Development of a methodology to reduce the order of a detailed lumped parameter transformer model
The transformer designer employs detailed electrical models to develop a reliable and cost effective transformer insulation structure. The power engineer must model not only the transformer, but the system, therefore the power engineer requires a smaller model of the transformer that accurately represents its behavior in order to investigate the effects of power system transients. Reduced models are generally obtained either from detailed design models or from measurements on fully constructed transformers. Reduced models constructed from design data generally act as low pass filters and are severely limited in accuracy at high frequencies. The latter technique has the major disadvantage that the model cannot be constructed until after the transformer has been built. Presently, both methods are subject to considerable error. The primary objective of this thesis is to develop a technique that reduces the order of a lumped parameter transformer model used in insulation design and that provides a reduced model of any specified size for transient studies of systems which contain the transformer. The reduction technique developed can be applied to any lumped parameter network which uses electric parameter analogs (i.e., FEM networks). The method of this thesis uses Kron`s reduction approach in time domain to obtain a reduced model. This reduced model is compatible with industry methods for transient studies (EMTP) and retains the accuracy and stability of the detailed model. Additionally, this reduced model can be used to predict the interaction between the transformer and the power system, via EMTP, giving a valuable tool to both power and design engineers. Application of the technique for the detailed model of a 765/345/34.5 kV, YYD, core form, 500 MVA single phase autotransformer is verified by frequency and time domain tests for the linear model. The nonlinear transformer model reduction technique is outlined and a proof of concept is provided by two examples.
- Research Organization:
- Rensselaer Polytechnic Inst., Troy, NY (United States)
- OSTI ID:
- 39365
- Resource Relation:
- Other Information: TH: Thesis (Ph.D.); PBD: 1993
- Country of Publication:
- United States
- Language:
- English
Similar Records
Linear, lumped parameter transformer model reduction technique
A nonlinear lumped parameter model for transient studies of single-phase core form transformers