Some Aspects of Generating Unit Availability Models

Generating unit availability data in North America are compiled by two national organizations. In the United States, utility data are processed and disseminated by the North American Electric Reliability Council (NERC) through the Generating Availability Data System (GADS). In Canada, these data are handled by the Canadian Electrical Association (CEA) in its Equipment Reliability Information System (ERIS). The CEA also processes major transmission equipment outage data as part of the ERIS program. The basic generating unit performance parameters from the GADS and ERIS systems are disseminated by NERC and CEA respectively in the form of comprehensive annual reports [1, 2]. The data bases of both these systems contain a considerable amount of raw data which can be interrogated to produce additional statistics and models for performance evaluation and system reliability assessment. This paper presents a brief comparison of basic data from the GADS and ERIS systems and illustrates some aspects of generating unit availability models using the CEA-ERIS data base. The paper also illustrates some of the additional information which can be produced from the continuous state monitoring procedure used by the CEA. This is illustrated by Weibull analysis of state residence time distributions, consideration of the effect of low operating factors for conventional fossil units and modified models for peaking units.