壓水式反應器之小破口冷卻水流失事故為一壓力邊界形成直徑約3/8-2英吋之事故。若未緩解，將會使爐心裸露進而導致爐心熔毀事故。在本研究中，利用RELAP5-3D模擬在馬鞍山電廠，即美國西屋公司設計之三迴路之2830 MWt壓水式反應器，之運轉員緩解行為。
根據電廠安全度評估報告，小破口冷卻水流失事故的緩解行為有高壓安全注水、降溫降壓、低壓注水系統再循環、更換燃料儲水槽再補水。這些行為將由RELAP5-3D做評估並判定其執行成功與否。熱流以及人為動作之相關參數將由程式模擬。不準度的量化結果將提供人為動作是否達到成功準則之依據以及誤失機率，即運轉員無法在某一時限內成功執行某動作之機率。
根據Wilks’ Formula，本研究選擇取樣了124個事件來獲得代表性的數據。量化不準度所選擇之參數由取樣自現象辨識及排序表以及上述所提之人為動作。結果將分為兩個面向去做分析，分別為：（１）爐心熔毀頻率方面及（２）系統熱流方面。
從爐心熔毀頻率方面分析，只有高壓安全注水之失效機率與安全度評估所得到之值不同。分析結果高壓注水由於運轉員因時間不足而導致執行失敗的機率為8.06×10^(-3)，相對於PSA中之值，增加了約80%，而此動作人為誤失機率，來到了1.74×10^(-1)，即約4.8%的提升。高壓安全注水在PSA中的總失效機率為5.51×10^(-3)，經過人為誤失機率的重新量化，總失誤機率為5.54×10^(-3)，提升了約0.54%。在量化過後，有兩個序列的爐心熔毀頻率發生了變化，分別是序列17與序列20，分別提升了4.65%以及4.63%。
不準度同時也可能導致熱流現象上的差異。在本研究中發現，執行降溫降壓時，一次側的壓力曲線可以從降溫降壓所需時間上大致分為兩類。此現象在統計上顯示出與執行降溫降壓的時間點有相關性。從熱流角度分析，此現象與管路封閉現象(Loop Seal)和一次側水之體積有關。

Small Break Loss of Coolant Accident (SBLOCA) of a Pressurized Water Reactor (PWR) is initiated by a rupture at the primary side pressure boundary with a size 3/8 to 2 in. If left unmitigated, the core may uncover and lead to a core melt accident. In the present study, the operator actions in mitigating a SBLOCA of Maanshan Nuclear Power Station of Taiwan Power Company, which deploys two Westinghouse designed three-loop PWR with rated thermal power of 2,830 MW, are simulated using RELAP5-3D code.
As delineated in the plant-specific Probabilistic Safety Assessment (PSA), the mitigation actions in SBLOCA include high head safety injection, cooling and depressurization of primary systems, low head safety recirculation, and Refueling Water Storage Tank (RWST) replenishment. These mitigation actions are analyzed by thermal hydraulic system analysis code RELAP5-3D to determine the successfulness of the action. The uncertainties of input parameters of the code are included in the assessment and action execution time of these mitigation actions is treated as one of the input uncertainty. The results of the uncertainties quantification give the probability of whether the success criterion of the mitigation action is satisfied. The probability is used to determine the human error probability of the actions, e.g. the action is not carried out within a certain time window.
124 cases were simulated based on Wilks’ formula to obtain representative values. Uncertainties were quantified by sampling from the phenomenon identification and ranking table (PIRT) using the Latin hypercube sampling technique. 18 thermal hydraulic parameters and 4 human action timing were selected for sampling. The results are analyzed from two perspectives: (1) core damage frequency perspective, and (2) system thermal hydraulic perspective.
From a core damage frequency perspective, only the human error probability of high head safety injection differs from that of the PSA. The failure probability due to insufficient execution time is, therefore, 8.06×10^(-3) which is about a 80% increase, and the total human error probability comes to 1.74×10^(-1) which corresponds to about 4.8% increase. The original failure rate of HHSI is 5.51×10^(-3) and after quantification the failure rate is 5.54×10^(-3), the increase in failure rate is about 0.54%. As a result, two specific sequences will be effected, they are sequence 17 and 20. Their respective increase are 4.65% and 4.63%.
The effects of uncertainties may also induce a thermal hydraulic phenomenon. Two types of depressurizing are identified, based on the time required to depressurize primary systems. This is discovered to be related to the execution time of cooldown and depressurization. This may also be approached from a thermal hydraulic perspective. It is discovered that the two types of depressurization are closely related to the loop seal clearance phenomenon and the average water level of the pressurizer during cooldown and depressurization (CND).

[1] “Safety Margin Action Plan – Final Report,” NEA/CSNI/R(2007)9.
[2] Institute of Nuclear Energy Research, Living Probabilistic Risk Assessment Maanshan Nuclear Power Plant, Institute of Nuclear Energy Research, Taiwan, 1995.
[3] “Emergency Operator Procedure 570.00: Reactor Scram or Safety Injection version 13”, Taiwan Power Company, February, 2016.
[4] “Safety Margin Evaluation – SMAP Framework Assessment and Application,” NEA/CSNI/R(2011)3.
[5] About RELAP5-3D(brief) https://relap53d.inl.gov/SitePages/About.aspx
[6] About RELAP5-3D(Detailed) https://relap53d.inl.gov/SitePages/About.aspx
[7] G.L. Mesina, “A History of RELAP Computer Codes,” Nuclear Science and Engineering, Vol 182, pp. V-IX, 2016.
[8] W. Keller, M. Modarres, “A historical overview of probabilistic risk assessment development and its use in the nuclear power industry - a tribute to the late Professor Norman Carl Rasmussen,” Reliability Engineering and System Safety, Vol 89, pp. 271-285, 2005.
[9] B. J. Garrick, “PRA-based risk management: History and perspectives,” American Nuclear Society, 2014.
[10] “60 FR 42622 - Use of Probabilistic Risk Assessment Methods in Nuclear Regulatory Activities; Final Policy Statement”, U.S. NRC, 1995.
[11] N, Aksan, “International Standard Problems and Small Break Loss-of-Coolant Accident (SBLOCA)”, Science and Technology of Nuclear Installations, 10.1155/2008/814572, 2008.
[12] “Thermohydraulics of Emergency Core Cooling in Light Water Reactors,” NEA/CSNI(1989)161
[13] “Small Break LOCA Methodology for US-APWR” , Mitsubishi Heavy Industries, 2014.
[14] Chen et al., “Effect of Power Uprate on Core Damage Frequency in a MBLOCA with Multiple Operator Actions,” Annals of Nuclear Energy, Vol. 109, pp.350-358, 2017.
[15] Bokyung Kim, Hee Cheon No, “Application of Wilks’ formula and concept of State Change Time to integrated Deterministic and Probabilistic Safety Assessment for evaluation of the safety margin of DEC accidents,” Nuclear Engineering and Design, Vol. 352, 2019.
[16] Helton et al., “A comparison of uncertainty and sensitivity analysis results obtained with random and Latin hypercube sampling,” Reliability Engineering & System Safety, Vol. 89, Issue 3, pp.305-330, 2005.
[17] Sukho Lee, Hho-Jung Kim, “Prediction of Loop Seal Formation and Clearing During Small Break Loss of Coolant Accident,” Journal of the Korean Nuclear Society, Vol. 24, Number 3, pp.243-251, 1991.