. The problem of convergence in actually due to the high value of the susceptance of the branch. I got the values from a public utility I’m working with, but I think that the the susceptance should be divided for 100 based on some checking of MV cables catalogs. With the modified values of the susceptance the load flow converge and the results are close to the ones I expected.
Best regards Simone Il giorno 22/ott/2014, alle ore 22:46, Sohail Khan <[email protected]> ha scritto: > Hi Mr. Cochi, > > I have tried to understand the files you sent. > > Regarding mpc.bus data: > > It is showing that all the buses are on the same level 20 kV. But if there > are transformers for example from 20 kV to 480 V then the LV side of > transformer should also be a bus? > > Regarding mpc.branch data: > > In the 9th column of mpc.branch in the section of the ratio. The value > represents the tap settings of the tap changing transformer. It seems that > there is a transformer in every branch. If there is no tap changing > transformer then the this value should be kept 0 in my opinion. Further the > formula of tap in the documentation apply only when the r & x of the branch > are 0. > > In mpc.gen data: > > There are 4 generators. 1 being the slack. > The others are on 26, 27 and 187. So are you taking the bus as PV. If so then > the corresponding bus type in the mpc.bus should be 2? > > please check if the 12 should be same as 09. ? > > > %% By making the above mentioned changes the "rundcpf" is successful. > Following command was used to generator the result. > > mpopt = mpoption( 'VERBOSE' , 1, 'OUT_ALL' , 1); > [RESULTS, SUCCESS] = rundcpf(mpc, mpopt); > > Also the modified mpc is attached with the file. > > Best wishes > Sohail > > PhD student, > TU Wien, Austria > ---------------------------------------------------- > Dear all, > I trying to run a lf on a MV network. I treat all the transformers as > simple branch. If I use the whole case file the load flow doesn't > converge, instead if I remove the first branch, that is the HV/MV > transformer, and put the slack bus on bus 2 (MV busbar of the > substation), the load flow converge. Also if I decrease the > transformer impedance of 10 the load flow does converge. The > transformer is a 25 MVA machine with a short circuit voltage of 14,6% > circa. The case file is attached if someone want to check > <matlab_result.mat>
