Abstract The management of reactivity hazards has long been a problem in chemical process and the indexing approach is believed to be an effective solution. However, while most of the reactivity indexes do not consider the runaway scenario and are often quite rough, the assessment criteria used to judge thermal runaway does not take the type of scenario or the severity of the consequences into account. At present, there are no index system that can represent the reactivity hazards in chemical process comprehensively. Therefore, this paper has developed a hazard index for reactivity management (HIRM) in chemical processes. The reactivity hazards could be identified, evaluated and managed through nine steps defined by this method. The HIRM method is able to assess various runaway scenarios and quantify the reactivity hazards based on the consequence and probability. The consequence is rated by the damage radius which represents the damage level caused by fire, explosion, and toxic release due to reaction runaway. The probability indicates the initial event frequency, the initiation time and the controllability of the entire runaway process. The reactivity risks could then be classified by the HIRM index uniformly and comprehensively, with methods to manage the risks proposed. Six sets of equipment in the dicumyl peroxide (DCP) production unit was used as an example to demonstrate the construction of the HIRM index system. The results showed good agreement with the actual risk distribution, meaning that the HIRM method is an effective tool for the management of reactivity hazards in chemical process.

中文翻译：

---

开发化学过程中反应性管理 (HIRM) 的危险指数

摘要 反应性危害的管理长期以来一直是化工过程中的一个问题，索引方法被认为是一种有效的解决方案。然而，虽然大多数反应性指标没有考虑失控情景，而且往往相当粗糙，但用于判断热失控的评估标准并未考虑情景类型或后果的严重程度。目前还没有能够全面反映化工过程中反应性危害的指标体系。因此，本文制定了化学过程中反应性管理（HIRM）的危害指数。可以通过该方法定义的九个步骤来识别、评估和管理反应性危害。HIRM 方法能够评估各种失控场景并根据后果和概率量化反应性危害。后果按破坏半径评定，该半径代表因反应失控引起的火灾、爆炸和有毒物质释放所造成的破坏程度。概率表示初始事件频率、发生时间和整个失控过程的可控性。然后可以通过 HIRM 指数对反应性风险进行统一和全面的分类，并提出管理风险的方法。以过氧化二异丙苯（DCP）生产装置的6套设备为例，对HIRM指标体系的构建进行了论证。结果与实际风险分布吻合良好，