Optimisation of distributed generation using simplex algorithm

by Kimberley Francis

Institution: Murdoch University
Year: 2014
Record ID: 1059921
Full text PDF: http://researchrepository.murdoch.edu.au/24192/


The traditional centralised power grid was based upon the concept that central power sources would feed power through the transmission system and then the distribution system where the power would be utilised by the connected loads (ITP 2010). The concept of the traditional power grid is changing with the increasing connection of distributed generation (DG) or embedded generation sources (Jenkins, et al. 2010). Grid controls and operations need to be redesigned for DG sources which introduce conditions, such as voltage rise and reverse power flow, for which a grid may not have been designed (Jenkins, et al. 2010). Case studies of DG implementation in Western Australia provide practical experience of the results of connecting DG sources to existing grids (Lewis 2012). This thesis describes a practical method for determining the optimum generation of power in a network and minimising line losses using DG sources while keeping bus voltages within required limits. While power flow simulations can be used to optimise networks, their use can be time consuming, particularly if the system is not simple. This paper uses minimal power flow simulations and provides mathematical models of the network conditions. The method proposed in this paper aims to reduce the time taken to produce a power flow optimisation problem in three steps: 1. performing power flow simulations to develop a linear approximation of a non-linear model of power and voltage relationships with injected power 2. performing further power flow simulations to develop a non-linear model of power and line loss relationships 3. then using both the linear power/ voltage and non-linear power / line loss relationships in a simplex algorithm to determine the optimum solution for generator location and capacity with minimum power loss while maintaining voltage stability Future work suggested includes extending the model to include thermal limits, transformer limits and short circuit ratings.