過去的核能電廠安全對於分析超過設計基準事件，一般使用PSA方法計算事故序列發生的CDF(core damage frequency)，然而其以大量假設條件及成功準則的評估並未充分考量參數的不確定性。RISMC(Risk-Inform Safety Margin Characterization)方法在分析事故序列時，加入其隨機不確定性的特質，將產生別於傳統方法的結果。在同時考慮事故嚴重性與發生頻率下，本文以台電核三廠廠內事件量化分析報告中事故序列CDF大於1.0E-07/y前十名中的第三名(前一二名分別是喪失4.16 kV緊要匯流排A串肇始事件與蒸汽產生器管束破裂) ─ 小破口爐心冷卻水流失(SBLOCA)序列S(2)S20為研究案例，驗證不同方法論的差異。在此事故序列中，儘管於台電核三廠人為可靠度分析報告中已充分估計人因條件失效概率，但在考量參數的隨機不確定性後，實際允許的救援時間將受到極大的影響，進而可能使失效概率有所變化，而非報告中假設固定的時間。本文希望能透過熱工水流程式RELAP5-3D來模擬納入電廠條件與人因參數不確定性的SBLOCA事故序列，並使用統計方法擬合密度機率函數以評估事故發生下運轉員進行「爐心冷卻再回復」的成功概率，進而與過去的CDF數據比較。

For analysis of Beyond Design Basis Event, PSA methods would be initiated to determine event sequences’ frequency of core damage. However, its estimation utilizing large numbers of conditional assumptions and criteria don’t have sufficient consideration of parameters uncertainty. RISMC method which addresses aleatory uncertainty would generate different result compared with traditional way. In light of both of sequences’ damage and frequency, this research uses Small Break Loss of Coolant Accident event sequence S(2)S20, 3rd of top ten internal initiating event sequences with highest core damage frequency over 1.0E-7/year ranked in Probabilistic Risk Assessment Maanshan Nuclear Power Station Final Report, as example (which 1st and 2nd are Loss of Safety 4.16kV Bus A-PB-S02 and Break of Steam Generator Tube respectively) to evaluate degree of difference between two methods. Although it already has a comprehensive estimation of Human Reliability Conditional Exceedance Probability in Maanshan Nuclear Power Plant Human Reliability Report, the actual operation time would be extremely affected if taking aleatory uncertainty as consideration. This research is expected to use RELAP5-3D to simulate SBLOCA event sequence with both plant operation factor and human factor parameters uncertainty. Then utilize statistical methods to approach probability distribution function, calculating operation success rate in step of Core Cooling Recovery. The result would be compared with past data in safety report.

[1]台灣電力公司.「核三廠人為可靠度分析報告」, May , 2013
[2]台灣電力公司.「台電MSPRASA分析報告201307」, July , 2013A. D. Swain, H. E. Guttmann , “ NURGE/CR-1278 Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications Final Report ” , August , 1983
[3]A. D. Swain, H. E. Guttmann , “ NURGE/CR-1278 Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications Final Report ” , August , 1983
[4]Atomic Energy Council , Executive Yuan , “ Probabilistic Risk Assessment Maanshan Nuclear Power Station Unit 1 Final Report ” , October , 1987
[5]Idaho National Laboratory ,“ RELAP5-3D Code Manual Volume II Appendices ”, June , 2014
[6]Slifker J.F., and Shapiro S.S. “ The Johnson system: selection and parameter estimation ” ,1980
[7]Milton , J. Susan. 吳榮彬譯. 「工程統計學原則與應用」, 2003
[8]K. Durga Rao, H.S. Kusheaha, A.K. Verma, A. Srividya , “ Quantification of epistemic and aleatory uncertainties in level-1 probabilistic safety assessment studies ” , July , 2007
[9]鄭玉濤、梁國興、楊惠君、蔡斐然，「風險告知安全裕度特性分析方法研究與應用- LBLOCA特定事故序列分析條件失效概率計算」