the basic strategy of case-based reasoning (CBR) is retrieving the most relevant cases and adapting them to fit new situation, and solutions would be stored for future reuse. A conceptual framework for a CBR system to scheme black-start plan is presented. The CBR has a hierarchical structure, cases in abstract level present the general characters of black-start, cases in the detail level include the restoration operations of the plants and substation in the subsystem. Conventional CBR system have had limited in revising ability. We developed "path finding" and "generation ability dispatch" algorithms for detail level case revising, and used qualitative process theory (QPT) to revise the abstract cases, the qualitative process model of black-start also can act as an explaining tool. This paper uses the CBR as the main frame, which combines qualitative analysis and quantitative analysis, and the theoretics research and field experience are well organized in this frame.