Dear shiyang, Thanks for your contributions. In fact I had thought in the direction of your first option. During the grid connection I allow the grid supply to be the only slack bus. The five generators bus location defined as PV bus. Then in Islanding I re-define the bus as the slack bus since the grid connection is now lost. I will have a look at other options you suggested. Once more many thanks. Idris
>-----Original Message----- >From: [email protected] [mailto:bounce-39129280- >[email protected]] On Behalf Of ??? >Sent: 10 January 2012 14:46 >To: 'MATPOWER discussion forum' >Subject: Re: AC power flow for multiple generators > >Dear Idris Musa, >As far as I know version 4.1 explicitly allows more than one slack bus >(bus it still cannot deal the situation when the Q output of any slack >bus reach its limit). 'pfsoln.m' defined how to distribute the >generations. From it we know Matpower assigned all active network loss >undertaken by each slack bus to its first generator (version 4.1 make >this available for all slack buses). >I agree with you that the situation you mentioned is not physically >meaningful. Thus, as my opinion: >Firstly, with the grid connected you can define the PCC bus only as the >slack bus, while without the connection, define the generator bus in >this island as slack bus (the reason to define only one slack bus is >that the electrical connection between PCC and the other slack bus is >usually not weak enough, so setting their voltage angles both as given >conditions is usually not reasonable -- how do you know the proper >angles difference between the two slack buses? Absorbing P may be a sign >that the setting angle of the slack bus is too small.); Secondly, if the >P loss is large and you don't think assigning it to the first generator >is sound, you can change the code of 'pfsoln.m', distribute P loss >equally to each generator on the slack bus; Thirdly, the premise of that >your slack bus P generation is reasonable, is that all other generators' >P settings are proper. Negative P generation may imply others' P >settings don't fit the real sistuation. Especially when you consider the >island re-connect to the grid, I think decreasing the P generations in >the island is quite normal. > >Shiyang, Li > >-----邮件原件----- >发件人: [email protected] >[mailto:[email protected]] 代表 Idris Musa >发送时间: 2012年1月10日 20:15 >收件人: 'MATPOWER discussion forum' >主题: AC power flow for multiple generators > >Dear all, >I want run AC power flow, considering multiple generators with >possibility of Islanding using MATPOWER 4.1. >In addition to the grid supply I have five generators of equal output >capacity at particular bus location. I defined this bus location as >slack bus. This means I have two slack buses (including the grid supply). >When I run MATPOWER AC power flow, with the grid connected, one of the >generators (the first generator of the five connected) is taking care of >system losses, showing that the chosen generator has unlimited capacity >and is also absorbing negative active power. I am assuming a distributed >generator (which could be synchronous generator, fuel cell etc) is this >physically feasible? I had expected the grid supply been the first slack >bus to take care of the system losses and provide other support for the >network. I would be glad if someone could advice on how to go about it. > >Idris Musa > > > > > >
