Abstract: To solve the key defect that Safety Integrity Level（SIL） verification of safety instrumented system and mechanism Failure Modes, Effects and Diagnostic Analysis（FMEDA） both lack in product（such as valves, transmitters and logic controllers and so on） failure rates obtained from special factory environment and interlock loops, a new method based on subjective estimation method and Pearson chi-square method is proposed to analyze failure data. The method is applied to estimate the life distribution and actual failure rate of pneumatic diaphragm control valve used in the interlock loop upon high exit temperature of acrylonitrile reactor. SIL verification of the loop is carried out by mechanism FMEDA and Fault Tree Analysis and further suggestions are presented. The result shows that SIL verification with data provided by the method is more accurate than others with foreign databases.

Key words: safety instrumented system, failure modes, effects and diagnostic analysis, safety integrity level verification, subjective estimation, Pearson chi-square

摘要： 针对安全仪表系统安全完整性等级验证（SIL）和机械装置的失效模式、影响及诊断分析（FMEDA）缺乏特定工厂环境、联锁回路中的产品（如阀门、变送器、逻辑控制器等）失效数据这一关键不足，提出了一种基于主观估计法和皮尔逊卡方检验法的失效数据分析方法，并以丙烯腈反应器出口温度高联锁回路中的气动薄膜调节阀为例，应用该方法估计产品寿命分布和实际失效率；应用机械FMEDA法和故障树（FTA）法对回路进行SIL验证，给出改进措施。结果表明，应用该方法获取的产品失效数据对相应联锁回路进行SIL验证比应用国外通用数据库更准确。

关键词: 安全仪表系统, 失效模式, 影响及诊断分析, 安全完整性等级验证, 主观估计, 皮尔逊卡方检验