發掘及降低核電廠隱藏性風險之探討

Abstract

有別於一般工業的工作安全，核電廠更需注意到核能安全。因為若核電廠發生核災事故則在生命、財產與環境的巨大損失是無法估計的。因此確保核電廠的核能安全是普為世人所重視的一件大事。要落實確保核電廠的核能安全有很多種方法，辨識出存在核電廠各種設備中的風險是其中之一。本文研究的目的是提供一種協助辨識營運中核電廠內隱藏性風險的方法，再針對發掘出來的隱藏性風險提供一些減少這些風險的參考對治方向。文中的研究假設是依據理論與文獻將安全績效與影響安全因素連結而推導出來。並採現場實際時間序列資料檢視檢定結果與理論的期望是否一致。如果檢定的結果是資料不支持研究假設則就可發掘出樣本核能發電廠中存在隱藏性風險與因素。吾人擷取樣本核電廠自營運開始後28年342台泵浦故障歷史資料，並採用潛藏成長模式(latent growth modeling, LGM)為分析工具。研究結果發現全廠泵浦故障平均成長趨勢呈開口向下二次式拋物線型與四個隱藏性風險。依此發現建議該核電廠除了要有充分的訓驗與適當的安全管理機制外，另建議加強團隊資源管理(crew resource management)減少隱藏性風險與提升該廠的核能安全。最後推薦本研究所採用的發掘營運中核電廠隱藏性風險的方法可使用到樣本核廠其他種設備與其他各核電廠，俾協助各廠辦識與解決各廠的隱藏性風險。

In this study, an approach is proposed to identify hidden contributors to risk in the operation of a nuclear power plant. Hypotheses linking safety performance to its influential factors are developed based on theory and literature. The failure to support the relational hypotheses with longitudinal safety performance data from a plant signifies hidden contributors to risk in the operation of the plant, which can be specifically identified by looking into the empirical results inconsistent with theoretical expectations. To demonstrate the use of the approach, we develop five hypotheses and analyze the longitudinal failure data for 28 years collected from 342 pumps in a large nuclear power plant by using latent growth modeling (LGM). The average trajectory of pump failures for the entire plant is inverted-U-shaped. Four hidden risk contributors in the operation of the sample plant have been identified. We suggest that, in addition to the adequacy of training and safety management systems, crew resource management be used to enhance nuclear safety for the sample plant. We recommend that the proposed approach be used for any other nuclear power plant to help identify hidden risk contributors and in other similar situations.