This thesis addresses the characterisation of the impact that changes in the design of a single generator can have on fault behaviour throughout a complex interconnected power system. The potential impact of generator design changes on network fault behaviour were considered, under both balanced three-phase and single line-to-ground fault conditions, using a new analytical procedure, described as the "breakpoint separation approach". This allowed the identification of the range of generator designs that could be used to improve network-wide fault behaviour. The general understanding obtained was applied to an evaluation of the feasibility of enhancing system fault performance through careful selection of the design of PowerformerTM, a directly connected high voltage generator, used to augment or replace existing generation capacity.
Given the complexity of modem power systems, it is inevitable that failures and faults will occur. Synchronous generators are the predominant sources of fault current within a power system. Their design and placement will have significant influence on the fault behaviour of the entire network. A clear understanding of the relationship between generator design and fault behaviour is essential for satisfactory system operation in the presence of novel generator configurations, such as PowerformerTM.
PowerformerTM is an innovative synchronous generator capable of producing electricity at transmission voltage levels. It is connected directly to the high voltage network, without need for a step-up transformer. The impact of replacing a conventional generator with this radically different generator configuration was addressed as a specific case of the over-riding relationship between generator designs and system fault behaviour developed in this thesis.
Initial work focussed on the selection of a generator model capable of representing wide-ranging design modifications. This model was used to illustrate the influence that the design of a single generator can have on the configuration of the entire network. Subsequently, an analytical representation of the relationship between generator design changes and network fault behaviour was developed.
This analytical representation, described as the "breakpoint separation approach" allowed unified treatment of the impact of generator design changes on a wide range of different fault parameters. The "breakpoint separation approach" was used to identify the system imposed limits on the ability of generator design selection to control fault behaviour. These limits control the maximum possible change in each fault parameter, the network locations where fault behaviour was most sensitive to generator design, as well as the comparative range of sensitive machine parameters. The effectiveness of the new technique was confirmed by comparison with the simulated fault behaviour of a small test system.
The analytical technique was used, along with traditional fault analysis methods, to characterise the potential impact of replacing a conventional generator with PowerformerTM on the fault behaviour of a realistic complex interconnected power system. The potential impact was contrasted with the behaviour observed when the conventional generator was replaced with a realistically designed PowerformerTM. In both cases, the manner of fault behaviour change was consistent with that predicted using the "breakpoint separation approach". In addition, it was shown that changes to the design of a single generator could have a pronounced effect on fault conditions hundreds of kilometres from the generator terminals. In contrast, the impact of realistic generator design changes was comparatively limited, and confined to the vicinity of the generator terminals, especially under balanced fault condition.
The significant disparity between the potential change in network fault behaviour, and the observed changes produced by more realistic generator modifications were found to result from the differences between the set of physically realisable generator parameters and the range of sensitive generator designs. The difference was less pronounced under single line-to-ground fault conditions, due to the greater degree of flexibility in the selection of generator grounding connections.
Overall, the results suggest that physically realisable changes to a single generator design have only limited ability to produce widespread modification to the fault behaviour of a complex interconnected power system. Conversely, this also implies that even pronounced generator design changes, such as the replacement of a conventional generator with the directly connected PowerformerTM, may be accommodated with only limited disruption in network fault behaviour. Finally, the presented study has demonstrated the effectiveness of new analytical approach to characterising the relationship between generator design and fault behaviour. In some respects, this development is of equal significance to the numerical results obtained.