The modular accident analysis program (MAAP) is a fast-running severe accident analysis tool with which the timing of key events and source terms in a severe accident are assessed. The idea of combining MAAP and an optimization algorithm to identify the realistic accident parameters in terms of minimizing the discrepancies between the plant data and the simulation results is straightforward. In 2008 Chien and Wang first compiled the combination of the MAAP4 source codes and a Simplex code as a computer-aided tool for the loss-of-coolant accident (LOCA) of the Kuosheng nuclear power plant (NPP). The break area and break elevation were successfully identified. However, in that approach to putting the idea into practice was that hard data dependence exists between MAAP and the optimization algorithm. Tedious tracing and modification work is required to ensure all plant variables in MAAP source codes with the exception of the adjusted accident parameters are identical at the beginning of every simulation. The plant- and accident-specific development features also easily limit the applications of this idea to the nuclear industry, like being boxed in.

In this study a so-called "out-of-box" approach is proposed that can omit the limits of the idea applications on severe accident management. A parameter identification tool developed in this approach for the same postulated LOCA of the Kuosheng NPP is carried out for verification and validation. It demonstrates the advantages of successful parameter identification, less programming efforts, and no plant- and accident-specific features.