ML23089A229

Pre-Application Ipwr PIRT Report for NuScale-like Design

ML23089A229
Person / Time
Issue date:	06/13/2011
From:	Peter Yarsky NRC/RES/DSA
To:
References
ML110130538
Download: ML23089A229 (42)

Text

NuScale proprietary information has been redacted from this document

Pre-Application Phenomena Identification and Ranking Table (PIRT) Report for NuS cale-like Integral Pressurized Water Reactors (iPWRs)

(Summary Report for Task 3 of iPWR User Need Response [ML110130538])

Manuscript completed: June 13, 2011

Division of Systems Analysis Office of Nuclear Regulatory Research

1 Pre-Application PIRT Report for NuScale-like Integral Pressurized Water Reactors (iPWRs)

Thermal-Hydraulic PIRT Panel Members:

D. Diamond (BNL)

S. Bajorek (NRC) - Panel Chair P. Griffith (MIT)

W. Krotiuk (NRC)

P. Lien (NRC)

B. Parks (NRC)

J. Staudenmeier (NRC)

Neutronics PIRT Panel Members:

S. Bajorek (NRC)

D. Diamond (BNL) - Panel Chair S. Frankl (NRC)

P. Griffith (MIT)

N. Hudson (NRC)

P. Yarsky (NRC)

Manuscript Preparations:

K. Tene (NRC)

AJ Nosek (NRC)

Project Manager: JR Skarda (NRC)

2 Executive Summa ry

This report de scribes t wo pre-application Plant P henomena Identificat ion and Rank ing Tables (PIRTs) developed for a NuScale-like desig n. The pur pose o f the PI RTs is to ident ify phenomena and pr ocess t hat will need to be cons idered when assessing safety r elated ev ents for I ntegral Pressurized Water Reac tors ( iPWRs). Two events were considered for t he desi gn; a Thermal-Hydraulic Driven Event and a Neutr onic Driven Event. Because the PI RTs hav e been developed early in the pre-application phase, neit her f inal desi gn data nor license application submittal data are av ailable. Conseq uently t he PIRTs hav e been based on best available information t hat has been obta ined from po tential applicants an d liberal use engineering/expert judgment that ha s been applied by t he PIRT panel members. A revision based on final des ign information is vital to t he de velopment and review of thermal-hydraulic and neutronic codes for l icensing decisions.

The accident scenario co nsidered for the t hermal-hydraulic PIRT is t he inadv ertent openin g o f a reactor recirculation v alve. An inadvertent control r od ban k w ithdrawal from ho t zero power (HZP) w as the scena rio consider ed for t he neutr onics PIRT. Both PIRTs identify phenom ena o f high im portance, those that must be modeled accurately or bounded in an appr opriate manner in a safety analy sis to ad equately assur e margin to regulatory lim its.

For t he thermal-hydraulic PIRT, the figures-of-merit ( FOMs) w ere t he Cr itical Heat Flux ( CHF),

mixture level in the vessel, and containment v essel pressure. The panel in general, did not feel that limits on t he depa rture from nuclea te boiling ratio (DNBR) w ould be challenged except possibly in the early par t of t he transient i f the co re flow w ere t o s tagnate while at or near full power oper ation. The m ixture level was selected as a FOM because the safety s ystems are designed t o prevent co re uncov ery and heatup. In the NuScale design, a conventional containment is r eplaced by a relatively sm all high-pressure c ontainment vessel. During blowdown and early in the transient, high p ressure w ithin the containment vessel is a s afety concern and ther efore included as a FO M.

Of par ticular int erest are those phenomena t hat are highly r anked in importance, bu t are poorly understood as indicat ed by a low k nowledge level. For t he NuScale-like d esign, the major concerns appear to be with t he Cont ainment Heat Removal System (CHRS), for which there were several phenomena the panel identified as having hig h im portance and less than satisfactory knowledge level. Phenomena in this cat egory include condensation heat transfer (both w ith and without no ncondensables), and mixing in t he containment as it a ffects t he liq uid pool and the gas space. There w ere additional concerns with choked flow f rom t he v essel through t he r eactor v ent valves and t he reactor r ecirculation valves. These are m ainly w ith t he prediction with flow t o the break and chok ed flow from the primary.

The neutr onic PIRT considered a scenario whereby a control ban k w as inadvertently continuously wit hdrawn from hot zero pow er cond itions. This would be expect ed to result in a power incr ease that would be offset initially by Do ppler feedback and t hen by r eactor tr ip.

Multiple FOMs were used to cov er as many phen omena as possible. The FO Ms w ere chosen by t aking into account what w ould be the acc eptance criteria for t his ev ent. Hence, the FO Ms 3

were: DNBR, reactor v essel pressure, f uel centerline temperature, and pellet -clad interaction (PCI) / pellet-clad mechanical interaction ( PCMI)

The panel listed 35 phen omena as hav ing s pecial significance in modeling this event. Mos t of these relat e to eit her t he neutronic initial conditions or neutr on kinetics. The results show t hat 22 of the phenom ena a re of high ( H) im portance (having si gnificant or do minant in fluence on one or m ore FOMs), 8 ar e of medium (M) im portance, and 5 are of low ( L) importanc e (having only a small influence on t he FO Ms). There are no phenomena t hat are of hig h im portance for which the knowledge level is low ( an H, L r anking). The nex t level of int erest is for pheno mena that have either a high importanc e with a medium knowledge level (H,M) or a medium importance with low k nowledge level (M,L). They ar e listed below w ith t heir r anking (importance, knowledge level).

• Shutdown bank speed (H,M)

• Assembly Interaction (H,M)

• Axial/radial reflector representation (H,M)

• Pellet burnup distribution (H,M)

• Core pin-by-pin burnup distribution (H,M)

• Gap conductance (H,M)

• Fuel conductivity and density (H,M)

• CHF correlations (H,M)

• Detector response (M,L)

The deficiencies in knowledge level for these phenomena are summarized as follows:

• Shutdown bank speed, which partially defines the termination of the scenario, is determined, in addition to Technical Specifications, by bowing of guide tubes, and this bowing is difficult to predict.

• Assembly interaction is always an important consideration in modeling and becomes more uncertain when, as in the NuScale design, the core is small.

• Axial and radial reflector representation is more important in the NuScale design than in large PW Rs, and the neutronics methods that are suitable for the large PW Rs are not expected to be sufficient for the iPW Rs.

• Pellet burnup distribution is important in determining the neutronic properties of the fuel and, because of self-shielding effects which are difficult to model, has a significant uncertainty.

• Core pin-by-pin burnup distribution is a factor in determining the thermo-mechanical properties of the fuel. Although power reconstruction methods exist, modeling is still uncertain.

• Gap conductance and fuel conductivity and density determine heat transfer to the coolant, but these properties may be uncertain due to uncertainties in the burnup of a given fuel rod and because of uncertainties in those properties as a function of burnup.

CHF correlations are directly related to one of the FOMs, but their use for short fuel assemblies is uncertain without measurements to confirm correlation applicability to these geometries. Detector response to changes in core power and delays between 4

signal receipt and the movement o f control banks is uncertain because of a lack o f design information rather t han bec ause the pheno menon is di fficult to m odel.

5 Table of Contents

1 INTRODUCTION................................................................................................................. 8 2 PLANT DESCRIPTION........................................................................................................ 8 2.1 Plant Overview............................................................................................................. 8 2.2 Reactor Core.............................................................................................................. 13 2.3 Pressurizer................................................................................................................. 1 3 2.4 Helical Coil Steam Generator..................................................................................... 13 2.5 High-Pressure Containment....................................................................................... 14 2.6 Decay Heat Removal System.................................................................................... 15 2.7 Containment Heat Removal System........................................................................... 16 2.8 Emergency Core Cooling System.............................................................................. 17 3 PIRT METHODOLOGY..................................................................................................... 17 3.1 General Approach...................................................................................................... 17 3.2 Application to Preliminary Designs............................................................................. 18 3.3 Ranking Scales.......................................................................................................... 18 3.3.1 Phenomena Importance Ranking........................................................................ 18 3.3.2 Knowledge Level Ranking................................................................................... 19 4 THERMAL-HYDRAULIC PIRT........................................................................................... 19 4.1 PIRT Panel Members................................................................................................. 19 4.2 Scenario..................................................................................................................... 20 4.3 Figures of Merit.......................................................................................................... 21 4.4 Systems, Subsystems and Components.................................................................... 22 4.5 Phenomena Ranking and Discussion......................................................................... 23 4.5.1 Core System Phenomena................................................................................... 23 4.5.2 Primary System Phenomena............................................................................... 24 4.5.3 Secondary System Phenomena.......................................................................... 27 4.5.4 CHRS System Phenomena................................................................................. 27 4.6 Phenomena / Knowledge Level Issues....................................................................... 32 4.6.1 Blowdown............................................................................................................ 32 4.6.2 RVV Depressurization......................................................................................... 32 4.6.3 Long-Term Cooling.............................................................................................. 32 5 NEUTRONICS PIRT.......................................................................................................... 33 5.1 PIRT Panel Members................................................................................................. 33

6 5.2 Scenario..................................................................................................................... 33 5.3 Figures of Merit.......................................................................................................... 34 5.4 Results and Discussion.............................................................................................. 34 5.4.1 Phenomena......................................................................................................... 34 5.4.2 Ranking............................................................................................................... 37 6.0

SUMMARY

AND CONCLUSIONS..................................................................................... 40 6.1 Thermal-hydraulic PIRT................................................................................................ 4 0 6.2 Neutronic PIRT.............................................................................................................. 40

7.0 REFERENCES

.............................................................................................................. 42

7 1 INTRODUCTION

The U. S. Nuclear Re gulatory Comm ission (NRC) has received notification of int ent to submit design certific ation applications for int egral pressurized water r eactors ( iPWRs) in 2012. The NRC has been requested to en gage in pre-application review act ivities for iPWR desi gns. The Office of New React ors ( NRO) is pr oceeding w ith ear ly ef forts to ident ify and examine key technical and policy issues important to licensing of s mall-o r medium-sized reactors desi gns.

The Of fice o f Nuclear Re gulatory Res earch ( RES) is supporting NR O w ith thes e early e fforts.

RES assistance focuse s on the ident ification o f unique and im portant phenomena, development of phenomena identification and rank ing tables (PIRTs), and assessment o f code capabilities that can provide insights t o the aforementioned issues and provide better preparat ion for review of design certific ation applications. Result s o f this t ask will be used as t he principal basis for a gap analy sis and code applicability assessment of NRC thermal-hydraulic and neutronic computational tools.

In Apr il 2011, PIRT Pane ls were conv ened to dev elop four iP WR PI RTs, two for a NuScale-like design and t wo for an mPower-like design. Two events or transients w ere considered for each design, a thermal-hydraulic driven event and a neutronic driven event. For t he NuScale-like design, the thermal-hydraulic and the neutr onics PIRT meetings w ere hel d on April 11, 2011 and April 13, 2011 res pectively. During these m eetings PI RTs w ere dev eloped. However, these activities have been performed early in t he pre-application phase w hen neither final design data nor license application s ubmittal data are available. Cons equently t he PIRTs have been based on best available information ob tained from pot ential applicants and t hat use of engineering/expert judgment has been applied by t he PIRT panel members. The applicabilit y of the PIRTs to subsequent iPWR applicat ion reviews is dependent on t he de gree of similarity o f the preapplication design in formation t o t hat of the final design informat ion submitt ed for a license application. This r eport describes t he t wo PIRTS developed for a NuScale-like design.

The PI RT panels w ere a ssisted by Dr. Raymond Skarda as Pr oject Manager for t he ex ercise and by M s. Kimberly T ene in preparation of the final report.

An overview and general descr iption of the NuSc ale-like plant design considered for this PIRT is provided in Chapter 2. Chapter 3 descr ibes t he methodology used for dev elopment o f the PIRTs. Chapter 4 des cribes the PI RT for t he thermal hydraulic driven event, and the neutronic driven event is discussed in Chapter 5. Conclusions are pr ovided in Chapter 6.

2 PLANT DESCRIPTION

2.1 Plant Overview The in formation and figures contained in this sec tion were drawn from r eferences 1 as s hown in the reference section.

The NuScale conceptual module, illus trated in Figures 2-1 through 2-4, is an integrated light water r eactor ( LWR) w ith passive safety featur es and a power r ating o f approximately 45 MWe

8 (160 MWt). The pressurizer, steam generator, hot leg, cold leg, and core are all housed in a shared reactor pressure vessel. A relatively sma ll steel containment (approximately 60' x 15' or 18 m x 4.6 m) compare d with that of a conventional PWR, envelopes the reac tor pressure vessel. The containment vessel is partially evacuated during power operation and is capa ble of relatively high press ures during accident conditions. The entire module and containment are submerged in a pool of water. The NuScale plant is intende d to be of modular design with up to 12 NuScale modules and containments at one site. Each module in the pool is supported by walls separating it from the other modules in the reac tor pool. The reactor pool is a stainless steel-lined concrete pool s hared by all of the operating modules. Each module is covered by an individual conc rete impact shield, and all of the modules and pool are enclosed in a single confinement building.

Containment Reactor Vessel Helical Coil Steam Generator Nuclear Core Containment Trunnion Figure 2-1 Two views of the NuScale module. The view on the left also depicts a portion of the walls supporting the containment in the reactor pool 9

Generator Steam Turbine Condenser Reactor Module and Containment Water-Filled Pool Below Ground Figure 2 A perspective view of a portion of the plant with the water filled pool and one reactor module and containment 10 Figure 2-3 Planar view of a plant, with maximum modules Figure 2-4 Cross sectional view of plant with elevations 11 The NuScale conceptual design relies on passive safety systems and incorporates all large piping paths into the reactor vessel. The vend or postulates that the use of passive safety systems for decay heat removal, emergency core cooling, and containment cooling will eliminate external power requirements under accident condi tions. The NuScale modules, control room, and spent fuel pool are all located below grade and housed in controlled-access buildings.

The primary side flow path is shown in Figure 2-5. The core is located inside a shro ud connected to the hot leg riser. Subcooled water enters the core, where it is heated and then flows vertically into the riser sectio n. Circulatio n continues as hot water exits the riser into the upper plenum and then turns downward into the annulus housing the steam generators. Hot water in the annulus between the riser and the inside wall of the reactor vessel is cooled by the steam generator tubes. The cooled, denser water descends throug h the downco mer into the lower plenum, then re-enters the core.

CONTAINMENT Control Rod Drives Reactor Vent _

Valves Control Rods SG Annulus/

Cold Leg __

Feed Header Downcomer Shroud -

Figure 2-5 Primary Side Flow Path 12 REACTOR PRESSURE VESSEL Upper Plenum Steam Header Steam Generator Tubes Hot Leg Riser Sump Reclr Valves Core 2.2 Reactor Core The NuScale module's nuclear core is shown in Figure 2-6 and consists of 37 fuel assemblies arranged in a 17 x 17 square array, -. The core includes 16 control rod clusters.

Each fuel assembly includes 264 fuel pins, 24 control rods, and one instrument tube.

Figure 2-6 Core Configuration 2.3 Pressurizer The pressurizer in the NuScale module's Nuclear Steam Supply System provides reactor coolant system pressure control. A baffle region is located above the steam generator region to provide a barrier between the saturated fluid within the pressurizer and the subcooled reactor coolant system fluid. This baffle region limits the temperature of fluid that may surge into or out of the pressurizer region by mixing and heating the fluid as it moves about this region.

2.4 Helical Coil Steam Generator The NuScale steam generator is a helical-coil, once-through heat exchanger located in the annular space between the hot leg riser and the reactor vessel's inside wall. Feedwater enters the tubes at the bottom, and superheated steam exits at the top. Two independent sets of steam generator tube banks occupy the steam generator region as observed in Figure 2-7.

13 Figure 2-7 Helical Coil Steam Generator 2.5 High-Pressure Containment

NuScales containment vessel is dry and partially evacuated under normal operating conditions.

The vendor expects this configuration to eliminate moisture problems that could cause component corrosion and impact the reliability of instrumentation and other systems within containment. The partial vacuum reduces convection heat transfer without the use of direct-contact reactor vessel insulation. Due to a lack of appreciable amounts of air, the vendor also expects the vacuum to enhance steam condensation rates during reactor vessel blowdowns and prevents the formation of combustible concentrations of hydrogen mixtures in the event of a severe accident. The containment vessel is roughly cylindrical in shape approximately 60 x 15 (18m length x 4.6m diameter).

14 Figure 2-8 High Pressure Containment 2.6 Decay Heat Removal System NuScale's decay heat removal system (DHRS) removes residual heat from the reactor core due to decay heat generation. The DHRS provides cooling for the core during normal shutdowns, station blackouts, and/or transients that result in a loss of normal feedwater. It has two independent piping trains, each intended to be capable of passively removing a sufficient fraction of the post-trip core power to prevent damage due to system heat-up.

During DHRS operation, cold water from the containment cooling pool is drawn from the inlet screen and sent to the steam generator tubes, where it transfers heat from the primary fluid and is evaporated. This steam is then vented and condensed in the containment pool. The steam generator removes heat from the reactor coolant in the reactor vessel annulus, creating a density difference between the hotter, lower-density coolant inside the riser and the cooler, higher-density coolant outside the riser. This density difference creates natural circulation of the reactor coolant in the same manner as during normal operation, but at a reduced flow rate. The check valves at various points in the DHRS prevent reverse flow.

15 Each DHRS train has an inlet screen, an inlet line that connects the cooling pool to the main feedwa ter line, an inlet valve, an inlet isola tion valve, an outlet line, an outlet isolation valve, and a vent sparger on the outlet. In addition, each feedwater line include s a pre-pressurize d, water filled accumulator to provide continual feedwater flow during natural circulation startup of the DHRS.

Spargers Decay Heat Removal Sump -

Line Figure 2-9 Decay Heat Removal System

2. 7 Containment Heat Removal System Following a postulated loss-of-coolant accident (LOCA), the containment heat removal system (CHRS) rapidly reduces the containment pressure and temperature, consistent with the functioning of other associated systems, and maintains them at acceptably low levels for extended periods of time. The CHRS, an engineered safety feature, is classified as a "system,"

even though it only consists of the containment cooling pool water and containment vessel. The containment cooling pool consists of a large, below-grade concrete pool that is designed to provide stable, ample cooling for the containment for at least 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> following any LOCA without any active heat removal from the pool. Following a postulated break in the primary system, steam released into the containment would be condensed on the inside surface of the containment wall, which, in turn, would be passively cooled by conduction and convection heat transfer with the cooling pool. Because the containment would be evacuated during normal operation, a low level of noncondensable gases would be present inside the containment, and condensation heat transfer rates would increase.

16 Sump Reclr Valvos (Inlet)

Figure 2-10 Containment Heat Removal System 2.8 Emergency Core Cooling System NuScale's emergency core cooling system (ECCS) consists of two independent reactor vent valves (RWs), two independent reactor recirculation valves (RRVs), and the CHRS. The ECCS provides a means of core decay heat removal in the event of loss of the main feedwater flow in conjunction with the loss of both trains of the DHRS. Long-term cooling is established through reactor recirculation cooling via the ECCS flow path in the event of a LOCA. The ECCS is initiated by open ing the RWs and RRVs. Opening these valves creates a path by which water condensed on the inside surface of the containment flows into the reactor coolant system via the RRVs. Opening the RWs establishes a natural circulation path whereby water that is boiled in the core leaves through the RWs, is condensed and collected in the containment, and then is reintroduced into the downcomer through the RRVs.

3 PIRT METHODOLOGY 3.1 General Approach Development of a Phenomena Identification and Ranking Table (PIRT) for a new reactor design can be characterized as a systematic, multi-step process to determine which physical 17 phenomena are of most i mportance to a pa rticular scenar io. Becaus e of the s ize and complexity of t he r eactor sy stem, and it s ex pected response t o v arious boundary condit ions and system int eractions, the system needs t o be ev aluated in terms o f its individual components such as t he cor e and steam generator. The ov erall objective of the PI RT i s t o de fine requirements and capabilities of an Ev aluation Model (EM). Spec ific steps that are par t of the PIRT process are:

1. Definition of Objectives: The first step is to specify the purpose of the analysis required by the EM and to identify the type of reactor system for which the PIRT applies.

2. Identification of Systems and Subcomponents: A reactor coolant system often contains several safety-related components. These components and interacting systems should be identified as part of the PIRT ranking process.

3. Scenario Selection: The accident type and general scenario are identified. The scenario can be sub-divided into several periods as phenomena evolve or as various sub-systems become activated or de-activated.

4. Specify Figure(s) of Merit: Figures of merit are quantitative standards of acceptance for the safety analysis. W hile the General Design Criteria (GDC) in Appendix A of 10CFR50.46 describe general requirements for safe reactor operation, specific parameters that are outputs of the EM are desirable as figures of merit for a PIRT.

5. Identification of Phenomena and Importance Ranking: Physical phenomena and processes that affect each component are identified and ranking in terms of importance with respect to the figure(s) of merit.

6. Knowledge Level Ranking: The supporting knowledge base and uncertainties associated with processes and phenomena should be characterized. The ranking of the knowledge base should be based on the state-of-the-art ability to understand and model a process, rather than a statement on whether the process is modeled acceptably in a specific thermal-hydraulic code.

7. Documentation: The documentation of a PIRT should include the phenomena and their rankings, along with the rationale by which the panel arrived at its ranking.

3.2 Application to Preliminary Designs The PIRT process described in the preceding section was applied to the plant as reported in Section 2. Additional information was reviewed by the panel in References 1-3. W hen the licensing basis design is available, assumptions, results, and overall applicability of this PIRT should be re-evaluated.

3.3 Ranking Scales

3.3.1 Phenomena Importance Ranking The panel determined the importance of a particular phenomenon based on expected effect the phenomena may have on the Figure(s) of Merit (FOMs) for the scenario. A summary of the importance rankings is provided in Table 3-1. Phenomena ranked high (H) are considered likely to have a dominant role in the transient. Phenomena with a high (H) rank would be expected to be modeled with a high degree of accuracy, or be addressed through sensitivity studies to determine the resulting uncertainty in the FOM. Phenomena receiving lower importance rankings [medium (M) or low (L)] are expected to have relatively less impact on the transient, and it may be possible to model such processes with lesser accuracy or lesser concern with the uncertainty in calculating the FOM. The inactive (I) ranking is reserved for those processes that either do not or cannot occur during parts of the transient.

18 Table 3-1 Phenomena Importance Ranking Table

Ranking Description High (H) Significant or dominant influence on one or more FOM.

Medium (M) Moderate influence on one or more FOM Low (L) Small influence on one or more FOM Inactive (I) Phenomena not present or possible

3.3.2 Knowledge Level Ranking The panel also commented on the state of knowledge for the processes receiving an importance ranking. The knowledge level ranking is meant to serve as an assessment of technical information available for the phenomenon as it may occur in the full-scale prototype during the transient. The knowledge level ranking is an assessment of the ability to model and simulate the phenomena accurately. It is not an assessment on the adequacy of how well the phenomenon is modeled and simulated by a particular analysis code. That type of an assessment should be made in a Code Applicability Report where the codes models and correlations are considered relative to the needs defined by the PIRT for a given transient scenario. Table 3-2 summarizes the knowledge level ranking criteria for this PIRT.

Table 3-2 Knowledge Level Ranking Criteria

Ranking Description

High (H) Phenomenon is well understood, and can be accurately modeled.

Medium (M) Phenomenon is understood, however may only be modeled with moderate uncertainty.

Low (L) Phenomenon is not well understood. Modeling the phenomena is currently either not possible or is possible only with large uncertainty.

4 THERMAL-HYDRAULIC PIRT

4.1 PIRT Panel Members

The following individuals participated on the panel to develop a thermal-hydraulic PIRT for the NuScale reactor:

• Dr. Stephen M. Bajorek (Chair), Senior Advisor for Thermal-Hydraulics, NRC/RES/DSA

• Dr. David Diamond (Facilitator), Brookhaven National Laboratory

• Dr. Peter Griffith, Professor Emeritus, MIT

• Dr. W illiam Krotiuk, NRC/RES/DSA/RSAB

• Dr. Peter Lien, Senior Reactor Engineer, NRC/RES/DSA/RSAB 19

• Dr. Benjamin Parks, Senior Reactor Engineer, NRC/NRR/DSS/SRXB

• Dr. Joseph Staudenmeier, Senior Reactor Engineer, NRC/RES/DSA/CDB 4.2 Scenario No large-break LOCA is possible in the NuScale reactor due to its integral design and absence of large penetrations in the reactor vessel. Thus, only small-break LOCAs need to be considered. Based on information on the NuScale design, there are a limited number of vessel penetrations that may result in significant inventory loss. The inadvertent opening of valves at the top of the pressurizer would result in a rapid depressurization, but the flow out of the vessel would be primarily steam. The lowest possible penetration to the reactor vessel is the reactor recirculation valve (RRV). An inadvertent opening of one of these valves would immediately result in the discharge of subcooled liquid, which would rapidly deplete vessel inventory and have a faster depressurization of reactor and faster pressurization of the containment vessel than a release from the top of the vessel. Therefore, the thermal-hydraulic accident scenario considered for this PIRT is the inadvertent opening of a reactor recirculation valve.

The assumed scenario for the inadvertent opening of a RRV follows. The LOCA is divided into three distinct periods, as outlined in Table 4-1.

Table 4-1 Periods Considered in NuScale Thermal-Hydraulic PIRT

Period Period Description Start 1 Blowdown Reactor recirculation valve opens.

2 RVV Depressurization Reactor vent valves open on low primary pressure (6 MPa) 3 Long Term Cooling Primary and containment pressure equilibrate.

Period 1, the blowdown phase begins with the inadvertent opening of one of the reactor recirculation valves. Subcooled critical flow is discharge from the vessel to containment during this initial period of the transient as the vessel depressurizes. Almost immediately, a reactor protection system trip signal occurs on loss of containment vacuum. The containment vessel begins to flood, the turbine and main feedwater pumps trip, and both the main steam line and the main feedwater lines are isolated. The pressurizer water level decreases and eventually reaches the level of the baffle plate. W hen the pressurizer pressure decreases to 6 MPa (870 psia), the reactor vent valves begin to open.

Period 2, the reactor vent valve depressurization period begins when valves at the top of the pressurizer open and release steam into the containment vessel. This increases the depressurization rate and during this period inventory is lost to containment from both the reactor vent valves and the inadvertently open reactor recirculation valve. Steam condenses on the containment vessel wall and drains into the pool of water collecting in containment. This containment pool level increases as level within the reactor vessel decreases. Heat is also removed by the helical coiled steam generators, from inventory held inside the tubes which eventually boil dry since the feedwater pumps have tripped. The second reactor recirculation

20 valve opens near t he end of this per iod following a timed delay and the pr essures in the r eactor vessel and containment near e quilibrium.

Period 3, t he final period, is long-term cooling. This period begins once both RRVs are open and the pr essure has equilibrated between reactor v essel and t he containment vessel. The water lev el in containment is expected t o be abov e the RRVs and higher than the lev el in the reactor v essel. The hy drostatic head will force water into the RRVs, w hile steam continues to be released from t he RVVs. Steam en tering c ontainment cont inues to be condensed on t he containment vessel wall and drain into the c ontainment water. Heat is remov ed from t he containment to the water in the con tainment vessel cooling pool which surr ounds that vessel.

4.3 Figures of M erit

The Fig ure(s) o f Merit (FOM) for an accident s cenario refer t o t he par ameters by w hich success or failure of t he s afety sy stems is determined. Because the acciden t scenario can have several periods, the FOM m ay c hange and depend on the specific period. For s mall modular r eactors, the FOM cr iteria as speci fied by 10 CFR 50. 46 (peak cladding temperature and t he maximum local clad oxidation) ar e not ex pected t o be challenged. The succes s o f the safety sy stems is therefore j udged on parameters that ens ure a w ide m argin to t he 10 CFR 50. 46 regulatory criteria.

A loss-of-coolant accident, such as the inadvertent openin g o f a RRV or RVV, is not expect ed to result in cor e uncov ery. To det ermine if cladding heat up oc curs as a result o f uncovery, the two-phase mixture lev el is select ed as a figure o f merit. T he mixture level refers to t he minimum level in the core o r v essel where sufficient liquid is pr esent to p revent d ryout. As long as t he mixture level, w ith uncer tainty, remains above the t op o f the cor e, there is little likelihood that cladding temperatures will exceed the local saturation temperature. The mix ture level is considered important during the RVV depressurization and long-term cooling periods of a LOCA transient in the NuScale-like design. Because the blow down per iod is short and t he w ater lev el remains in t he pr essurizer, the m ixture lev el was not conside red important in that early per iod.

If a signif icant temperature increase in t he fuel and cladding occurs du ring a LOCA, it is considered more lik ely t o be t he result of hig h-power locat ions exceeding the critical heat flux (CHF). The cr itical heat flux is that heat flux at which there is sufficient vapor blan keting or bubble clotting on a f uel rod such t hat heat is not effectively r emoved and a r apid temperature rise occurs. Since t his phenomenon was considered possible only w hen the power w as high, CHF was selected as a FOM only dur ing t he first two periods of t he s cenario.

NuScale transients, with coolant dischar ge from the reactor v essel to t he containment v essel, are considered tightly c oupled. That is, the tr ansient in t he reactor v essel can be strongly influenced by phenomena and conditions in the containment. The cont ainment pressure is expected to incr ease r apidly dur ing a RRV or RV V opening, and flow t o the reactor v essel depends on when the pressur es bet ween the t wo vessels equilibrate. The NuScale containment vessel is also an important barrier to t he release of fission products, and it must be demonstrated that pressures and t emperatures remain within accept able design limits.

21 Therefore, during the blowdown and RVV depressurization periods, containment pressure is considered an im portant FOM.

For t he inadvertent opening of an RVV in a NuSca le system, the F OMs by per iod are list ed in Table 4-2.

Table 4-2 Figures of Merit for NuScale T/H PIRT

Period CHF Mixture Level Containment Pressure Blowdown X X RVV Depressurization X X X Long-Term Cooling X

4.4 Systems, Subsystems and Components

The panel divided the NuScale design into systems, subsystems and components in order to better identify significant phenomena that influence the FOM. The major systems were considered to be the Core and the Primary System, which is composed of the riser, upper plenum, pressurizer, steam generator annulus, downcomer, and the lower plenum. The reactor vent valves and reactor recirculation valves were also considered to be parts of the primary system. The steam generator secondary side was considered to be an individual system. The containment heat removal system (CHRS) is composed of the containment vessel and containment vessel cooling pool. Table 4-3 summarizes the main systems, subsystems and components considered in the NuScale PIRT.

22 Table 4-3 Systems and Subsystems System Subsystems

Core Fuelrods Coolant Subchannel Barrel/Baffle Control Rods/ Guide Tube Corewide flow

Primary Riser Upper plenum Pressurizer Reactor vent valves Steam generator annulus Downcomer Reactor recirculation valves Lower plenum

Secondary Steam generator tubes (outside)

Containment Heat Removal Containment vessel System (CHRS) Containment cooling pool

4.5 Phenomena Ranking and Discussion This section presents the ranking of each phenomenon, and provides a brief rationale on how or why the panel arrived at the ranking. Each system, core, primary, secondary, and CHRS, is discussed. Table 4-4 lists the phenomena, along with the importance rank and the knowledge level assessment.

4.5.1 Core System Phenomena

4.5.1.1 Fuel Rods The fuel in the NuScale core is conventional PWR reactor fuel. However, the core and rods are approximately one-half height compared to most operating PW Rs. The panel considered the fuel to be composed of the UO2 fuel pellets, the gap between the pellets and the cladding, and the zircalloy-based cladding. Some of the design parameters, such as initial internal pressure, were not available at the time this PIRT was developed.

The panel considered Decay Heat to be a highly ranked process throughout the transient due to its effect on critical heat flux and mixture level. The knowledge level for Decay Heat was considered to be high. Other fuel-related parameters and phenomena such as stored energy, gap conductance and fuel thermal conductivity, were not considered to be of much importance since the core is covered with liquid and most of the initial energy will be transferred to the coolant early in the event. The total peaking factor, both magnitude and location, was given an M importance rank during blowdown because of its potential impact on CHF. Core design methods were considered sufficiently accurate such that the value and location of the peak power location would be known and therefore given a high knowledge level ranking. Because

23 the cladding is expected t o rem ain covered and to r emain near sa turation temperature, cladding oxidation was not con sidered an act ive process and therefore was assigned an I ranking.

4.5.1.2 Coolant Subchannel Coolant s ubchannel processes and phenom ena a re t hose that occur in the fluid. These include single-a nd two-phas e pressure drop and fl uid mixing. During the blow down period, pr ocesses that were c onsidered m ore im portant were thos e that de termined the velocity and void distributions as the cor e flow st agnates ear ly in the event. Two-phase pres sure drop, natural circulation, CHF (correlations), flow r egime transition, c ross flow / mixing, a nd void distribution were assigned an M im portance r ank as these were considered processes that w ould determine the onse t of CHF. If CHF, by bubble clotting o r v apor blanketing, w ere t o occu r in an RRV or RVV inadvert ent opening in a NuScale reactor, it is much m ore li kely t o occur dur ing blowdown as opposed to lat er per iods. Therefore, most of the phenomena im portant for blowdown CHF were consider ed of low im portance during the RVV Depressurization and Long-Term Cooling pe riods. Natural circulation retained an M im portance ranking during the RVV Depressurization period and Interf acial Drag was ranked M f or both t he RVV Depressurization and Long Term Cooling periods due to t heir im portance on mixture level. The knowledge level of most processes w as considered M or H throughout t he transient. The CHF (correlations) were g iven a low k nowledge lev el ranking be cause available correlations were consider ed likely be out of range at the lo w f low r ates ex pected during RVV Depressurization and Long-Term Cooling.

4.5.1.3 Barrel/Baffle The Bar rel/Baffle subsy stem r efers to t he region surrounding the core. This regi on may r esult in a bypass of coolant from t he cor e, and lar ge str uctures m ay con tribute to heat release during the event. The panel considered processes asso ciated with this regi on t o be unimportant because the hea t transfer is insignificant relative to other co mponents, and assi gned an L importance ranking. The knowledge level of t he p rocesses w as considered high.

4.5.1.4 Cont rol Rods / G uide Tube The e ffect of the control rods and guide t ubes on the flow dist ribution and h eat release by t he structures w as considered by t he panel. In each cas e the po tential impacts were considered small, and each process was as signed a low im portance r anking. Knowledge lev el was considered high.

4.5.1.5 Corewide Flow Corewide flow r efers to the dis tribution of fluid velocity lat erally acr oss t he core. The panel considered this as a potential c ontributor to flow-induced stability, bu t decided that a low importance ranking was appropriate.

4.5.2 Primary S ystem Phenomena

4.5.2.1 Riser The hot leg ris er con tains flow t hat goes from t he c ore t o t he upper plenu m. Since th is fluid is at Thot, flashing will occur in t his component be fore it occurs elsewhere in the vessel. The panel considered flas hing to be a process o f high im portance for t he riser due to its effect on void formation. Primary nat ural circulation, and its e ffect on t he bul k coolan t flow, w as considered as

24 a process of medium im portance in th e riser du ring the init ial blowdown p eriod but of low importance for lat er per iods as the lev el dropped below t he top of t he riser and the path for natural circulation was broken. Interfacial drag between the phases, and two-phase level swell were considered to be o f medium im portance during the RVV Depressurization and Long-Term Cooling per iods as these pr ocesses contribute to the determination of t he mixture level.

Vertical and radial natural cir culation refers to na tural circulation internal to t he r iser it self in which hot fluid in the cen ter o f the r iser flows up while cooler fl uid flows down along t he r iser wall. Sinc e a large temperature difference ex ists acros s t he r iser w all during most of the transient, this process w as given a low im portance ranking by the panel. Inlet flow and temperature distribution refers t o v ariations in velocity and temperature enter ing the r iser f rom the core. These were no t considered significant or to hav e a major e ffect on t he m ixture lev el in the riser and t herefore wer e given a low im portance ranking.

4.5.2.2 Upper Plenum The upper plenum in t he NuScale vessel is the relatively sm all region above the t op o f the r iser but below t he pressurizer. Flow in t his region is at T hot, and th e flow i s tu rning towards the annulus between the ris er and t he v essel outer su rface.

4.5.2.3 Pressurizer The pr essurizer pr ovides reactor coolant system pressure con trol. The pr essurizer is separated from t he rest of the vessel by a baffle t hat serves as a barrier bet ween the saturat ed fluid w ithin the pressurizer and t he s ubcooled reactor coolant system fluid. Th is baffle r egion limits the temperature of fluid that may surge int o or ou t of the press urizer r egion by mixing and heating the fluid as it moves about this region.

The liq uid level in the pressurizer and the vessel pressure ar e im portant setpoints that determine the openin g o f the r eactor v ent v alves. Therefore, du ring blowdown, f lashing w as considered a hig hly r anked phenomenon, and ph ase separation and int erfacial drag/relative motion between the phases were both consider ed to be of medium im portance. During the Reactor Vent Valv e Depressurization period, these processes remain im portant, but because the rate o f depressurization slows somewhat, flashing has lesser im pact than in blowdown.

During the React or Ven t Valve Depressurization per iod flow int o t he pr essurizer t o the vent valves occurs, and two-phase pressure dr op t hrough t he pr essurizer is considered to hav e medium i mportance. Flooding a t the baffle plat e is ex pected to oc cur, limiting the drain rate o f the pressurizer, and is c onsidered to be o f high im portance.

When the Long-Term Co oling per iod occur s, the pres surizer is expected to be empty and t here will be little flow t hough the pressurizer t o t he v ent v alves. All pressurizer proc esses at this t ime were therefore cons idered to hav e low im portance.

4.5.2.4 Reactor Vent Val ves ( RVV)

The r eactor v ent v alves are located at t he top of the pressurizer, and pr ovide a relief path for steam t o ex it t he v essel and condense on the walls of t he containment vessel. The v ent v alves are inactive dur ing the blow down per iod in an inadvertent RRV opening e vent, but affect the transient once t hey open on a low pr essure o r low w ater lev el signal. During the RVV

25 Depressurization and Long-Term Cooling periods, the RVVs act as a seco nd break to t he vessel. Early in time, the flow is expec ted to be choked. As the ves sel level decreases and the pressure beco mes low, the flow m ay becom e unc hoked and single pha se. The panel therefore considered choked flow and single phase press ure drop to be o f high im portance dur ing the RVV Depressurization period, and t hen hav ing le ss importance dur ing Long-Term Cooling as the importance ranks were decr eased to L and M r espectively.

4.5.2.5 St eam Generator Annulus The st eam generator an nulus refers t o phenom ena that occur in the fl uid space outside of t he helical coiled tubes. Because the s team generator is isolated ear ly in the transient and later boils dry, only sev eral processes in t he blowdown per iod were considered to have much importance on t he FO Ms. In par ticular, processes that affect natural circulation through the vessel, flashing, two-phase pressure dr op, and na tural circulation flow w ere considered t o hav e medium i mportance. Th e feed header and steam header ar e pot entially i mportant resistances to flow, and t hese w ere ther efore also given an M im portance r anking. Heat removal during the blowdown period is expected to be by conv ective heat transfer t o t he S G tubes. Since t he rate of heat transfer may r emain significant during blowdown, t his process was given a medium importance ranking.

Following t he blowdown period, flow t hrough t he steam generator annulus is expect ed to be very lo w and heat transfer to and from t he s team generator also very low. Therefore, all steam generator p rocesses w ere considered t o hav e low importance for the RVV Depressurization and Long-Term Cooling Periods.

4.5.2.6 Downcomer The dow ncomer is t he a nnular r egion between the core bar rel and the v essel inner su rface t hat is below t he helical coil steam generator. Flashing, two-phase level swell, int erfacial drag/relative motion between phases, and t he effect of t his r egion on t he natural circulation of bulk flow in the vessel were cons idered to hav e medium i mportance due t o their e ffect on flow t o the core and t he r esulting effect on CHF dur ing the blowdown per iod. During the RVV Depressurization period two-phase level swell, inter facial drag/relative motion between phases and stored energy r elease from t he v essel wall were consider ed to hav e medium i mportance due to t heir e ffect on t he distribution of v oid and fluid temperatures due to the potential impact of flow t o t he r eactor r ecirculation valves. During Long-Term Cooling, only stored ener gy r elease from the vessel w all was consider ed to hav e medium im portance. All other pr ocesses considered were assigned low im portance r ankings.

4.5.2.7 Reactor Recirculation Valves (RRV)

The r eactor r ecirculation valves (RRV) act as t he break for t he inadv ertent opening ev ent.

During blow down, s ingle-and t wo-phase pressure drop, as w ell as choked flow, w ere all considered phenomena of hi gh importance due to their im pact on all three FOMs. After t he RVVs open, flow out the RRVs is expected to r emain choked and w as therefore c onsidered along w ith the t wo-phase pr essure drop to hav e high im portance. During the Long-Term Cooling per iod, flow in and out of t he RRVs is ex pected to be sin gle-phase and unchoked, so single-phase pr essure dr op was considered to have medium im portance with other p rocesses ranked low.

26 4.5.2.8 Lower Plenum Processes in the low er plenum that may hav e an impact on the FO Ms included flashing, single-phase pressure dr op, and the e ffect of that region on natural circulation bul k flow r ate. Each of these phenomena w as assigned a medium import ance rank ba sed on t he possibility that t hese could affect the cor e flow r ate and t herefore the CHF. During the RVV Depressurization and Long Term Cooling periods, ene rgy r elease from the vessel wall w as considered a pr ocess of medium i mportance.

4.5.3 Secondary S ystem Phenomena For an inadv ertent opening of an RRV, the secon dary o f the NuScale design has only a limited involvement. Following loss of con tainment vacuum, the m ain feed pum p trips, and the m ain feed line is i solated. Inventory w ithin the steam generator tubes is bo iled off. Eventually the inside of t he tubes is em pty, and t he generator ce ases to r emove heat from the vessel.

Assuming that most o f the inventory boils away dur ing blowdown or shor tly t hereafter, the phenomena of f lashing, sing le-phase convection heat t ransfer and two-phase c onvection heat transfer w ere given a medium im portance r anking during blow down and a low im portance ranking for la ter per iods. Other phenom ena associated with the secondary side were considered to be o f low i mportance.

4.5.4 CHRS System Phenomena The c ontainment heat removal system (CHRS) is int ended to mainta in the cont ainment environment (pres sure and temperature) a t acceptable levels throughout the entir e transient and for ex tended periods o f time. The CHRS is classified as a sy stem, even though it only consists o f the cont ainment v essel itself and t he containment coolin g pool t hat surrounds it.

The cont ainment cooling pool consists of a lar ge, below-grade conc rete po ol that is designed t o provide stable, a mple cooling for t he c ontainment for a t least 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> s following any LO CA without any act ive heat remova l from the pool.

4.5.4.1 Containment Vessel Following a postulated break in the pr imary sy stem, steam r eleased into the containment would be condensed on the inside surface of the cont ainment wall, w hich, in turn, would be passively cooled by conduct ion and convection heat transfer w ith the c ooling pool. Because the containment would be evacuated during normal operation, a low lev el of n oncondensable gases would be present inside the containment.

Since condensation heat transfer is cr itical to removing heat from t he v essel, phenom ena that affect condensation rates were considered o f high importance t o t he det ermination of the containment pressure F OM. During all three per iods of the tr ansient, processes of high importance were c ondensation heat transfer on the wall of the containment vessel, nat ural circulation flow in the gas-spac e o f the c ontainment v essel, and t he e ffect of non-condensables.

While the NuScale design asserts t hat a hig h v acuum w ill be maintained during normal operation, the panel considered it unlikely t hat all noncondensables could be eliminated or prevented during a L OCA. Non-condensables m ay be pr esent from disso lved nitrogen com ing out o f solution from t he c oolant, for ex ample. During blow down, the e ffect of thermal properties for fluid below at mospheric pressure w as assigned a medium im portance r anking due to the potential uncertainties it may cr eate in an analy sis. Conduction heat transfer t hrough the

27 reactor vessel wall and the containment vessel wall were cons idered processes of medium importance due to the resistance to heat transfer that they represent.

By the start of the Long-Term Cooling period, a liquid level will have developed in the containment vessel. Thermal stratification, mixing by internal natural circula tion, and heat transfer between the pool and the gas space will affect the containment pres sure as well as the temperature of fluid returning to the reactor vessel through the RRVs. Therefore, the panel assigned a medium importance rank to stratification, natu ral circula tion, and interfacial heat transfer at the pool surface for this period.

4.5.4.2 Containment Cooling Pool The containment cooling pool was assumed to contain a large volume of water that would require a very long period of time to heat significantly. The panel assigned low importance rankings to all processes for the blowdo wn and RW Depressurization periods since natural convection heat transfer should be sufficient to remove heat from the containmen t vessel. Late in the transient, during the Long-Term Cooling period, some other heat transf er processes may become significant and important to the FOMs. Therefore, in Long-Term Cooling, nucleate boiling, single phase convective heat transfer, natu ral circulation in the cooling pool, and thermal stratification in the pool were considered to be of medium importance.

Table 4-4 Phenomena and Knowledge Level Rankin gs System Component Process/Phenomena BLD RVV AP LTC KL CORE FUEL RODS Decay Heat H H H H CORE FUEL RODS Fission Power L L L H CORE FUEL RODS Stored Energy L L L M CORE FUEL RODS Gap Conductance L L L M CORE FUEL RODS Fuel Conductivity L L L M CORE FUEL RODS Initial Gap Pressure L L L M CORE FUEL RODS Cladding Conductivity L L L M CORE FUEL RODS Cladding Oxidation (CORRE LATIONS) I I I H CORE FUEL RODS Total Peaking Factor M L L H CORE FUEL RODS Burnup Distribution L L L H CORE FUEL RODS Boron Precipitation L L L M System Component Process/Phenomena BLD RVV LTC KL AP CORE SUBCHANNEL Single Phase Pressure Drop L L L H CORE SUBCHANNEL Two Phase Pressure Drop M L L M CORE SUBCHANNEL Flashing L L L H CORE SUBCHANNEL Natural Circulation M M L M CORE SUBCHANNEL lnterfacial Drag L M M M CORE SUBCHANNEL Single Phase Convection L I I H CORE SUBCHANNEL Two Phase Convection L L L H CORE SUBCHANNEL CHF (correlations) M L L L CORE SUBCHANNEL Flow regime transition M L L M Grid spacer effects CORE SUBCHANNEL ( entrainment/ deentrainment) I I I M 28 CORE SUBCHANNEL Grid spacer effects (heat transfer) L L L M CORE SUBCHANNEL Cross flow / mixing M L L M CORE SUBCHANNEL Clad ballooning L L L M CORE SUBCHANNEL Void distribution M L L M CORE SUBCHANNEL Turbulent mixing L L L L CORE SUBCHANNEL Boron blockage in subchannels L L L M System Component Process/Phenomena BLD RVV LTC KL AP CORE Barrel/Baffle Stored energy L L L H CORE Barrel/Baffle Bypass flow L L L H System Component Process/Phenomena BLD RVV LTC KL AP CORE Control rods/GT Effects on flow L L L M CORE Control rods/GT Effects on heat transfer L L L H System Component Process/Phenomena BLD RVV AP LTC KL CORE COREW IDE FLOW Stability L L L M System Component Process/Phenomena BLD RVV AP LTC KL PRIMARY Hot leg riser Flashing H H L H PRIMARY Hot leg riser Two-phase level swell L M M M PRIMARY Hot leg riser Two-phase pressure drop L L L M PRIMARY Hot leg riser Primary natural circulation flow/bulk flow M L L M PRIMARY Hot leg riser lnterfacial drag/relative motion of phases L M M M PRIMARY Hot leg riser Vertical/radial natural circulation L L L L PRIMARY Hot lee riser Inlet flow/temp distribution L L L M PRIMARY Hot leg riser Convection heat transfer to shroud/riser L L L H PRIMARY Stored energy release/conduction of Hot leg riser shroud/riser L L L H PRIMARY Hot leg riser Radiation heat transfer from shroud/riser L L L H PRIMARY Control rod drives/supports structures affect Hot leg riser on flow L L L H PRIMARY Hot lee riser Riser Bypass Flow L L L H PRIMARY Hot leg riser Mixing L L L H System Component Process/Phenomena BLD RVV LTC KL AP PRIMARY Upper plenum Flashing H L L H PRIMARY Upper plenum Two-phase level swell L L L M PRIMARY Upper plenum Two-phase pressure drop L L L M PRIMARY Upper plenum Single-phase pressure drop L L L H PRIMARY Upper plenum Primary natural circulation flow/bulk flow M L L M PRIMARY Upper plenum lnterfacial drag/relative motion of phases L L L M PRIMARY Upper plenum Vertical/radial natural circulation L L L M PRIMARY Upper plenum Convection heat transfer to reactor vessel L L L H PRIMARY Stored energy release/conduction of vessel Upper plenum wall L L L H PRIMARY Radiation heat transfer from reactor vessel to Upper plenum conta inment vessel L L L H System Component Process/Phenomena BLD RVV LTC KL AP PRIMARY PRESSURIZER Flashing H M L H 29 PRIMARY PRESSURIZER Phase separation M M L M PRIMARY PRESSURIZER Flooding at baffle plate L H L M PRIMARY PRESSURIZER Two-phase pressure drop L M L M PRIMARY PRESSURIZER Single-phase pressure drop L L L H PRIMARY PRESSURIZER lnterfacial drag/relative motion of phases M M L M PRIMARY PRESSURIZER Primary natural circulation flow/bulk flow L L L H PRIMARY PRESSURIZER Convection heat transfer to vessel L L L H Process/Phenomena BLD RVV LTC KL System Component AP PRIMARY Reactor vent valves Two-phase pressure drop I M L M PRIMARY Reactor vent valves Single-phase pressure drop I H M M PRIMARY Reactor vent valves Choked flow I H L M System Component Process/Phenomena BLD RVV AP LTC KL PRIMARY SGANNULUS Flashing M L L H PRIMARY SG ANNULUS Two-phase level swell L L L M PRIMARY SGANNULUS Two-phase pressure drop M L L M PRIMARY SG ANNULUS Single-phase pressure drop L L L H PRIMARY SGANNULUS Primary natural circulation flow/ bulk flow M L L M PRIMARY SG ANNULUS lnterfacial drag/relative motion of phases L L L M PRIMARY SGANNULUS Vertical/radial natural circulation L L L L PRIMARY SGANNULUS Convection heat transfer to SG tubes M L L M PRIMARY SG ANNULUS Convection heat transfer to vessel L L L H PRIMARY SGANNULUS Convection heat transfer to riser L L L H PRIMARY SG ANNULUS Stored energy in the steam generator tubes &

fluid L L L M PRIMARY SG ANNULUS Stored energy release/conduction of vessel wall L L L H PRIMARY SG ANNULUS Stored energy release/conduction of riser L L L H PRIMARY SGANNULUS Radiation heat transfer from vessel L L L H PRIMARY SG ANNULUS Radiation heat transfer from shroud/riser L L L H PRIMARY SGANNULUS Tube bypass flow M L L M PRIMARY SG ANNULUS Feed header effect on flow M L L M PRIMARY SGANNULUS Feed header stored energy M L L M PRIMARY SG ANNULUS Steam header effect on flow M L L M PRIMARY SG ANNULUS Steam header stored energy M L L M Process/Phenomena BLD RVV LTC KL System Component AP PRIMARY DOWNCOMER Flashing M L L H PRIMARY DOWNCOMER Two-phase level swell M M L M PRIMARY DOWNCOMER Two-phase pressure drop L L L M PRIMARY DOWNCOMER Single-phase pressure drop L L L H PRIMARY DOWNCOMER Primary natural circulation flow/ bulk flow M L L M PRIMARY DOWNCOMER lnterfacial drag/relative motion of phases M M L M PRIMARY DOWNCOMER Vertical/radial natural circulation L L L L PRIMARY DOWNCOMER Convection heat transfer to vessel L L L H PRIMARY DOWNCOMER Convection heat transfer to shroud/riser L L L H PRIMARY Stored energy release/conduction of vessel DOWNCOMER wall L M M H 30 PRIMARY DOWNCOMER Stored energy release/conduction of riser L L L H PRIMARY DOWNCOMER Radiation heat transfer from vessel L L L H PRIMARY DOWNCOMER Radiation heat transfer from shroud/riser L L L H System Component Process/Phenomena BLD RW LTC KL AP PRIMARY Reactor Recirc.

Valves Single-phase pressure drop H L M H PRIMARY Reactor Recirc.

Valves Choked flow H H L M PRIMARY Reactor Recirc.

Valves Two-phase pressure drop H H L M System Component Process/Phenomena BLD RW LTC KL AP PRIMARY Lower plenum Flashing M L L H PRIMARY Lower plenum Two-phase level swell L L L M PRIMARY Lower plenum Two-phase pressure drop L L L M PRIMARY Lower plenum Single-phase pressure drop M L L H PRIMARY Lower plenum Primary natural circulation flow/bulk flow M L L H PRIMARY Lower plenum lnterfacial drag/relative motion of phases L L L M PRIMARY Lower plenum Vertical/radial natural circulatio n L L L L PRIMARY Lower plenum Convection heat transfer to vessel L L L H PRIMARY Lower plenum Stored energy release/ conduction of vessel L M M H System Component Process/Phenomena BLD RW AP LTC KL SECONDARY SG tubes Void distribution L L L L SECONDARY SG tubes Single-phase convection heat transfer M L L M SECONDARY SG tubes Two-phase convection heat transfer M L L M SECONDARY SG tubes Flashing M L L H SECONDARY SG tubes Stored energy of tubes L L L H System Component Process/Phenomena BLD RW AP LTC KL CHRS Containment vessel Condensation heat transfer H H H M Vertical/radial natural circulation flow (gas CHRS Containment vessel space) H H H M CHRS Containment vessel Single-phase convection heat transfer L L L H Conduction heat transfer (vessel and CHRS Containment vessel conta inment) M M M H CHRS Containment vessel Effect of noncondensables H H H L CHRS Containment vessel Fluid properties for pressures < 0.10132 MPa M I I H CHRS Containment vessel Thermal strat ification L L M L Vertical/radial natural circulation flow (liquid CHRS Containment vessel space) L L M L CHRS Containment vessel lnterfacial heat / mass transfer (pool to gas) L L M L System Component Process/Phenomena BLD RW LTC KL AP Containment CHRS cooling pool Nucleate boiling L L M H Containment CHRS cooling pool Single-phase convection heat transfer L L M H Containment CHRS cooling pool Vertical/radial natural circulation L L M L Containment CHRS cooling pool Thermal strat ification L L M L 31 4.6 Phenomena / Knowledge Level Issues Of par ticular i mportance ar e the processes and phenom ena that are consi dered important, yet are also considered poorly under stood. This com bination can lead to difficulties in analysis of the transient and/or hig h uncertainty in the predicted results. Th is section summarizes the PIR T rankings.

4.6.1 Blowdown The panel considered 12 0 processes and pheno mena during the Blow down period for t he NuScale design. Of th is 120, 11 ( 9%) w ere consider ed of high ranking im portance. Only t wo were consider ed of high or medium i mportance, but with low k nowledge level. These were CHF (correlations) in the co re and the e ffect of non-condensables in the con tainment vessel. The lack of knowledg e for CH F in the cor e is due t o the relatively poor under standing o f CHF in par t-length r od bundles with low v elocity coolant. The panel did not ex pect this t o becom e a m ajor safety issue be cause the lik elihood of exceeding CHF seemed t o be remote and conser vative estimates of CHF would probably be sufficient. However, t he M knowledge lev el is an acknowledgement of a spare database at c onditions of interest.

During blow down, t he major conce rn may be that of t he e ffect of non-condensables on condensation heat transfer in the containment v essel. While the NuScale design int ends to eliminate this as a factor, the panel was concerned with t he effect of tr ace amounts on the FOMs. Because o f the d ifficulty in determining co ndensation rates w ith the pr esence o f non-condensables, the panel identified t his as a pr ocess with high im portance but low k nowledge level.

4.6.2 RVV Depressurization For t he RVV Depressurization period, 125 pr ocesses and phenomena w ere considered. Of these 10 ( 8%) w ere considered of high im portance. Of these, only t wo were considered to be o f high o r medium im portance, but with low k nowledge lev el. These were the effect of non-condensables on containment condensation, and the effect o f stored energy heat r elease in the lower plenum. The e ffect of non-condensables is due to it s e ffect on t he containment pressure FOM in this period. T he effect o f stored energy r elease from t he low er plenum w as due to its impact on the CHF FOM and the di fficulty in determining the heat transfer from a complex geometry.

4.6.3 Long-Term Cooling During the Lon g-Term C ooling per iod, only 4 (3%) o f 122 act ive processes w ere considered to be of high im portance. However 6 processes were identified as having high or medium importance with a low k nowledge level. The e ffect of non-condensables on cont ainment condensation rate w as found to be important with low k nowledge level. There were 5 process es in the CHRS with m edium im portance and low k nowledge level. These w ere thermal stratification and int ernal natural circulat ion (in both the cont ainment liquid and the cooling pool),

and interfacial heat transfer between the cont ainment liquid and t he gas s pace. As these processes ar e ass ociated with the ultimate heat sink for t he NuScale design, the panel concluded that this is an area of possible high unc ertainty.

32 5 NEUTRONICS PIRT

5.1 PIRT Panel Members

The panel consisted of e xperts w ith an understanding o f computational neutronics and thermal-hydraulics, and reactor a nd plant sy stems and their r esponse to various upsets. They m et for one day to carry out t he PIRT. Several of t he pa nel members had also p articipated in the thermal-hydraulic PIRT. The panel members and their a ffiliations were:

• Stephen Bajorek (Facilitator), NRC/RES

• David Diamond (Panel Chair), Brookhaven National Laboratory

• Istvan Frankl, NRC/RES

• Peter Griffith, Consultant

• Nathanael Hudson, NRC/RES

• Peter Yarsky, NRC/RES

Other NRC staff, from the Office of New Reactors (NRO) were also present to provide their support when needed. In addition, a telecom with Kent W elter from NuScale Power was held to help the panel understand how reactor startup would take place and to answer other general questions about the design.

5.2 Scenario The focus of this PIRT was on an event in which the neutronics was important. The design-basis reactivity accident for PW Rs has usually been the control rod ejection accident. However, in the specific case of NuScale it is expected that the frequency of occurrence of this event will be so small that it will no longer be a design-basis accident. Hence, the scenario considered by the panel was an inadvertent control rod bank withdrawal from hot zero power (HZP) condition at any time during the fuel cycle (expected to be a multi-batch, two-year cycle). This scenario would result in a power excursion mitigated first by Doppler reactivity feedback, and then by reactor trip. It was assumed that the reactor would be tripped by a high power trip setpoint (e.g.,

118% of nominal power), rather than by any low power or period trips that might be present, in order to make the power excursion as severe as possible for the purposes of the PIRT.

The initial conditions at HZP for a NuScale-type reactor are expected to be achieved by

Decay heat would be determined by how long the fuel had been operating and by taking into account at least one day of shutdown before startup. The pressure of the vessel at HZP would be at its nominal value for normal operation as the result of using pressurizer heaters. Coolant temperature would be determined by those heaters and any decay heat that might be present.

33 5.3 Figures of Merit It was decided that multiple figures of merit (FOMs) would be used to cover as many phenomena as possible. The FOMs were chosen by taking into account the acceptance criteria for this event. Hence, the FOMs were:

• Departure-from-nucleate-boiling ratio (DNBR)

• Reactor vessel pressure

• Fuel centerline temperature

• Pellet-clad interaction (PCI) / Pellet-clad mechanical interaction (PCMI)

For this event, the acceptance criteria are that DNBR remain above the 95/95 (i.e. 95%

probability at 95% confidence level) DNBR limit everywhere in the core, peak (pressurizer) pressure must be maintained below 110% of the design value, and the fuel centerline temperature must remain below the melting point.

The difference between PCI and PCMI is subtle, and it is sometimes difficult to differentiate the two types of failure. PCI is generally caused by stress-corrosion cracking due to fission product (iodine) embrittlement of the cladding, while PCMI is primarily a stress-driven failure. Generally, transient safety analysis considers a one percent uniform (elastic and inelastic) strain acceptance criterion to preclude PCI/PCMI failure.

Gaseous release and fuel thermal expansion are responsible for cladding strains at moderate to high exposures. Therefore, PCI or PCMI analyses of cladding strain for transients and accidents should take into consideration: fuel thermal expansion, gaseous fuel swelling models, and irradiated cladding properties

5.4 Results and Discussion

5.4.1 Phenomena The results of the PIRT panels deliberations are shown in Table 5.1. The list of phenomena was developed by the panel taking into account the important neutronic and fuel thermo-mechanical effects during this event as well as thermal-hydraulic phenomena. The thermal-hydraulic phenomena had been found to be relevant to the PIRT for a LOCA event. Note, however, that many fewer phenomena need to be taken into account in the neutronic PIRT relative to the thermal-hydraulic PIRT.

The phenomena are listed in the table according to the system or component in which they are found and according to what general behavior they affect. For this PIRT the three systems/components being considered are the core, reactor coolant system, and instrumentation and control system. Most of the phenomena occur in the core, as the bank withdrawal scenario is primarily a localized event. The focus groups are primarily in the core and include neutron kinetics (recognizing that some phenomena affect initial conditions rather than dynamic effects), fuel thermo-mechanical behavior, and subchannel thermal-hydraulics.

The pressurizer is also present as pressurizer pressure is one of the FOMs. Lastly, instrumentation and controls are part of the reactor protection system and have a direct effect on the neutronic response to the bank withdrawal scenario.

34 Because this w as a limited PIRT, there was no attempt to de fine in writing all of t he pheno mena listed, a pr actice which is commonly a part o f a PIRT. It w as assumed that the r eader w ould understand w hat w as meant by t he pheno menon. However, t here were several instances where clarification is useful and is given below.

• Shutdown bank speed is listed as a phenomenon; it is not simply something determined by design (as for example assumed for the control bank being withdrawn). The speed is primarily expected to be determined by a Technical Specification that specifies that banks must be able to be inserted a specified distance in a specified time. However, the speed beyond that point may be determined by bowing of guide tubes and this may be quite uncertain. It is, therefore, considered a phenomenon which must be considered and possibly taken into account in some way in the modeling of the event.

• Moderator density and temperature feedback are separated because they are accounted for differently in modeling the neutronics. The former refers to the effect of changes in density due to a change in temperature, while the latter refers to the change in scattering properties as a result of changes in moderator temperature.

• Fuel temperature feedback generally takes into account the Doppler effect and the change in scattering by oxygen within the pin. However, it generally does not also take into account the expansion/contraction of the pin with increasing/decreasing temperature.

• Assembly interaction and axial/radial reflector representation are two phenomena that are related. The small size of the integral PW R core necessitates consideration of neutronic phenomena that are generally of low or moderate importance for large commercial PW R cores. In particular, the small size of the reactor and high leakage fraction increases the importance of the axial and radial reflector, and in particular, the treatment of the reflector region.

Generally, PW R cores are analyzed using a two-step method whereby detailed transport calculations are performed for fuel assembly lattices to calculate homogenized, few group, nuclear cross-sections to be used in downstream nodal codes. The lattice calculations are generally performed presuming that the lattice may be depleted in an infinite array of like lattices to generate the exposure dependent nuclear parameters.

However, for high leakage cores, the inter-assembly coupling and coupling to the reflector may be of sufficient importance that this approach may not be fully applicable.

For example, for assemblies near the core periphery, which is a large fraction of the total number of fuel assemblies, the influence of the spectral thermalization of the return current in the reflector is likely to introduce strong gradients in the sub-assembly spectrum for the edge assemblies. Under such conditions, use of the standard two-step approach without modification may result in substantial bias in the calculated power distribution and kinetic behavior.

35

• Pin-to-pin burnup distribution and gap conductance and fuel conductivity are related to each other.

There is interplay between the important phenomena during the transient such that it would be difficult to demonstrate that in all cases any one particular phenomenon could be treated conservat ively. In particular, one can consider the influence of the average heat resistance of the fuel. The heat resistance here refers to a combination of the fuel pellet conductivity and gap conductance (both of these parameters are a function of the fuel pin exposure). The FOMs include the peak fuel centerline temperature as well as DNBR. Consider a bias in the heat resistance such that the event is analyzed with a higher, average fuel heat resistance. In this case, the transient increase in cladding heat flux is dampened by the higher heat resistance and this arguably will result in the prediction of a greater marg in to DNBR. However, the higher heat resistance will result in a greater propensity to hold heat up in the fuel, which will result in a more rapid fuel temperature increase, and thus a more rapid arrest of the power excursion by Doppler worth. This may or may not be conservative when then considered against the resultant increase in fuel temperature, which would reduce predicted margin to the fuel centerline melt FOM. Given this interplay, those parameters affecting the accurate prediction of the average fuel heat resistance were considered highly important.

The average fuel heat resistance, however, is a function of the fuel rod exposures (pin by-pin burnup distribution) and the instantaneous power distribution. The power distribution apportions the power across a variety of different heat resistances and thus influences the dynamics of transient cladding heat flux, average fuel temperature increase, and the overall transient event progression.

• Time in fuel cycle refers to the fact that feedback and bank worth change with burnup. It is necessary to do the analysis with the most conservative initial conditions.

• The pellet burnup distribution refers to the radial distribution of actinides and fission products and any burnable poison within the pellet. The fact that fission power peaks at the surface of the pellet has an effect on this distribution.

• Stored energy is a phenomenon that determines the initial fuel temperature.

• Cladding conductivity includes the effect of any oxide layer.

• Core pressure drop and loop pressure drop are important for calculating the flow rate through the core.

Table 5-1 PIRT Results for NuScale Neutronic Event System/ Focus Process / Phenomena Importance Knowledge Component Level CORE Neutron kinetics Withdrawn control rod bank worth H H CORE Neutron kinetics Shutdown bank worth H H 36 CORE Neutron kinetics Shutdown bank speed H M CORE Neutron kinetics Moderator density feedback H H CORE Neutron kinetics Moderator temperature feedback M H CORE Neutron kinetics Fuel temperature feedback H H CORE Neutron kinetics Assembly interaction (e.g., through ADFs) H M CORE Neutron kinetics Soluble boron reactivity feedback H H CORE Neutron kinetics Initial nodal power distribution H H CORE Neutron kinetics Transient nodal power distribution H H CORE Neutron kinetics Axial/radial reflector representation H M CORE Neutron kinetics Pin to pin power distribution H H CORE Neutron kinetics Time in fuel cycle H H CORE Neutron kinetics Pellet burnup distribution H M CORE Neutron kinetics Delayed neutrons H H CORE Neutron kinetics Decay heat M H CORE Neutron kinetics Xe/Sm Concentrations L H CORE Neutron kinetics Pellet/structure/coolant direct energy M H CORE Fuel Thermo-Pellet radial power distribution M M CORE Fuel Thermo-Core pin-by-pin burnup distribution H M CORE Fuel Thermo-Stored energy L H CORE Fuel Thermo-Fuel heat capacity H H CORE Fuel Thermo-Gap Conductance H M CORE Fuel Thermo-Fuel conductivity and density H M CORE Fuel Thermo-Cladding conductivity L H CORE Fuel Thermo-Cladding strain H H CORE Subchannel Subcooled boiling M H CORE Subchannel Natural Circulation L H CORE Subchannel Convection H H CORE Subchannel Core pressure drop M H CORE Subchannel CHF (correlations) H M RCS Loop pressure drop M H RCS Pressurizer Water expansion H H RCS Steam Generator Tubes Heat transfer L M I&C Reactor Protection Detector Response M L

Abbreviations: RCS - reactor coolant s ystem; ADF - assembly discontinuity factor; I&C - instrumentation and control; CHF - critical heat flux

5.4.2 Ranking The results show that 22 of the phenomena are of high importance (having significant or dominant influence on one or more FOMs), 8 are of medium importance, and only 5 are of low importance (having only a small influence on the FOMs). There are no phenomena that are of high importance for which the knowledge level is low (an H,L ranking). A low knowledge level 37 ranking indicates t hat the phenom enon is not w ell understood and t hat modeling t he phenomenon is cur rently eit her not pos sible or is possible only with lar ge uncertainty. If t here had been highly im portant pheno mena with low knowledg e level, t hese w ould have the highest priority w ith respect to as suring that computer cod es could model them pr operly.

The nex t level of int erest is with t wo other sco ring combinations. The phe nomena at this level have either a high im portance with a medium knowledge level (H, M ) or a medium i mportance with low k nowledge level (M,L). They ar e list ed b elow wit h their r anking (importance, knowledge level) and an explanation of each with respect to it s im portance and knowledge level follows.

• Shutdown bank speed (H,M)

• Assembly Interaction (H,M)

• Axial/radial reflector representation (H,M)

• Pellet burnup distribution (H,M)

• Core pin-by-pin burnup distribution (H,M)

• Gap conductance (H,M)

• Fuel conductivity and density (H,M)

• CHF correlations (H,M)

• Detector response (M,L)

The shutdown bank speed, along with its differential worth, determines the shutdown reactivity as a function of time and is directly responsible for terminating the bank withdrawal scenario.

Hence, bank speed is ranked as highly important. As mentioned above, the speed is for the most part expected to be determined by a Technical Specification requirement that specifies that banks must be able to be inserted a specified distance in a specified time. However, the speed beyond that point may be determined by bowing of guide tubes. This may be quite uncertain and is therefore ranked with medium knowledge level.

Assembly interaction is always an important consideration in modeling and becomes more uncertain when, as in the NuScale design, the core is small. Hence, it is ranked of high importance. The small reactor size introduces strong spatial and spectral flux gradients, resulting in a higher degree of uncertainty in the analysis of such reactors. Assembly interaction is usually accounted for with a synthesis approach in which the assembly is represented without any interaction with neighbors in order to obtain homogenized properties to represent the assembly within a core representation. The isolated representation also provides assembly discontinuity factors used to partially correct for the fact that the assembly is not part of an infinite array of identical assemblies. This synthesis approach adds to the uncertainty, and the phenomenon is ranked with a medium knowledge level.

The core of this reactor is expected to have only 37 fuel assemblies, with appearance similar to 17x17 PW R assemblies currently in use except for a reduced height. The reduced radial and axial dimensions mean that the leakage in this core will be high relative to that expected in a typical large PW R. Hence, axial and radial reflector representation will be highly important. For example, significant spectral gradients may be present between the core and reflector, and 38 given a large leak age fraction, the in fluence of this spectr al gradient o n o ther si gnificant parameters ( e.g. power distribution) should be ac counted for in the overall analysis method.

Standard approaches which generate nuclear par ameters for as sembly lat tices may not fully capture t he si gnificant spectral interaction between the cor e and reflector. Therefore, the panel ranked t he knowledge level here as only m edium. Although cur rent modeling of r eflectors for large P WRs has been sa tisfactory, there is little experience t o fully under stand how success ful these models will be when leak age is more si gnificant

The pellet bur nup distr ibution determines neutronic properties of t he fuel and the pow er distribution within the pellet and is r anked as hig hly im portant. The f uel management may eit her be for a sing le 24-month fuel cycle or for multi-batch reloads. If gadolinia is used in t he fuel, the burnup distribution becomes even more im portant because sel f-shielding will caus e the gadolinium to bu rn p referentially at t he surface of the f uel rod. There ar e challenges in modeling the bur nup distr ibution and indeed, some analysis in the past has ignored t he effec t altogether. These challeng es r educe the knowledge lev el to a medium r anking.

The core pin-by-pin burnup distribution is a factor in determining the t hermo-mechanical properties of the fuel. Henc e, it is important in determining fuel behavior a nd the w ay in which heat gets from t he f uel to t he coolant and is ranked as highly im portant. Frequently, only assembly-average bur nup is used t o det ermine pr operties and this introduces uncertainty int o the analysis. In par ticular the dynamic cladding h eat flux is sensitive to the combination of t he radial power dist ribution at the sub-assembly lev el as well as the distribution of heat resistance, which is determined according to pin-by-pin burnup. Pin-by-pin burnup, a nd the as sociated influence on fuel thermo-mechanical properties ( such as pellet conductivity and gap conductance) can be obtained via an online flux r econstruction methodology w ith a core depletion calculation. Nevertheless, the pin-by -pin burnup distribution is s till considered to hav e only a medium knowledge level.

Gap conductance and fuel conductivity and density de termine heat transfer t o the coolant.

Since the bank withdrawal event m ay lead to a power-c ooling m ismatch, the heat flux at the clad surface is a key pa rameter and t hese phenomena are ranked as hi ghly im portant. T he fuel thermal properties ar e u ncertain because of uncertainties in the bur nup o f a giv en fuel rod and because of uncertainties in the properties as a function of bur nup and henc e, t he knowledge level is considered medium.

Although DNBR was identified as a FOM, t he on set o f DNB is not expec ted as any sa fety analysis would need to d emonstrate marg in to the onset. T he calculat ed margin is based on the use of CHF c orrelations. CHF correlations are directly r elated to one o f the FOMs and, are therefore, ranked as hi ghly im portant. They ar e s omewhat uncer tain for s hort fuel assemblies as the m easurements that a re available have been done with geometries t ypical of large P WRs.

Hence, t he ranking of knowledge level is medium.

Detector r esponse h as been ranked as hav ing medium im portance. The s ignal for r eactor tr ip is expected to com e from n eutron det ectors w hich may be placed within the cont ainment. The detector response t o cha nges in c ore power and delays between signal receipt and the movement of control banks is uncertain becaus e of a lac k o f design in formation. Hence, 39 although t his phenom enon has been r anked as h aving a low k nowledge level, t his ranking is expected to be easily upg raded when more in formation is av ailable from the vendor.

6.0 SUMMAR Y AND CONCLUSIONS

This report do cuments two PIRT s for the NuScale-like design t hat is st ill preliminary. The PI RT panels made us e of information that was available as April 2011 and base d the r ankings on that information. Because t he design is preliminary a nd is expected t o change be fore submittal for Design Cer tification, it is important t hat the PIRTs in th is report also be v iewed as preliminary. A revision, based on final design in formation is vital to the development and review of thermal-hydraulic and neutronic codes for licens ing decisi ons.

6.1 Thermal-hydraulic PIRT

The accident scenario c onsidered for the t hermal-hydraulic PIRT is t he inadv ertent openin g o f a reactor recirculation v alve. An inadvertent control r od ban k w ithdrawal from ho t zero power (HZP) w as the scena rio consider ed for t he neutr onics PIRT. Both PIRTs identify phenom ena o f high im portance, those that must be modeled accurately or bounded in an appr opriate manner in a safety analy sis to ad equately assur e margin to regulatory lim its.

For t he thermal-hydraulic PIRT, the figures-of-merit ( FOM) w ere t he Critical Heat Flux (CHF),

mixture level in the vessel, and containment v essel pressure. The panel in general, did not believe it lik ely t hat limits on the departur e from n ucleate boiling ration ( DNBR) w ould be challenged except possibly in the early par t of the transient i f the c ore flow were t o s tagnate while at or near full power ope ration. The mixture lev el was selected as a FOM because the safety s ystems are designed to pr event c ore unco very and heatup. I n t he NuScale design, a conventional containment is replaced by a r elatively sm all high pressure co ntainment vessel.

During blow down and early in t he transient, high pressure w ithin the containment v essel is a safety conce rn and t herefore included as a F OM.

Of par ticular int erest are those phenomena t hat are highly r anked in importance, bu t are poorly understood as indicat ed by a low k nowledge level. For t he NuScale-like d esign, the major concerns appear to be with t he Cont ainment Heat Removal System (CHRS), w hich is where there w ere several phenomena in which the pane l identified as having high importance and less than satisfactory knowledge lev el. Phenomena in t his category include condensation heat transfer ( both with and without nonco ndensables), and m ixing in the cont ainment as it affects the liquid pool and t he gas space. There w ere a dditional concerns with chok ed flow from the vessel through t he reactor v ent v alves and the r eactor r ecirculation valves. These ar e mainly with the pr ediction w ith flow t o t he br eak and choked flow from the pr imary.

6.2 Neutronic PIRT

The neutr onic PIRT considered a scenario whereby a control ban k w as inadvertently continuously wit hdrawn from hot zero pow er cond itions. This would be expect ed to result in a power incr ease that would be offset initially by Do ppler feedback and t hen by r eactor tr ip.

40 Multiple figures-of-merit (FOMs) w ere used to c over as m any pheno mena as possible. The FOMs w ere c hosen by t aking in to accoun t what would be the accep tance c riteria for t his event.

Hence, the FOMs w ere:

• Departure-from-nucleate-boiling ratio (DNBR)

• Reactor vessel pressure

• Fuel centerline temperature

• Pellet-clad interaction (PCI) / Pellet-clad mechanical interaction (PCMI)

There were only 35 phenomena that the panel listed as having special significance in modeling this event. Most of these relate to either the neutronic initial conditions or neutron kinetics. In addition, there were phenomena that relate to fuel thermo-mechanical properties and subchannel thermal-hydraulics.

The results show that 22 of the phenomena are of high (H) importance (having significant or dominant influence on one or more FOMs), 8 are of medium (M) importance, and only 5 are of low (L) importance (having only a small influence on the FOMs). There are no phenomena that are of high importance for which the knowledge level is low (an H,L ranking). A low knowledge level ranking indicates that the phenomenon is not well understood and that modeling the phenomenon is currently either not possible or is possible only with large uncertainty. If there had been highly important phenomena with low knowledge level, these would have the highest priority with respect to assuring that computer codes could model them properly.

The next level of interest is with two other scoring combinations. The phenomena at this level have either a high importance with a medium knowledge level (H,M) or a medium importance with low knowledge level (M,L). They are listed below with their ranking (importance, knowledge level), and a detailed explanation of each phenomenon with respect to its importance and knowledge level is found in Section 5.

• Shutdown bank speed (H,M)

• Assembly Interaction (H,M)

• Axial/radial reflector representation (H,M)

• Pellet burnup distribution (H,M)

• Core pin-by-pin burnup distribution (H,M)

• Gap conductance (H,M)

• Fuel conductivity and density (H,M)

• CHF correlations (H,M)

• Detector response (M,L)

The deficiencies in knowledge level for these phenomena are summarized as follows:

• Shutdown bank speed, which partially defines the termination of the scenario, is determined, in addition to Technical Specifications, by bowing of guide tubes and this bowing is difficult to predict.

41

• Assembly interaction is always an important consideration in modeling and becomes more uncertain when, as in the NuScale design, the core is small.

• Axial and radial reflector representation is more important in the NuScale design than in large PW Rs and the neutronics methods that are suitable for the large PWRs are not expected to be sufficient for the iPW Rs.

• Pellet burnup distribution is important in determining the neutronic properties of the fuel and, because of self-shielding effects which are difficult to model, has a significant uncertainty.

• Core pin-by-pin burnup distribution is a factor in determining the thermo-mechanical properties of the fuel but although power reconstruction methods exist, modeling is still uncertain.

• Gap conductance and fuel conductivity and density determine heat transfer to the coolant, but these properties may be uncertain due to uncertainties in the burnup of a given fuel rod and because of uncertainties in the properties as a function of burnup.

CHF correlations are directly related to one of the FOMs, but they are somewhat uncertain for short fuel assemblies as measurements that are available have been done with geometries typical of large PW Rs. Detector response to changes in core power and delays between signal receipt and the movement of control banks is uncertain because of a lack of design information rather than because the phenomenon is difficult to model.

7.0 REFERENCES

1. NuScale Preliminary Loss-of-Coolant Accident (LOCA) Thermal-Hydraulic and Neutronics Phenomena Identification and Ranking Table, NuScale PIRT Licens ing Topical Report, NuScale Power, Inc. NR-TR0610-289, June 2010.

2. Vasavada, S., Zavisca, M., and Khatib-Rahbar, M., Modeling and Analy sis of Selected Loss of Cooling Accidents in an Integrated Light Water Reactor Using the TRACE Computer Code, ERI/NRC 210-203, Energy Research, Inc., Rockville, MD, May 2010.

3. Diamond, David, NuScale Reactor Overview, Presentation to NRC Staff, Rockville, MD, August 11, 2010.

42