Semiautomated Model Validation of Power Plant Equipment Using Online Measurements

Model validation of power plant equipment is expected to be mandated imminently by the North American Electric Reliability Corporation. These standards will result in a great need for performing model validation for a large fleet of existing and future power plants. The historic approach to model validation of synchronous generators has typically been a battery of offline tests. In addition, the process of validation depends on a trial and error series of simulations, comparing the results of such simulations with measured test data, to gradually tend toward a validated model and its parameters. Such a method presents significant challenges for application on a large fleet of units. This paper documents a practical approach to model validation. The method is based on collecting online data from the generating unit, typically near or at base-load. Data are collected either from digital-fault recorders capturing system disturbances, or from digital excitation system for an imposed small voltage-reference step. The data are then used in a software tool to perform model validation using an automated optimization algorithm-the entire process may be said to be semiautomated. The benefits are explained in the paper through actual examples of the application.