Design of MIDA, a Web-Based Diagnostic Application for Hydroelectric Generators

Up to 95% of Hydro-Quebec´s electrical power is produced by hydroelectric generators. The remainder comes from conventional thermal and nuclear generators and wind turbines. Implementing a cost-effective general maintenance program for generators is of utmost importance for Hydro-Quebec. This paper presents MIDA, a Web-based application for diagnosing hydroelectric generator. It allows Hydro-Quebec maintenance personnel to better establish maintenance priorities based on analysis and trending of data coming from several diagnostic instruments following an integrated generator diagnostic methodology (MIDA). Generator maintenance is so crucial that almost from the start of the project, Hydro-Quebec managers needed the initially-incomplete diagnostic information it produced. This paper presents the evolutionary prototyping methodology we used for the development of the MIDA application, which is very well suited to this type of research and development project. It also presents the programming language we used: MDI (for Dynamic Information Modeling), an object-oriented knowledge-based prototyping tool.