عدم اطمینان و آنالیز حساسیت LBLOCA در یک نیروگاه هسته ای PWR: نمایش نتایج از فاز V از برنامه BEMUSE

This paper presents the results and the main lessons learnt from Phase V of BEMUSE, an international programme promoted by the Working Group on Accident Management and Analysis (GAMA) of OECD to address the issue of the capabilities of best-estimate computational tools and uncertainty analysis. The scope of Phase V is the uncertainty analysis of a Large Break Loss-Of-Coolant-Accident (LBLOCA) in a Pressurized Water Reactor. Fourteen participants from twelve organizations and ten countries participated in the Phase V of BEMUSE. The paper starts with a general description of the BEMUSE programme including the objectives, structure, and the outline of the Phase V specification. Then it summarizes some general aspects on the uncertain model parameters and the results for the uncertainty analysis and for the sensitivity evaluation. To end with, general recommendations and conclusions are presented as practical guidance for uncertainty analysis performance.

The use of best estimate codes in combination with evaluation of uncertainties, the so-called BEPU methodologies, is an accepted procedure by the regulatory authorities to carry out a deterministic safety analysis. Since the promulgation of the USNRC ECCS regulatory requirements in 1988, several BEPU approaches have been developed in this scope. Examples of activities promoted by the OECD/CSNI to assess such methodologies are the “Uncertainty Analysis Methods” workshop held in London (UK) 1994 (Wickett and Yadigaroglu, 1994), the “Best Estimate Methods in Thermal Hydraulic Safety Analysis” seminar held in Ankara (Turkey) 1998 (Aksan et al., 1998), and the “Uncertainty Methods Study” (UMS) study (Wickett et al., 1998). The UMS was the first international study on the uncertainty methodologies, a step by step comparison of the BEPU methods that was organized in response to the recommendations concluded in the UAM 1994 workshop. Furthermore, the IAEA Safety Report series No.23 “Accident analysis for Nuclear Power Plants” (Allison, 2002), issued in 2002, recommends sensitivity and uncertainty analysis if best estimate codes are used in licensing analysis. A comprehensive overview about uncertainty methods can be found in the IAEA Safety Report Series No.52 “Best Estimate Safety Analysis for Nuclear Power Plants: Uncertainty Evaluation”, issued in 2008 (IAEA, 2008). The BEMUSE Programme has been promoted by the Working Group on Accident Management and Analysis (GAMA) of OECD/NEA and endorsed by the Committee on the Safety of Nuclear Installations (CSNI). BEMUSE (Best Estimate Methods – Uncertainty and Sensitivity Evaluation) main objectives are: • To evaluate the practicability, quality and reliability of Best Estimate methods including uncertainty evaluations in applications relevant to nuclear reactor safety. • To develop a common understanding in this domain. • To promote/facilitate the use of these methods by the regulatory bodies and the industry. The application of these methods to a Large-Break Loss of Coolant Accident (LBLOCA) constitutes the main activity of the programme, organized into two main steps (six phases): - Step 1: Best-estimate and uncertainty and sensitivity (BE + U + S) evaluations of the LOFT L2-5 test. It includes Phase I the description of the methods (Micaelli et al., 2005), Phase II the BE calculation of the test (Petruzzi et al., 2006) and Phase III the uncertainties and sensitivities analysis (de Crécy et al., 2007). - Step 2: BE + U + S evaluations of a PWR Nuclear Power Plant. The aim of Phase V (Reventos, 2009) is the uncertainty analysis of a LB-LOCA based on a reference calculation performed in Phase IV (Reventos et al., 2008). For this particular and for the rest of the article it is understood that the reference case is a calculation using nominal best-estimate input values and default values for the computer code options and input data for models. Finally Phase VI will summarize conclusions and recommendations of the whole exercise. Background from previous phases has been used to produce final uncertainty results in Phase V. The objectives of the activity are: • To obtain uncertainty bands for the maximum cladding temperature (time trend), upper plenum pressure (time trend), maximum peak cladding temperature (scalar), 1st peak cladding temperature (scalar), 2nd peak cladding temperature (scalar), time of accumulator injection (scalar), time of complete core quenching (scalar). • When using a probabilistic approach methodology: to evaluate the influence of the selected parameters on maximum cladding temperature (time trend) and upper plenum pressure (time trend). • To compare procedures with experience gained in previous Phase III. Fourteen groups from twelve organizations and ten countries have participated in BEMUSE Phase V (see Table 1). All participants have experience in the involved fields. Notable works of the organizations devoted either to comparative exercises, uncertainty evaluations or model verification are Glaeser et al. (1994), D’Auria (1998), Guba et al. (2003), de Crécy et al. (2008), Reventos et al. (2008), Song et al. (2010), Bukin et al. (2009) and Perez et al. (2010). Table 1. List of participants in BEMUSE Phase V. Numb. Organisation Country Name E-mail Code 1 AEKI Hungary A. Guba guba@aeki.kfki.hu ATHLET 2.0A I. Tóth tothi@susserv.kfki.hu 2 CEA France T. Mieusset thomas.mieusset@cea.fr CATHARE2 P. Bazin pascal.bazin@cea.fr V2.5_1 (r5_567) A. de Crécy agnes.decrecy@cea.fr 3 EDO Russia S. Borisov borosov_sl@grpress.poldolsk.ru TECH-M-97 4 GRS Germany T. Skorek Tomasz.Skorek@grs.de ATHLET 2.1B H. Glaeser Horst.Glaeser@grs.de 5 IRSN France J. Joucla jerome.joucla@irsn.fr CATHARE2 P. Probst pierre.probst@irsn.fr V2.5_1 mod6.1 6 JNES Japan A. Ui ui-atsushi@jnes.go.jp TRACE ver4.05 7 KAERI South Korea B.D. Chung bdchung@kaeri.re.kr MARS 3.1 8 KINS South Korea D.Y. Oh k392ody@kins.re.kr RELAP5/mod3.3 9 NRI-1 Czech Republic R. Pernica per@ujv.cz RELAP5/mod3.3 M. Kyncl milos.kyncl@ujv.cz 10 NRI-2 Czech Republic Jiri Macek mac@ujv.cz ATHLET 2.1 A 11 PSI Switzerland A. Manera annalisa.manera@psi.ch TRACE5rc3 J. Freixa jordi@freixa.net 12 UNIPI-1 Italy A. Petruzzi a.petruzzi@ing.unipi.it RELAP5/mod3.2 F. d’Auria d5808@ing.unipi.it 13 UNIPI-2 Italy A. Del Nevo a.delnevo@ing.unipi.it CATHARE2 F. d’Auria d5808@ing.unipi.it V2.5_1 mod6.1 14 UPC Spain M. Perez marina.perez@upc.edu RELAP5/mod3.3 F. Reventos francesc.reventos@upc.edu L. Batet lluis.batet@upc.edu Table options BEMUSE Phases IV and V deal with a generic PWR plant. Zion 1 was selected as a reference for practical reasons but there was no detailed documentation available concerning the state of the plant as initial and boundary conditions, fuel properties, etc. To limit the spread of the results due to the scarcity of data, during Phase IV it was agreed to provide common PWR-generic information about geometry and modelling. In a similar way for Phase V, some information about uncertainty was supplied to the participants following a probabilistic approach (see Sections 2 and 2.2). Zion Unit 1 was a four loop Westinghouse design PWR, located in Illinois, USA. The plant rated thermal power was 3250 MW. It ceased operation in 1998. It is now permanently shut down. The steady-state conditions are listed in Table 2, extracted from BEMUSE phase IV specifications (Reventos et al., 2008). Table 2. Steady-state main parameters. Parameter Steady-state value Power MW 3250.0 Pressure in cold leg MPa 15.8 Pressure in hot leg MPa 15.5 Pressurizer level m 8.8 Core outlet temperature K 603.0 Primary coolant flow kg/s 17357.0 Secondary pressure MPa 6.7 SG downcomer level m 12.2 Feed water flow per loop kg/s 439.0 Accumulator pressure MPa 4.14 Accumulator gas volume m3 15.1 Accumulator liquid volume m3 23.8 RCP's velocity rad/s 120.06 Table options The scenario simulated is a cold leg LBLOCA without actuation of the high pressure injection system (HPIS). Extended information on the plant and scenario can be found in the Phase IV report (Reventos et al., 2008) as well as in Perez et al. (2010). The conditions imposed for the scenario can be summarized as: • Low pressure injection system (LPIS): 1.42 MPa pressure set point. Driven by a flow-pressure table. • Accumulators’ injection: 4.14 MPa pressure set point. • Containment pressure imposed as a function of time after the break. • Reactor coolant pumps velocity imposed as a function of time after the break. • Power after scram imposed by means of a reactor power multiplier as a function of time after the break. Uncertainty and sensitivity analysis had to be performed for seven output parameters – two time trends and five single-valued parameters, described in Table 3. Table 3. Output parameters for uncertainty and sensitivity analysis. Type Definition Criterion Time trend Maximum cladding temperature: MCT Maximum cladding temperature at each time step without location dependency (neither axial nor radial) Pressure in the upper plenum: Pup No criterion Scalar quantities 1st PCT (blowdown phase)a MCT and t < tinj 2nd PCT (∼reflood phase) MCT and t > tinj Time of accumulator injection: tinj Time of beginning of injection Time of complete core quenching: tque Tclad ≤ Tsat + 30 K Maximum peak cladding temperature: MPCT Maximum temperature value reached on the cladding surface during all transient, independently of its locations (radial or axial) Tclad: cladding temperature, Tsat: saturation temperature. a It was agreed that the definition of the blowdown phase follows the one given in LBLOCA PIRT (Shaw et al., 1988) as in phase III. Table options Comparing to previous Phase III, a Maximum Peak Cladding Temperature (MPCT) scalar quantity has been added. The reason of including it is because it is the main parameter which is compared with its design safety limit in LOCA licensing analyses. The form of the uncertainty of the output parameters are a lower and upper bound giving an estimation of the 5% and 95% quantiles, respectively, with a confidence level of 95% for both quantiles. Following previous Phase III procedure, the exercise also included sensitivity analysis. Except for the two groups of UNIPI with its CIAU method all the participants used a fully probabilistic approach based on the non-parametric statistics originally incorporated in the GRS methodology for LOCA analysis. A brief description of the different methods used by participants can be found in another paper (de Crécy et al., 2008).