A statistics based approach to enhancing safety and reliability of the batch-reactor charging operation

In general there are two critical decisions that must be made in the charging sequence of batch reactors, i.e. target setting and alarm generation. In this paper, a number of statistics-based strategies are proposed to perform these tasks. Specifically, both off-line and on-line target setting procedures are presented and either one of them can be adopted to increase the profit margin of any given batch process without sacrificing reliability. A synthesis method for building optimal alarm logic is also described in detail. The monitoring systems constructed according to this suggested approach are effective in reducing the probability of undetected faulty batches. These techniques have been tested with extensive simulation studies. The results show that the proposed strategies are suitable for application if high-value-added products are manufactured in the plant, which in fact is a prevailing situation of batch processes.