ML20065S779

Best Estimate Analysis of Small Break LOCA in RESAR-3S PWR, Interim Rept

ML20065S779
Person / Time
Site:	05000545
Issue date:	09/30/1982
From:	Blakeley J, Cozzuol J EG&G, INC.
To:	Guttman J Office of Nuclear Reactor Regulation
References
CON-FIN-A-6468 EGG-NTAP-6032, NUDOCS 8211020015
Download: ML20065S779 (97)

Text

._

(

CS/Sfonce W p}/G GentthanlOf

.H icd EGG-NTAP-6032 September 1982 BEST ESTIMATE ANALYSIS OF A SMALL BREAK LOCA IN A RESAR-3S PRESSURIZED WATER REACTOR s?

J. E. Blakeley J. M. Cozzuol Idaho National Engineering Laboratory Operated by the U.S. Department of Energy y

amy cc

.-----_[ggy ummary %

T.,,

-j

, seemur summe ammer ums.w

.wt/* -g b

'U%dh i

',; _ w=~ ~t;4h.,we.. %

=r::n e..

G U

l L a, m y.y.>

T~

~

s

^ ~

+

2

~

This is an informal report intended for use as a preliminary or working document 0

Prepared for the l

U.S. NUCLEAR REGULATORY COMMISSION Under DOE Contract No. DE-AC07-76ID01570 Q

6 6 E 6 Maho FIN No. A6468 g

8211020015 820930 PDR RES PDR

hEGnG....

FORM EG4G 396 mu sa T

INTERIM REPORT Accession No.

Report No. EGG-NTAP-6032 Centract Program or Project

Title:

NRC Technical Assistance Program Division

, Subject of this Document: BEST ESTIMATE ANALYSIS OF A SMALL BREAK LOCA IN A RESAR-3S PRESSURIZED WATER REACTOR Type of Document: Technical Report Author (s):

J. E. Blakeley, J. M. Cozzuol Drt) of Document:

September 1982 O) t\\~.ponsible NRC Individual and NRC Office or Division: Jack Guttman, NRC-DSI This document was prepared primarily for preliminary or internal use. It has not received full review and approval. Since there may be substantive changes this document shou!d not be considered final.

[

EG&G Idaho. Inc.

Idaho Falls. Idaho 83415 Prepared for the U.S. Nuclear Regulatory Commission j

Washington, D.C.

I Under DOE Contract No. DE-AC07-761D01570 NRC FIN No.

A6468 O

i V

INTERIM REPORT l

1 ABSTRACT The RELAP5/M001 computer code was used to calculate the system response of a Westinghouse RESAR-3S plant to a limiting small cold leg break.

Best estimate assumptions were used in the calculation for the purpose of verifying and quantifying the conservatisms inherent within the analytical methodology required by 10 CFR 50.46 and Appendix K to 10 CFR 50.

Results of the analysis of the best estimate small break calculation indicate a continuous primary system depressurization with only brief periods of dryout of the top third of the core, and with peak cladding temperatures remaining less than steady state full power cladding temperatures. Comparisons of the RELAP5 results with results obtained from a Westinghouse evaluation model calculation for the limiting small break in a RESAR-3S plant show that the requirements for evaluation model small break calculations specified by 10 CFR 50.46 and Appendix K to 10 CFR 50 result in significant conservatisms in calculated system response relative to best estimate calculations.

FIN No. A6468--RESAR-3S "Most Probable" Best-Estimate LOCA Analyses in Support of FSAR Reviews O)

(

ii

SUMMARY

The RELAP5/M001 systems thermal-hydraulic code was used to calculate the response of a Westinghouse RESAR-3S pressurized water reactor (PWR) to a small cold leg break.

The Westinghouse RESAR-35 system is a large, 3411 MW thermal, 4-loop PWR.

The system was modeled with the RELAPS i

computer program with three intact loops combined into a single loop, and 2

with the fourth loop containing a break which represented a 0.08727 ft crack in the primary system piping where the accumulator line connects with the cold leg piping.

The chosen break size is the limiting small break size, as documented in the Westinghouse RESAR-3S Final Safety Analysis Report. Best estimate assumptions were used in the RELAP5 calculation for the purpose of providing a basis for verifying and quantifying the conservatisms inherent in the Westinghouse evaluation model (EM) RESAR-35 calculations as required by 10 CFR 50.46 and Appendix K to 10 CFR 50.

Analysis of the results of the RELAP5 calculation indicate that the p

system response to the small cold leg break is characterized by a continuous primary side depressurization, with only brief periods of dryout of the top third of the core. The periods of dryout occur just prior to blowout of the loop seal, and again after initiation of accumulator injection.

However, fuel rod cladding temperature increases are limited, and peak cladding temperatures during periods of dryout are substantially less than the steady state full power cladding temperature.

The comparisons of results from the RELAPS best estimate calculation with the results froc the Westinghouse avaluation model calculation illustrates the conservatisms inherent in the evaluation model analytical methodology.

In particular, the EM calculation shows a relatively prolonged period when the upper half of the core is uncovered, and during which cladding temperatures increased to a maximum of about 1760 F.

This compares to the best estimate calculation which shows only brief periods when the top of the core is uncovered and maximum fuel rod cladding temperature increases during these periods of only about 30 F.

Oi O

iii

T FOREWORD The RESAR-3S "Most Probable" Best Estimate LOCA Analyses in support of FSAR Reviews project was conducted under the direction of NRC's Division of Systems Integration, Roger Mattson, Director; Themis Speis, Assistant Director for Reactor Safety; Brian Sheron, Branch Chief for Reactor Systems; Norm Lauben, RSB Section Leader; and Jack Guttmann, Project Manager / Technical Monitor.

EG&G Idaho personnel involved in the project were Tom Charlton, Branch Manager, Reactor Simulation and Analysis Branch; Dr. Andy Peterson, Supervisor, PWR Systems Analysis; Tom Laats, Supervisor, Fuels Analysis and Data Bank; James Cozzuol, Jeb Blakeley, and Don Fletcher, Engineers; Joan Mosher, Glada Gatenby, and Brenda Hendrickson, Word Processing; Erma Jenkins and Sindi Crowton, Data Technicians.

This project was completed in September 1982 under FIN Number A6468 and NRC B&R Number 20 19 40 42 3.

I i

l l

9

CONTENTS ABSTRACT..............................................................

11

SUMMARY

iii FO R EWO R D..............................................................

IV T

1.

INTRODUCTION...................'..................................

I 2.

PLANT AND POSTULATED ACCIDENT DESCRIPTION........................

3 3.

COMPUTER CODE AND MODEL DESCRIPTION..............................

4 4.

ASSUMPTIONS FOR BEST ESTIMATE AND EM CALCULATIONS................

9 5.

INITIAL AND BOUPDARY CONDITIONS..................................

Il 5.1 Initial Conditions.........................................

11 5.2 Boundary Conditions 14 6.

ANALYSIS RESULTS.................................................

19 6.1 RELAPS Calculation--General System Behavior................

19 6.1.1 System Pressure Respotse...........................

21 6.1.2 System Mass Inventory / Distribution.................

24 6.1.3 Core Thermal / Hydraulic Response....................

33 6.1.4 B re a k Fl ow Re s po n se................................

45 6.1.5 Similarity of RELAPS Calculation to Experimental Results............................................

49 6.2 RELAP5/ Westinghouse Calculation Comparisons................

52 7.

CONCLUSIONS.....................................................

62 8.

REFERENCES.......................................................

63 APPENDIX A--RELAP5 UPDATES............................................

A-1 APPENDIX B--QUALITY ASSURANCE PROCEDURE FOR DEVELOPMENT i

0F THE RELAPS RESAR-3S SMALL BREAK MODEL.........................

B-1 t

APPENDIX C--FUEL STORED ENERGY CALCULATION............................

C-1 i

FIGURES 1.

RELAPS nodalization diagram for the RESAR-3S small break 5

s calculation......................................................

l v

i

2.

Comparisons of actual and desired radial temperature profiles for the hot and average fuel pins in the RELAPS model...........

13 3.

Comparison of normalized core power for RELAP5 and EM small break calculations.........................................

15 4.

Comparison of pumped ECC flow rates for the RELAP5 and EM calculations..................................................

18 f

5.

Comparison of pump speeds for RELAP5 and EM calculations.........

18 6.

Upper plenum pressure from the RELAPS calculation...............

22 2

7.

Upper plenum fluid temperature and corresponding saturation temperature from the RELAP5 calculation (0 to 100 s).............

22 8.

Steam generator inlet plenum fluid temperature and corresponding saturation temperature from the RELAP5 c a l c u l a t i o n ( 0 t o 10 0 s )........................................

23 9.

Total primary system mass inventory from the RELAPS calculation......................................

25

10. Collapsed liquid levels in the upflow and downflow sides of the intact loop steam generator tubes from the RELAPS 27 calculation (0 to 600 s)........................................

11. Collapsed liquid levels in the upflow and downflow sides of the broken loop steam generator tubes from the RELAPS 27 calculation (0 to 600 s)........................................

12. Collapsed liquid level in the vessel upper plenum for the 28 RELAP 5 ca l c ul a ti on ( 0 to 600 s ).................................

13. Void fraction in the intact loop hot leg from the RELAP5 28 calculation (0 to 500 s).......................................

14. Void fraction in the broken loop hot leg from the RELAPS 29 calculation (0 to 500 s).......................................

15. Downcomer and core collaosed liquid levels relative to the bottom of the core barrel from the RELAPS calculation (0 to 750 s).......

29

16. Collapsed liquid levels in the upflow and downflow sides of the intact loop pump suction from the RELAP5 calculation (0 to 750 s) 31

17. Collapsed liquid levels in the upflow and downflow sides of the broken loop pump suction from the RELAP5 calculation (0 to 750 s) 31

18. Collapsed liquid levels in the downcomer and core relative to the bottom of the core barrel from the RELAPS calculation........

32 vi 1

m 19.

Collapsed liquid levels in the upflow and downflow sides of the broken loop pump suction from the RELAPS calculation......

34 20.

Cladding surface temperature for the average power fuel pins at the 10 to 12 foot elevation from the RELAPS calculation......................................................

35 21.

Cladding surface temperature for the average power fuel pins at the 8 to 10 foot elevation from the RELAP5 calculation......................................................

35

22. Cladding surface temperature for the high power fuel pin at the 10 to 12 foot elevation from the RELAPS calculation......................................................

36 23.

Cladding surface temperature for the high power fuel pin at the 8 to 10 foot elevation from the RELAPS calculation......................................................

36 24.

Cladding surface temperature for the high power fuel pin at the 6 to 8 foot elevation from the RELAP5 calculation......................................................

37

25. Collapsed liquid level in the core relative to the bottom of the active fuel from the RELAP5 calculation........

37 T

26. Void fraction in the core for the six volumes y

encompassing the active fuel region from the RELAP5 calculation (40 to 260 s)........................................

38 27.

Comparison of the collapsed liquid level in the core region to the mass flow into the bottom of the core from the RELAP5 calculation (0 to 500 s).........................

40

28. Comparison of the cladding temperature at the top of the high power fuel pin to the void fraction at the top of the core from the RELAPS calculation (400 to 550 s)..............

41 29.

Comparison of core collapsed liquid level to the collapsed liquid levels in the broken and intact loop pump suctions from the RELAPS calculation (1100 to 1400 s).....................

41 30.

Comparison of the cladding temperature at the top of the high power fuel pin to the void fraction at the top of the core f rom the RELAPS calculation (1100 to 1400 s)............

43 31.

Comparison of core collapsed liquid level to the collapsed liquid levels in the broken and intact loop pump suctions from the RELAP5 calculation (1400 to 1900 s).....................

43 32.

Comparison of the cladding temperature at the top of the high power fuel pin to the void fraction at the top of p

the core f rom the RELAPS calculation (1400 to 1900 s)...........

44 U

vii

33. Mass flow out the break from the RELAPS calculation......

46 34.

Mass flow in the broken loop cold leg from the RELAP5 calculation (100 to 500 s)......................................

47

35. Comparison of break mass flow to the temperature of the fluid in the break volume from the RELAPS calculation....

47

36. Comparison of the break mass flow to the break volume void fraction from the RELAPS calculation (490 to 590 s).................................

48 37.

Comparison of the break mass ficw to the temperature of the fluid flowing to the break from the RELAPS calculation (860 to 1080 s)..................................................

48

38. Comparison of the break flow to the flow on the vessel side of the break volume from the RELAP5 calculation (860 to 1080 s)......

50 39.

Mass flow on the vessel side of the break volume from the RELAP5 calculation (1000 to 1900 s).........................

50 40.

Comparison of the break volume fluid temperature to the corresponding saturation temperature from the RELAPS calculation (800 to 1800 s) 51 41.

Comparison of pressurizer pressures for the RELAPS and EM calculations.................................................

54 42.

Comparison of the total amount of ECC injected into the primary system for the RELAPS and EM calculations...............

56 43.

Comparison of the break flow rates for the RELAP5 and EM calculations....................

56 44.

Comparison of primary system total mass for the RELAP5 and EM calculations..................

58 45.

Comparison of core mixture levels for the RELAPS and EM calculations 58 46.

Comparison of high power fuel pin cladding temperatures for the RELAP5 and EM calculations at the 10 to 12 foot elevation....

60 47.

Comparison of the high power fuel pin cladding temperatures for the RELAPS and EM calculations at the 8 to 10 foot elevation..........

60 48.

Comparison of the heat transfer coefficient at the 10 to 12 foot elevation on the high power fuel pin for the RELAP5 and EM calculations...................

61 viii

i

49. Comparison of the hot spot fluid temperatures for the RELAPS and EM calculations..............................................

61 TABLES

{

r 1.

Comparison of assumptions used in the RELAP5 best estimate calculation and the evaluation model calculation.................

10 2.

Initial conditions...............................................

12 3.

Comparison of boundary condition setpoints for RELAPS and EM calculations..................................................

16 4.

Rated pump parameters for RESAR-3S pumps.........................

17 5.

Sequence of events for RELAPS small break calculation............

20 f

l 1

O Y

iX

N BEST ESTIMATE ANALYSIS OF A SMALL BREAK LOCA IN A RESAR-35 PRESSURIZED WATER REACTOR 1.

INTRODUCTION Current requirements for determining the acceptability of emergency core cooling (ECC) systems in light water reactors during postulated loss-of-coolant accidents (LOCA) incorporate conservatisms developed to bound the uncertainties in the analysis of the phenomena that occur. These conservatisms are codified in 10 CFR 50.46 and Appendix K to 10 CFR 50.

Calculations have been performed with sophisticated analytical computer programs as part of an effort to verify and quantify the conservatisms inherent in the requirements of 10 CFR 50.

This report documents the results of a postulated 4 inch cold leg break LOCA analysis of a Westinghouse pressurized water reactor (PWR)

~

assuming most probable operating parameters and utilizing the best available analytical methodology. The PWR design selected was the d

Westinghouse RESAR-3S nuclear steam supply system (NSSS). Drawings and plant data supplied by Westinghouse were used to construct a computer model representing the system geometry and operating conditions. The computer l

code selected for the analysis was the RELAPS/ MODI computer code developed at the Idaho National Engineering Laboratory for small break LOCAs.

Section 2 of this report contains a description of the plant and the small break LOCA calculation that was performed. Details of the RELAP5/ MOD 1 nodalization used to represent the RESAR-3S system are given in Section 3 along with the code options selected for the calculation.

Section 4 highlights the differences in the assumptions used for the RELAPS most probable best estimate calculation and those used in the Westinghouse evaluation model (EM) calculation. Section 5 compares the boundary and initial conditions used in the RELAP5 calculation to those used in the Westinghouse EM calculation. The results of the RELAP5 calculation are v

1

presented in Section 6.

Qualitative and quantitative comparisons to the EM analysis performed by Westinghouse under the guidelines of 10 CFR 50 are also presented in Section 6.

Section 7 details the conclusions reached concerning the RELAPS analysis and the comparisons made to the licensing analysis.

.l l

O 1

)

9 O

2 l

l

O 2.

PLANT AND POSTULATED ACCIDENT DESCRIPTION The RESAR-35 NSSS is a Westinghouse pressurized water reactor consisting of a pressure vessel and 4 separate coolant loops. The vessel contains 193 nuclear fuel assemblies, each containing 264 fuel rods in a 17 x 17 array. There are 61 full length control rods for reactor control.

Each coolant loop consists of a vertical, single-stage, centrifugal pump; a Type F steam generator with Inconel tubes; auxiliary feedwater systems; a steam dump system: ECC systems; and the connecting piping. A pressurizer and associated surge line is attached to one loop.

The postulated accident is a 4 inch break at the location where the accumulator line is welded to the cold leg in a loop without the pressurizer. This corresponds to the most limiting small break in terms of highest peak cladding temperature, as presented in the RESAR-35 Reference Safety Analysis Report.2 Depressurization of the primary system causes a reactor trip, steam dump actuation, auxiliary feedwater flow, and ECC 4

system actuation.

Loss of primary coolant through the break eventually caused approximately 75% voiding of the system, but no significant or prolonged heatup was calculated in the core. The calculation was terminated at 1870 s after 110 s of RHR injection during which system mass inventory increased rapidly.

l I

l

3.

COMPUTER CODE AND MODEL DESCRIPTION The computer code chosen for the small break calculation was RELAP5/M001 Cycle 18 with updates. RELAP5/M001 is an advanced best estimate computer code developed at INEL for small break calculations. The code has the capability for modeling all systems in an NSSS required for small break calculations as well as an extensive control system package.

Features unique to RELAPS include subcooled and two phase nonequilibrium and nonhomogeneous choked flow models, horizontal stratified flow and stratified choked flow models, noncondensibles in the vapor phase, and a two phase mechanistic abrupt area change model.

RELAPS 4

calculations have been compared to other code calculations and to experimental data.4-7 Conclusions of these assessment efforts show that RELAPS is an appropriate computer code for calculating the trends of the phenomena occurring in small break transients.

The model used in this calculation consists of 150 volumes, 159 junctions, and 181 heat structures. Figure 1 shows a nodalization diagram of the model.

Three of the four loops were lumped together as the intact loop and the broken loop was modeled separately. The length of corresponding components in each loop was the same, wit':

Ne intact loop components having three times the volume and flow area of tne oroken loop components. The pressurizer surge line connects into the intact loop hot leg. Charging pumps, safety injection pumps, accumulators, and residual heat removal (RHR) pumps were modeled in each loop.

The charging flow was injected directly into the cold leg. The safety injection and RHR flows were injected into the accumulator line which was connected to the cold leg upstream of the charging line connection. The break was located in the single loop at the same location as the accumulator line connection with the intent of modeling a crack in the weld of the accumulator line to cold 2

leg connection. The flow area of the break was 0.08727 ft ; the equivalent of a 4 inch diameter circular hole.

O 4

M e,s O

J{ w a r a

j g ll i

! fl,,.~

R 4

ii

@;g @,5 @

ge py-i Or, b.

t 1 p 4. _,.-

j i 1 mi,

g g=g-m s

Z ~l ~ -

rw a

s b

D l@', (5),

k

(=:

i La

_~ -

e 3

@.f 3

5 4

C l

l

=

g G3 g

L MMMhWN i! 6%wusswwwwsst j 1 l g.i L

e

'.-. l t

j ti p l L ) d ti

}, Iy mwwwMM h g

E

. i

]

.s s: %%swesse

@ * @ h g: s :

e

+ _! @

E gI I G) i c

3 E muwww sw iH;

'j a

hyf*lFj' l i @l i i

2 sN g awww I$ E nwwwwwww E

-S 5 e

p e;

I

.a i

%5 d%

Q~'

eu 4,

WI

-.ama.

i l

y; e>gw, i'y,".,-

3 s

= =m

!!'ly,

,i 3

I y-

@O i

!.!:,l1 I

~

i.

S

.c'e W

!k 3.!..

(k m

a w ;

i:

- I

'I l

jl !'

l "i==s

$I$ $

a: -

[.c 5

The steam generator primary and secondary sides were modeled including internals metal mass and heat transfer area, main and auxiliary feedwater systems, steamline, relief valves, turbine stop valves, main steam isolation valves, and steam dump system. Since the RELAPS separator component is an ideal separator (it allows only dry steam out the top) the steam separators and driers were lumped into one component. The steam dump system had the capability of relieving 40*4 of full load steam line flow.

The atmospheric dump system and one of the spring loaded safety relief valves was modeled. More r: lief valves would have been added had they been needed for the calculation. The relief valves flow areas were sized to give the correct flow at the valve actuation pressure.

The reactor coolant pumps were modeled using PUMP components.

Homologous curves, two phase difference curves, and two phase multiplier tables for head ard torque for Westinghouse PWR pumps were used for all input requirements except the two phase difference curves for the energy dissipation region of the head and torque curves. Semiscale two phase difference data was the only information available to use for this input.

The pressurizer tank and surge line were modeled with PIPE components. Eight nodes were used in the pressurizer tank so the draining could be followed. The heaters and cooling spray were not modeled since they were not needed to reach steady state and would not be utilized in the transient. The power operated relief valve was not modeled since it would not be challenged in this transient.

The vessel model includes a downcomer, lower plenum, core region, core bypass, upper plenum, and upper head with an " inverted top hat."

The model included the leakage path from the vessel inlet nozzles to the upper plenum, and the upper head spray nozzle flow path. The core region had six volumes so the licuid level could be tracked and the hydraulic conditions could be accurately calculated.

Heat structures were used to model the stored energy and heat transfer surfaces of the primary system loop piping; steam generator walls, O

6

b internals, and tubes; and vessel walls, internals, and fuel pins. One fuel pin was modeled as a hot pin, but there was no separate fluid volume for that pin. Heat losses to the ambient were not modeled. Heat transfer coefficients were miculated by the code for all heat structures except one in the vessel tnat was used to simulate gamma heating. This heat structure was modeled as a thin piece of steel with a large surface area and a high heat transfer coefficient so all the energy generated within it would be immediately transferred to the fluid in the vessel.

Other user relected code options include the following. Nitrogen was the noncondensible gas in the accumulator. Wall friction and unequal temperature options were implemented for all primary and secondary volumes except pumps and the steamline beyond the main steam isolation valves (MSIV). These options are not allowed for pumps. The main steamlines beyond the MISVs were not regions critical to the calculation so the options were chosen to minimize problems in achieving steady state conditions. With a few exceptions to be discussed later, the following g

modeling criteria were applied at the junctions. As recommended by the RELAPS development group, choking was allowed at all junctions except at the separators; the geometry determined whether a junction had a smooth or j

abrupt area change; and the full inertia treatment option was selected for all junctions. Liquid and vapor could have unequal velocities except at the separator inlets and the accumulator line to cold leg connections.

These selections gave more realistic void distributions in the steam generators and prevented problems that could occur by injecting cold water into a two phase system.

At 1522.2 s, a restart was performed which changed two upper nead bypass flow junctions to mitigate a mass error problem. As suggested by the RELAP5 code development group, the junctions from the downcomer to the upper head and from the upper head to the guide tubes were given increased flow areas and loss coefficients. These changes wculd provide the same steady state flow through the junctions, but allow the code to calculate two phase flow with less mass error. The choking and two velocity options were turned off for these junctions at this time.

If this technique had been used for the entire calculation, the system mass error would have been l

7 M

smaller but the system depressurization and core thermal hydraulics would not have been significantly altered.

Another restart was performed at 1722.2 s.

The flow from a time dependent volume to the volume that feeds the pumped ECC trains was choking when RHR pumps came on.

Since the area of this junction was arbitrarily chosen, it was increased to prevent the unrealistic choking problem. With these changes, the calculation ran smoothly to completion at 1870 s.

The updates used in this calculation include the following. One update fixes an error so tha; all restart variables are defined. Two updates cause the velocity term and mass terms in the momentum flux equation to be donored. Another update allowed mass error and total system mass to be used as minor edit and plot variables. At the 1722.2 s restart, an update was added to reduce the mass error. All updates were recommended by the code development group and are listed in Appendix A.

O O

8

4.

ASSUMPTIONS FOR BEST ESTIMATE AND EM CALCULATIONS The assumptions for the initial and boundary conditions in the RELAPS calculation are the result of getting best estimate conditions of a plant at typical operating conditions. Table 1 compares the assumptions used in the RELAPS calculation to the assumptions used in the EM calculation. The conservatisms in the EM calculation are apparent in the lower ECC and auxiliary feedwater flow rates, longer delay times, higher water temperatures, and higher core power.

i l

l l

TABLE 1.

CCMPARISON OF ASSUMPTIONS USED IN THE RELAPS BEST ESTIMATE CALCULATION AND THE EVALUATION MODEL CALCULATION RELAPS EM ECC All trains working (2 safety One train working (one safety injection pumps, 2 charging injection pump, one charging pump) pumps, 2 RHR pumps) 4 accumulators injecting into the 3 accumulators injecting into the primary system primary system Pumped ECC is based on design Pumped ECC flow degraded 5% from head / flow curve with 10% of design head design head added uniformly over curve All ECC water at 90 F Accumulator water at 120 F Pumped ECC water at 100 F Safety injection delay = 10 s iafety injection delay = 25 s Steam Generator Secondary Side Steam dump system tctuated at No steam dump system reactor trip 2 motor driven auxiliary feed-1 motor driven auxiliary feedwater water pumps, 500 gpm each, pump, 470 gpm, 60 s delay, 30 s delay, 90 F water 120 water 1 turbine driven auxiliary feed-No turbine driven auxiliary feed-water pump, 1000 gpm, 60 s water pump delay, 90 F water Core Power Decay 76% of ANS + 20*.' power decay ANS + 20% power decay curve curve 2.2 s rod drop time 3.4 s rod drop time Total core power of 3411 Mw Total core power of 3206.88 Mw Total peaking factor of 1.678 Total peaking factor of 2.32 Reactor Coolant Pumps Tripped at 1300 psia Tripped at 0 s O

10

O 5.

INITIAL AND BOUNDARY CONDITIONS U

This section compares RELAPS initial conditions to the desired initial conditions for a RESAR-35 plant.

Differences between the boundary conditions of the RELAPS best estimate calculation and the Westinghouse EM calculation are discussed.

5.1 Initial Conditions When the model was assembled and quality assured (see Appendix B) the process of arriving at a steady state condition was initiated. The sum of the elevation changes of all flow loops was checked to ensure that all loops close. Estimates were made of initial fluid conditions throughout the system and these were applied to each volume.

Estimates of mass flows were input for each junction. Power was applied to the core, the pumps were turned on to rated conditions, a time dependent volume was connected to the top of the pressurizer to force a constant pressure boundary condition on the primary side, main steam valves were opened, and feedwater was pumped into the steam generators. Feedwater flow was set equal to steam flow to maintain the desired secondary side inventory. A pump speed controller adjusted the pump speed to get the desired mass flow rate through the core. A steam valve controller adjusted the steam flow rate to get the correct cold leg temperature. The secondary side pressure had to be lowered by 61 psi below specified conditions and the recirculation flow ratio lowered to 3.1 from 3.7 to obtain a high enough heat removal from the primary to get the desired cold leg temperature.

Pressure drops around the system were compared with desired values and loss coefficients were changed to obtain desired pressure drops.

Loss coefficients in the vessel bypass paths were adjusted to give correct flows in these areas.

<~

n i

n

The maximum stored energy in the fuel was determined from FRAPCON-2 runs (Appendix C) and a radial temperature profile was generated for the average and hot fuel pins.

The radial temperature profile in the RELAP5 fuel pins was forced to match the FRAPCON-2 temperature profile. This procedure gave the correct stored energy in the fuel since the fuel properties and the fuel geometry were the same for both codes. The temperature profile in the RELAPS model was changed by changing the gap conductance and fuel pellet radial power profile in an interative procedure with constant fluid conditions.

Figure 2 comnares the initial temperature profile used in RELAPS to the desired (FRAPCON-2) temperature profile for the hot pin and average pins.

Steady state was reached when it was determined that selected pressures, temperatures, and flow rates were at their desired values and were not changing in time.

Table 2 shows important system parameters at steady state conditions compared to des' red conditions. After steady state was reached, the pump speed and sterm valve position controllers were removed and those parameters were hold constant until conditions of the system during the transient dictated a change.

TABLE 2.

INITIAL CONDITIONS RELAPS RESAR-3S Parameter Value Value Cold leg temperature ( F) 557.7 557.6 Hot leg temperature ( F) 617.7 617.8 Pressurizer pressure (psia) 2255.

2250.

Primary coolant flow (1bm/s) 39091.

39111.

Total core power (MW) 3411.

3411.

Secondary side pressure (psia) 935.

1000.

Feedwater temperature ( F) 440.

440.

Steam mass flow rate (1bm/s) 1047.

1051.

Vessel AP (psi) 45.48 45.41 Steam generator AP (psi) 31.4 33.2 O

12

~_

... -. __. _ _.. -. -. ~..

J

?

l 1

1

(

l I

28 "

ACTUAL TEMPERATURE

[2250 DESIRED TEMPERATURE v

L. 2000 l

s s

?, f750 t

ca.

E 8,

1500 Avenge sin T

8 1250

\\

-e d m@

~

1 750 0.00 0.02 0.04 0.06 0.05 0.10 0.12 0.14 0.16 0.18 Rcdlus (in)

Figure 2.

Comparisons of actual and desired radial temperature proffles for the hot cad overage f uel pins in the RELAPS model 13

5.2 Boundary Conditions This section discusses the boundary conditions applied in the RELAP5 calculation, including fuel rod power, ECC injection parameters, and steam generator secondary side control. The setpoints and delays for the

~

application of the boundary conditions are presented in Table 3.

These boundary conditions are compared to the EM calculation boundary conditions where appropriate.

Heat losses to the environment were not modeled since they would be a negligibly small fraction of the system power during this calculation. The containment pressure was held constant at 14.7 psia throughout the calculation. The subcooled and two phase break flow multipliers were each set to 0.84 as recommended by the code development group and the LOFT program code users.

In both the RELAP5 and EM calculations, after the pressurizer pressure had dropped to 1860 psia there was a 2.0 s signal processing delay before a reactor trip signal was generated.

From the time of the trip signal until the control rods were fully inserted was 2.2 s in the RELAPS calculation and 3.4 s in the EM calculation. Reactor power did not begin to decay until the rods were fully inserted. A plot of the normalized power decay curve for both calculations is shown in Figure 3.

The reactor trip signal initiated the closing of the turbine stoo valves which had a 0.5 s closing time. A steam dump system that directs up to 40% of the full power steamline flow directly to the condenser was enabled after the reactor trip signal was generated. The flow rate of the steam dump system was based on an average temperature that had been processed in a lead-lag controller.

The lead time constant was 10 s and the lag time constant was 5 s.

The EM calculation did not model a steam dump system.

O 14

- _.. ~ ~

I I

i 1

I 4

1 J,

l i

t I

t 1d C

3

~

RELAPS

~*

EW l

u.

1(f o

ca.

3

,s e

i N=

{

.l l

O e

E 10" l

{

9.

\\

%.'~...----.-.-.---....._._._.___.]

in-.

0.0 500.0 1000.0 1500.0 2000.0 Time (s)

Figure 3.

Comparison of normalized core power f or RELAPS and EW small break calculations.

15

TABLE 3.

CCMPARISON OF BOUNDARY CCNDITION SETPOINTS FOR RELAPS AND EM CALCULATIONS a

RELAP5 Setpoint EM Setpoint Event (Time Delay)

(Time Delay)

Open break time = 0.0 s time = 0.0 s Pressurizer low pressure p = 1860 psia p = 1860 psia Close turbine stop valves p = 1860 psia p = 1860 psia (2.0 s)

(2.0 s)

Begin power decay p = 1860 psia p = 1860 psia (4.2 s)

(5.4 s)

D Pressurizer low-low pressure p = 1760 psia p = 1760 psia b

Safety injection ("S")

p = 1760 psia p = 1760 psia signal (2.0 s)

(2.0 s)

Charging flow initiated "S" signal "S" signal (25.0 s)

Close feedwater valve "S" signal "S" signal (2.0 s)

Safety injection flow "S" signal "S" signal initiated (10.0 s)

(25.0 s)

Motor driven auxiliary "S" signal "S" signal feedwater flow started (30.0 s)

(60.0 s)

Turbine driven auxiliary "S" signal N/A feedwater flow started (50.0 s)

Reactor coolant pumps p = 1300. psia time = 0.0 s tripped off Accumulator pressure 600 psia 600 psia RHR flow initiated p = 215 psia N/A a.

Time delay after setpoint is shown in parenthesis below the setpoint parameter.

b.

Values obtained from Westinghouse data package for RESAR-3S plants.

O 16

A safety injection ("S") signal was generated 2.0 s after the pressurizer pressure reached 1760 psia.

The main feedwater valves started closing 2.0 s after the "S" signal.

In the RELAP5 calculation, the motor driven auxiliary feedwater pumps came on 30.0 s after the "S" signal, and the turbine driven auxiliary feedwater pumps came on 50.0 s after the "S" signal.

In the EM calculation, the motor driven pump came on 60.0 s after the "S" signal. The auxiliary feedwater pumps were controlled to cover the steam generator tubes and keep them covered throughout the calculation without completely filling the generators.

In the RELAP5 calculation, the "S" signal initiated the pumped ECC injection with no delay for charging pump injection and a 10 s delay for safety injection pump actuation.

In the EM calculation, no pumped ECC was injected until 25 s after the "S" signal. The accumulators were initiated at 600 psia and the ficw rate was controlled by system pressure at the injection point for both calculations.

RHR was ramped on from 200 to 215 psia in the RELAP5 calculation and was not considered in the EM calculation.

Plots of pumped ECC flow rates vs. time are shown in Figure 4.

In the RELAPS calculatic,n, the reactor coolant oumps were tripped off at 1300 psia and allowed to coastdown.

Reverse rotation was not allowed, but they were allowed to spin in a turbine mode.

Reactor coolant pump speed is compared in Figure 5 for the RELAPS and EM calculations.

In the EM calculation, the pumps were tripped off at the initiation of the accident. Rated pump parameters used in the RELAPS calculation are shown in Table 4.

TABLE 4.

RATED PUMP PARAMETERS FOR RESAR-3S PUMPS Speed (rpm) 1186.

Flow (gpm) 94400.

Head (ft) 304.

Torque (1bf-ft) 28015.

~

2 Moment of inertia (lb-ft )

95000.

3 Density (lbm/ft )

47.18

,0 17

O 1200 RELAP5

-- EM 1000 l} -

m N

E a00

B i

v y

600 O

C n

400 a

02 200 -

,_______,,,...~ ~ - ~ ~ ~ - - - ~ ~~~ ~~~~~~~

0 C

500 1000 1500 2000 Tim. (3)

Figure 4.

Comocrison of pumped ECO flow rates for the REL APS and EW calculottons.

1C i

EM PUWPS 1

C RELAP5 INTACT LOOP PU;dP 7

O RELAPS BROKEN LCCP DUMP o

0.8 ec.

\\

i 3'

c.E M 3

I c.

I D

o i

N 0.4

=

O t

E w

I C

02 -\\

\\

% M L. '_. :

~

A 0

0 500 1000 1500 2000 Time (s)

Figure 5.

Comocrisen of rormalized pwmo speeds for RELAPS and EW calculatiovis.

O 18

(~N 6.

ANALYSIS RESULTS U

The following two sections present the results of the analysis of the RELAP5 small break (i.e., 4 inch cold leg break) calculation for a RESAR-3S plant. The first section describes the general system behavior as calculated by the RELAP5/M001 program, and includes a detailed analysis of the factors which influenced the response. A brief summary of how the RELAPS results compare with experimental data is also included in this section. The second section presents comparisons of results obtained from the RELAPS calculation with the results obtained from the limiting small break calculation performed by Westinghouse for the RESAR-3S document.

Comparisons are included in this section for each of the graphical outputs presented in the RESAR-3S document. These comparisons quantify the influence of the conservatisms imposed by 10 CFR 50.46 and Appendix K to 10 CFR 50 on the Westinghouse RESAR-3S calculation.

6.1 RELAPS Calculation--General System Behavior Table 5 presents a sequence of events highlighting important u

operations and thermal-hydraulic events which occur during a 4 inch cold leg break LOCA, as calculated by the RELAP5/M001 program. The system response to the small cold leg break is characterized by a continuous primary side depressurization, with only brief periods of dryout of the top third of the core. The periods of dryout occur just prior to blowout of the loop seal, and again after the initiation of accumulator flow.

[

Following the initiation of the transient, voiding of the primary system progresses from the upper elevations downward. The continued loss b

of fluid from the system via the break, in conjunction with the formation

(

'of liquid seals in the; pump suction piping, causes a gradual depression of j

the r:.txture level in the vessel below the top of the core. The

~ corresponding dryout of the, upper portion of the fuel rods results in a imited rod, temperature increase (temperatures remain less than steady state full power temperatures).

However, blevout of the loop seal shortly

\\

after'the core begins to uncover leads to a recovering of the core, and the o,

l ',

19

TABLE 5.

SEQUENCE OF EVENTS FOR RELAP5 SMALL BREAK CALCULATION Time Event (s)

Blowdown initiated 0.0

~

Pressurizer low pressure trip setpoint 15.2 (1860 psia) reached Intact / broken loop turbine stop valves 15.3 begin to close Steam dump system actuated 15.3 Core power decay initiated 19.5 Pressurizer safety injection signal 20.4 setpoint (1760 psia) reached Charging pump flow initiated 22.5 Intact / broken loop main feedwater 24.5 valves begin to close High pressure injection pump flow 32.5 initiated Intact / broken loop pump coastdown 34.9 initiated Pressurizer emptied 40 Upper plenum / hot leg fluid saturates 38-42 Auxiliary feedwater flow initiated 52.6 Upper head begins to drain 200 Broken loop pump suction legs clear 520 and break uncovers Intact / broken loop accumulator injec-858 tion initiated

~

RHR pumped injection initiated 1755 Calculation terminated 1870 0

20

mixture level remains above the top of the core until after accumulator V

injection begins. The accumulation of liquid in the pump suction piping after accumulator injection starts, results in two futher periods (again of brief duration) in which the mixture level in the vessel is depressed below the top of the core. Again, however, rod temperature increases are limited, and peak cladding surface temperatures remain well below the full power steady state temperatures. Once injection from the low head residual heat removal pumps begins, the rate of liquid addition to the primary system (via all active components of the ECC system) is sufficient to cause a rapid increase of the primary system liquid inventory, thus assuring adequate and continued cooling of the core.

The system thermal-hydraulic response is discussed in more detail in the following paragraphs, and the factors that influence the system behavior as the transient progresses are identified. System pressure response, system mass inventory / distribution, core thermal-hydraulic response, and break flow behavior (including the influence of ECC injection and break flow) are the primary phenomena of interest.

V 6.1.1 System Pressure Response The vessel upper plenum pressure for the small break transient is shown in Figure 6.

The timing of events which influence the depressurization rate are also indicated in the figure.

Immediately following rupture, the primary system fluid (exclusive of the pressurizer fluid) is subcooled and the depressurization is rapid. By between 38 and 42 seconds, the system depressurizes to the saturation point of the fluid occupying the volume between the vessel upper plenum and the inlet plena of the steam generators, as indicated in Figures 7 and 8 which compare the sessel upper plenum and intact loop steam generator inlet plenum volume equilibrium temperatures with their corresponding volume saturation temperatures. Bulk boiling of the hot fluid at this point (beginning near the inlet of the steam generator and working its way back to the vessel upper plenum) is sufficient to slow the depressurization rate (Figure 6) as steam generation tends to offset the effect of coolant volume loss at the break. The system then continues to depressurize at a a

21

O 2400 2200 Hot I*1/uccer clenum fluid saturates 2000 G342s; f

1800 ~

9 *500 f'

/

Steam generator secondaries t

ecome rest sources (530 s)

O 1400

• 1200 S 1000 '-

Acc:.mulater flow

~

W tegins (858 s) 8 800 Q-M ; umped flow 500 r-begins (1755 s) -

400 200 0

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Tim. (s)

Figure 6.

Ucper plenum pressure from the RELAP5 calculation.

660

'g-s

-- SATURATlON TEMPERATURE VOLUME ECUILIBRIUW TEMPERATURE 650 p640

's, v

s 630

's,

.$.620 m 8 610 I

3 600

-g 8.590 k,

s 2 5m

\\-

3

[ $70

\\ -

s

$50 i

g 0

10 20 30 40 50 60 70 80 90 10 0 Time (s)

Figure 7.

Upper plenum fluid temperature end corresponding saturation temperature from the RE' AP5 calculation (0 TO 100 s).

O 22

e a

0 9

660 i

650 VOLUWE ECUILlBRIUW TEMPERATURE

\\.

-- SATURATION TEMPERA URE

[ 640 - '

• g v

-\\-N

$30 3

$ $20 7

3 610 s.

S $00 N.

0

\\-

1, e

t,

- sw 3m i

3 l

~ 570 l

i 540 t

I f

?

e e

e 9

0 10 20 40 40 50 60 70 50 90 10 0 Time (s)

Figure S.

Stecrn generator inlet plenum fluid temperature and corresoonding saturation tem 9erature from the RELAPS calculation (O TO 100 s).

4 i

O i

23 i

I

gradually increasing rate until between 530 and 590 seconds when the steam generator secondaries become a source of heat to the primary system. After this point, the system continues to depressurize but at a gradually decreasing rate of depressurization.

During the accumulator injection period, the system depressurization exhibits a stepped response that is due to the combined effects of the accumulator injection, break flow, and loop hydraulic response. The mechanism for the stepped pressure response is discussed in more detail in Section 6.1.4.

By 1755 seconds the system is depressurized to the point that allows ECC injection from the residual heat removal pumps to begin. The system continues to depressurize further through the termination of the calculation at 1870 seconds.

6.1.2 System Mass Inventory / Distribution The primary system transient mass inventory for the small break calculation is snown in Figure 9.

During the first 500 seconds (or prior to clearing of the loop seal), depletion of the primary system liquid inventory is 0,uite rapid, as the system pressure is high and conditions in the break volume are primarily single phase liquid. Blowout of the loop seal at 520 seconds causes the fluid conditions in the break volume to change from single phase liquid to a relatively high void fraction liquid / steam mixture, which when combined with the steadily falling primary system pressure, results in a conside'rably reducea break flow rate and a leveling off of the system mass inventory. Accumulator injection (beginning at 858 seconds) initially results in a minor increase in system mass inventory, although a significant recovery of inventory is prevented because of the on/off nature of the accumulator flow and because of an increase in the break flow rate when subcooled ECC is present in the break volume ( see Section 6.1.4).

However, as the transient progresses, further reductions of the primary system mass inventory occur (between 1160 and 1325 seconds, and again between 1450 and 1750 seconds), even with the accumulators continuing to inject. These reductions in mass inventory occur as a result of the accumulation of liquid in the broken loop pump suction leg piping which leads to a significant increase in the subcooling of the liquid present in the break volume and a corresponaing increase in the break flow rate. The loop hydraulics during these periods of increased system mass loss are discussed in more detail below.

Initiation of liquid 24

- _~

=.. _ -. -.

4 I

4 1

l

?

l 2

i 500000 450000 i

4C0000 i

350000 7j 300000

• 250000 nno 200C00 2

~

150000 100000 50C00 l

0 O

200 400 600 800 1000 t200 1400 1400 1800 2000 Time (s)

Figure 9.

Total primary system moss inventory from the RELAP5 calculation.

1 l-e 25

injection from the residual heat removal pumps at 1755 seconds leads to a relatively rapid increase in the primary system mass as the combined ECC injection rate (from all components) becomes significantly greater than the break flow rate.

The increase in mass inventory then continues through the termination of the calculation.

The primary system mass distribution for the small break calculation is characterized by the voiding of the upper elevations of the system, with liquid collecting in the lower elevations.

Immediately following rupture, liquid lost out the break is made up by the liquid draining from the pressurizer, and the primary system remains essentially liquid solid. When the pressurizer has nearly emptied (i.e., at about 39 seconds), flashing of the hotter fluid in the primary system begins. Voiding of the primary system fluid occurs first on the upflow side of the intact loop steam generator (as fluid in the intact loop hot leg is at a slightly higher temperature than in the broken loop hot leg due to mixing with the hot pressurizer fluid), and then progresses through the remainder of the upper portions of the system as the transient continues.

Figure 10 and 11 show the collapsed liquid levels in the upflow and downflow legs of the intact and broken loop steam generators, respectively. As indicated in Figure 10, voiding of the intact loop steam generator begins at about 40 seconds, and the upflow and downflow sides are essentially empty by about 450 and 350 seconds, respectively. Voiding of the broken loop steam generator (Figure 11) begins somewhat later (at about 60 seconds) and the upflow and downflow legs are emptied by about 440 and 300 seconds, respectively.

Figure 12 shows the collapsed liquid level in the upper plenum portion of the vessel (covering the distance from the top of the core to the hot leg centerline), while Figures 13 and 14 show the void fraction in the intact and broken loop hot legs near the vessel. Again, each figure indicates a gradual depletien of liquid inventory beginning at about 60 seconds.

Figure 15 compares the collapsed liquid levels in the downcomer and across the core, covering the region between the hot / cold leg centerline and the l

bottom of the core barrel. The downcomer remains essentially full until about 350 seconds. However, depletion of the core liquid begins at about 60 seconds and continues at a relatively constant rate until about 225 seconds. At this point a reversal of the core flow causes a sudden decrease in the core level, although the flow reversal also allows liquid 26 l

30

\\!'\\,

UPFLOW S10E N.

-- 00WNFLOW SIDE 25 N

i 2 20

.\\.

S v

\\.

=

15

\\.,

O

't.

O>

8 I 'y ^.

l Q 10 -

\\!\\,\\

5

't

% N

's! v..-._,s L s

0 O 50 10 0 15 0 200 250 300 350 400 450 500 550 600 Tim. (3)

Figure 10.

Collapsed liquid levels In the upflow cnd downflow sides of tt e Intact loop steam generator tubes from the RELAP5 calculation (O TO 600 s).

30 I

\\.

UPFLOW SIDE

-- 00WNFLOW SIDE 25 s.

l s

@ 20

• g v

-g C

'\\

O 13

(.g o

I 10 J'.

t I

5

\\

. -.....L M w

^

0 l

0 50 10 0 15 0 200 250 300 350 400 450 500 550 600 l

Time (s)

Figure 11.

Oo Acosed IIquid levels in the uptiow and downflow sides of the broken loop steem generator ti.nes from the RELAPS calculation (O TO SCO s),

w 27

O 5

i 4

OO 3

J 8=

2 2

L.s 1

kbJhdMM/

0 100 200 300 400 SCO 6C0 700 800 Time (s)

Fiqure 12.

Collapsed liquid level in the core upper plenum from the RELAP3 calculation (0 TO 600 s).

0.9 k

e 0.8 1

0

]

0.7 c

& 18 3

0.s o"

0.4 L

10.3 Q>

0.2 0.1 0

0 50 10 0 15 0 200 250 300 50 400 450 500 Time (s)

Figure 13.

Vold fraction in the intact loon hot leg from the RELAP5 calculation (O TO 500 m).

O 28

1 ie i

i i

}

c 0.8 f.

I l

l'-]

0.7 i

C II l

J O.S 8

2 0.5 d

o 0.4 6

10.3 0>

0.2 0.1 0

O 50 10 0 15 0 200 250 300 350 400 450 500 Tim. (s)

Figure 14.

Void fraction in the broken loop hot leg from the RELAPS calculation (0 TO 500 s).

l n

21 DOWNCCWER fy

-- VESSEL (CORESIDE)

I ts

s.'.s 20

'3 h ll" Vh

,j

, II i

1 $

d d

Al 1

a i

t.!

f

'.,k

$e t

\\b

\\

Y 1

I

• g f

I 17

\\

v i

16

• g g

I e

10 ~

y 0 14 tes.

E v,,1 y-j, 10

.f 9

i i

~

g 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 Time (s)

Figure 15.

Downcomer and core collapsed :Iquid levels rela tive to the bottom of the core barrel frcm the RELAP5 calculation (O TO 750 s).

l 29

in the upper portions of the system to begin to drain back into the vessel. Thus between 225 and 400 seconds, the rate of decrease of the core level is reduced as liquid from the upflow sides of the steam generators and from the upper head and guide tube volumes drains back into the upper plenum and core regions.

As the upper portions of the system empty, liquid in the pump suction piping forms seals which impede the flow of steam from the core region around the loops to the break. The resulting higher pressure in the core / upper plenum region, relative to the break location, causes a depression of the liquid levels in the downflow legs of both the intact and broken loop pump suction piping and in the core.

Figures 16 and 17 compare the collapsed liquid levels in the upflow and downflow legs of the intact and broken loop pump suction, respectively, and show the decrease in the downflow leg liquid level below that of the upflow leg in both loops after about 400 seconds.

Referring again to Figure 15, a further depression of the core liquid level occurs after about 410 seconds (i.e., once liquid stops draining back into the vessel from the high points of the system).

The depression of the core liquid level at this time leads to an intermittent uncovering of the top of the core between about 450 and 520 seconds (see Section 6.1.3).

However, by 520 seconds the liquid level in the broken loop pump suction downflow leg reaches the bottom of the suction piping, and a rapid clearing (blowout) of the liquid in the upflow leg then begins (see Figure 17). The resulting flow path through the broken loop from the vessel to the break allows equalization of the upper plenum / break region pressure, and a rapid increase in the c re liquid level occurs (see Figure 15).

Between the blowout of the loop seal and about 1150 seconds, the core level remains relatively constant and the downcomer level exhibits a gradual increase as the combined ECC injection rate from all sources during this period is somewhat greater than the break flow rate. However, between 1160 and 1325 seconds and again between 1450 and 1750 seconds, the core level exhibits a significant decrease as indicated in Figure 18 which compares the long term downcomer and core levels.

The core level depression during these periods is a result of the reformation of the loop 30

m 14,

- DOWNFLOW S10E 12

)

-- UPFLOW SIDE J

N.,,,. e.,.,.. ; - - (J

{

,0

.4 v

8 I'

)

O O

S l

J I'e i l

4 1

I

.=

( )

j l i

4 l l

'(

i 2

)

0 O 50 100 ?SO 200 250 300 350 400 450 500 550 600 650 700 750 Tim. (s)

Figure 16.

Collapsed liquid levels in the upflow and downficw sides of the intact loop pump suction from the RELAPS

]

eciculation (O TO 750 s).

14 i

i i

i i

i i

i i

i i

i i

- DOWNFLOW SIDE 12

- UPFLOW SIDE t

/*.

&o N.

I r

~.,....e.<*'.,,,,..]

'1

,n f

j v

8 l

j eo I.

=

i O

5

'g

=

i U

l 4

l {l, l.

2 I

g I

$$A 'y_.f 0

0 50 1C0 150 200 250 300 350 400 450 500 550 600 650 700 750 Time (s)

Figure 17.

Collopsed liquid levels In the upflow cod downflow sicos of the broken Icop purep suction from the RELAP5 calculation (O TO 750 s).

l l

l s

l l

31

O' 1

l 22 i

21

-I 00WNCCMER 20 pl1' p'

-- VESSEL (CORESIDE) i,l

'ip.f;i, 'l "II l

!,Ih }

d

, ]

I, f

is -

i li I

l

/

c-ci

<i Il l

[

y' r

18 i

[

e

.! o it, c 14 j

t.'

i 11 y

i I,

j 10 9

e g

0 200 400 600 800 1000 1200 1400 16CO

'800 2000 Time (s)

Figure 18.

Collapsed Uquid levels in the cowncomer and core re lative to the bottom of the core borrel from the RELAP5 calculation.

l 32

i seal. Figure 19 compares the collapsed liquid levels in the upflow and downflow legs of the broken loop pump suction piping and illustrates the presence of the seal during the periods of level depression in the core.

The depression of the core liquid level during these periods again results in an intermittent dryout of the top of the core, although temperature increases are limited. Referring again to Figure 18, a rapid increase in both the downcomer and core levels occurs once injection from the low head residual heat removal pumps begins at about 1755 seconds.

6.1.3 Core Thermal / Hydraulic Responses The vessel liquid inventory for the small break calculation remains sufficiently high that the rod cladding temperatures stay below the steady state full power cladding temperatures throughout the transient, although as indicated previously, there are brief periods when dryout of the upper portion of the core leads to minor cladding temperature increases.

Figures 20 and 21 show the rod cladding temperatures at the 10 to 12 and 8 to 10 foot elevations above the bottom of the core, respectively, for an

, /

average power rod, while Figures 22 and 23 show the cladding temperatures at the same elevations for the hot pin.

Except for the period between 320 and 520 seconds and again between 1200 and 1250 seconds, and 1670 and 1740 seconds when intermittent dryout of the top third of the core gives rise to cladding temperature increases of a maximum of about 30 degrees Farenheit, the cladding temperatures generally remain a few degrees above the fluid saturation temperature. Figure 24 shows the cladding temperature on the hot pin at the 6 to 8 foot level above the bottom of the core, and does not indicate any rod dryout at this elevation.

The factors which influenced the hydraulic response of the core are indicated on Figure 25 which shows the collapsed liquid level in the core covering the active fuel region. As the system depressurizes following rupture, boiling of fluid in the core causes the initial decrease in core liquid inventory shown in Figure 25.

Figure 26 compares the vapor void fraction at different elevations in the core, and indicates the onset of I

boiling starting at about 60 seconds at the top of the core and progressing to the bottom of the core by about 229 seconds. By 225 seconds, the J

i 33

O 14

.s a

p

- 00WNFLOW SIDE U,

-- UPFLOW SIDE 12

.g 20 ;- * \\

,f rs

-- x ;t;

'\\. c/g) -).'

z, Il !

}

j I

8 1

v I

j l

5-i I

=

3 6

1 I

l l l!

4 l!ll, i

i i

c i

2 L '. !

!. j w

l 1

1 t

j I

ti l i

f)..

t i

, [MMd8h hh-fi h' O

0 200 400 600 800 1000 ?200 1400 1600

'800 2000 Time (s)

Figure 19.

Collapsed liquid levels in the upflow and downflow sides of the broken loop pump suction from the RELAP5 calcul ation.

I I

e O

I 34

(

l

575 l

-j i

es0 i

^

l'- 625

~

v

,500

~

5 575 d

~

0550l-

~

[525r-

~

_jW

~

~

475 m

~

450

~

@ 425

]

3 400 O

~

375

~

g 0

200 400 600 800 1000 1200 14C0 400 1800 2000 Tim. (3)

Figure 20.

Clodding surf cce temperature for the oversgo power fuel pins at the 10 to 12 foot eievoticn frem the RELAPS calculation.

650 )

4 i

i 4

a 525 m

~

DW

~

o 575 w

~

35s0 G

" 525 3

$ 500

~

475

~

5? 450

~

J 425 -

m 3 400

~

375

~

i e

i 3g 0

200 400 600 800 1000 1200 14C0 1600 800 2000 Time (s)

Figure 21.

Cicdding surf ace terrporature for the average power fuel pins at the 8 to 10 foot elevatten from the RELAP5 calculation.

35

O 575 N

650 o

a I

b 825 v

,SGC 5 575 i

O 550 w

i

[ 525 h500

~

475 m

C 450

$425 O-400 r o

375 350 O

200 400 600 800 1000 1200 14C0 16C0 1800 2000 Tim. (s)

Figure 22. Cladding surf ace temperature f or the high power f uel pin at the 10 to 12 f oot elevation from the RELAPS calculotton.

575

$50 m

b 625 v

, 600 j575 j

O 550

[ $25 h500

~

475 Cn c 450 3 425 0- 400 O

375 350 0

200 400 600 800 1000 1200 1400 1600

• 800 2000 Time (s)

Figure 23. Clodding surf ace temperature for the high power fuel pin at the 8 to 10 foot eleva tion from the RELAP5 calculation.

O1 36

1 l

875 1

850 p g s25

, 500

~

5 575

~

C 550 s

I

[525

.)

h500

~

~

475

~

._$450

~

$ 425 5 4C0 o

375 350 O

200 400 600 800 1000 1200 1400 1600 '800 2000 Time (s)

Figure 24.

Clodding surf ace terrporoture f or the nigh cower fuel pin at the 6 to 3 f oot elevation frem the RELAPS calculation.

is i

6 i

Dre 11auid begins g

to boil (60 s)

U

~

Qre flow reversal occurs (225 s) 1 ~

gg, ;3,,,,,o gje, llll Accumulator flow begins (1755 s) v 10 y

tegins

S58 s)

\\

l 28

{

l' f

i l

-}

I h

fi q

I i

7 i

k f

I i

ij

{

6

} bLeco seal elewout l

occurs (520 s) 5 C 200 400 600 800 t000 1200 1400 1600 '830 2000 Time (s) l Figure 25. CoGaosed liculd level in the core relative to the i

bottom of the active fuel from the RELAPS calculation.

N 37

O'

~

0.8 O TO 2 FT 0.7 ---- 2 TO 4 FT C 4 TO 6 FT C

O 6 TO 5 FT X 8 TO 10 FT o

0 10 TO 12 FT 0.3 V

0.4 1

0.3 t

..fc o

0.2 ~

l 0.1 e

{*

~'

~ - ' '"

'I 0.0 "

40 60 80 10 0 12 0 14 0 16 0 18 0 200 220 240 250 Tirne (s)

Figure 26.

Vold fraction in the core f or the six volumes encompossing the ocitve f uel region from the RELAP5 calculation (40 to 260 s)

O 38 l

O effective head of the primary coolant pumps is reduced sufficiently (pump coastdown begins at 35 seconds) to allow reversal of core flow. At this point, fluid in the upper parts of the system (i.e., in the upflow sides of the steam generators and in the hot legs) begins to drain back into the vessel upper plenum and core regions.

In addition, fluid in the vessel upper head becomes saturated at about 210 seconds, and also begins to drain into the upper plenum region.

Figure 27 which compares the flow rate at the inlet to the core with the core collapsed liquid level, indicates an increase in level shortly after the core flow reversal occurs, followed by a period (lasting until about 430 seconds) when the liquid depletion rate is greatly reduced. By 430 seconds, most of the liquid in the upper parts of the system is depleted, and the core inventory once again begins to decrease as boiloff, combined with the presence of the loop seals (discussed earlier) causes a depression of the vessel / core liquid level.

Figure 28, which compares the cladding temperature of the hot pin at the top of the core (10 to 12 foot elevation) with the vapor void fraction of the adjacent volume, shows the effect of the resulting dryout on the cladding temperature as the mixture level is depressed below the top of the O]

core. The oscillatory nature of the core level at this point, however,

(

causes the cladding to rewet shortly after each dryout, thus limiting the magnitude of the temperature increase. This period of dryout is terminated at 520 seconds by blowout of the loop seal, which leads to a rapid increase of the core liquid level (Figure 25) as liquid in the pump suction and cold leg piping is redistributed to the vessel.

As indicated previously, the period of dryout between 1200 and 1250 seconds is a result of the gradual accumulation of liquid in the broken loop pump suction piping which leads to a depression of the mixture level below the top of the core.

Figure 29 compares the collapsed liquid levels in the upflow leg of the intact and broken loop pump suction piping and include the collapsed liquid level in the core. As shown in the figure, the liquid level in the broken loop pump suction begins to increase at about 1150 seconds (due to liquid draining back into the suction piping through the pump). As the level continues to increase, the flow path between the vessel upper plenum and the break (through the broken loop) becomes blocked and a reduction of the core level occurs. (Note that the v

39

i 9

45000 15 CORE INLET FLCW

-- CORE COLLAPSED LICUID LEVEL 35C00 r7-

^

N n

E

~ ~.s 3 25000 L3,

E

' l'N

i. ;

- ?O

=

N C t5000 g

1 0

t

,g4 '

\\li$ & '

'J Q

tI4 S

h' I! '

} (

5000 I

' rg!h de%

l f

5

-5000 0

50 iOO ISO 200 250 300 350 400 450 500 Time (s)

Figure 27. Comparison of the collapsed IIquid level in the core region to the mass flow into the bottom of the core from relco5 calculation (0 to 500 s)

I O

40

1.5 700 i

i VOID FRACTICH

-- CLACDING TEMPERATURE c

O n

y o

1-f%

v T

2 r600 o

y eo.

0.5 l1 E

^

L~ ' \\

\\

- J L Jq r+ l\\

!\\

L.,j\\.]

I,

[

0 500 40C 450 500 S50 Figure 28. Corrporison of the cladding temperature at top of the hign power fuel pin to the vold fraction ct the top of the core from the RELAPS (400 to 550 s) 30 15 6

i i

i INTACT LOOP UPFLOW LEG

-- BROKEN LOOP UPFLOW LEG Q CORE C 20

+ 10 D

fm Mp,r

^

gj b

A b

j g-------

g

__s

- to 4--T i

I

\\

,,, i

~~l'd,

,u,y,&.

0 0

1?00 115 0 1200 1250 1300 1350 1400 Time (s)

Figure 29. Comparison of the core colaosed liculd level to the coBapsed liculd levels in the broken and intact loop pun p e6ctions from the RELAPS calculation (1100 to 1400 s) 41 n.-

,..,-,-n

intact loop pump suction has remained blocked with liquid since the beginning of the transient.) The reduction in core level is terminated at about 1240 seconds when liquid in the intact loop pump suction begins to clear.

Figure 30 compares the rod cladding temperature of the hot pin at the 10 to 12 foot elevation above the bottom of the core with the void fraction of the adjacent fluid volume, and illustrates the cladding temperature increase which occurs as the volume becomes voided. Again, the cladding temperature increase is limited by rewet resulting from the oscillatory nature of the core liquid level.

The period of dryout between 1670 and 1740 seconds is once more a result of accumulation of liquid in the broken loop pump suction, but the dryout is not caused by a blockage of the steam flow paths around the loops as occurred earlier.

Figure 31 compares the collapsed liquid levels in the upflow legs of the intact and broken loop pump suction piping, and again includes the core collapsed liquid level. As indicated in the figure, the liquid level in the broken loop suction upflow leg begins to increase at about 1450 seconds, and the leg is full by 1465 seconds. However, the intact loop suction leg begins to clear shortly after the broken loop leg begins to fill.

(The intact loop suction leg piping refilled previously at about 1340 seconds--see Figure 29.) Thus a path for steam flow from the vessel to the break is maintained and the depression of the core level is miminal. The effect of blockage of the broken loop pump suction piping, however, is to cause an increase in flow rate from the vessel toward the break, which in turn causes a considerable increase in the degree of fluid subcooling in the break volume as cool ECC liquid is carried toward the break. As a result, a corresponding increase in the break flow rate occurs which leads to the further gradual depletion of the vessel inventory shown in Figure 31, and the intermittent dryout of the top third of the core between 1670 and 1740 seconds. Figure 32 compares the cladding temperature of the hot pin at the top of the core with the void fraction of the adjacent volume, and again illustrates the effect of the dryout. The reduction in core level is terminated once RHR pumped injection begins at about 1760 seconds, and a further rapid increase in the core level occurs at about 1790 seconds as the broken loop pump suction piping clears for a final time.

42

t5 MO VolD FRACTION

-- OLACD1NG TEMPE9ATURE

-850 C

G m

O le g.

U 1

~~750 w

2 j

-650

[

I k

j i

0.5 4'

i l

~ 550 E

]

c.

I e

i U

L'I gnf.f,ft,q,Jgir,} g, 9't wA j

j,- 450 g

0 350 It00 ftSO 1200 1250 1300 1350 1400 Tim. (s)

Figure 30. Comparison of the cladding temperature of the top of the high power fuel pin to the void fractlen at the top of the core from tPe 9ELAP5 e?lculation O

(1100 to 1400 s) 30 15 a

INTACT LOOP UDFLOW LEG BROKEN LOOP UPFLOW LEG 25 O cope

_q 2 20

[

V

[

-9 V

15 y

yj

-6 g

g

• 0 r-g r

r r-]

k

-h'

--3 S

M.We#WY hm 0

0 1400 1450 1500 1550 1600

'650 1700 1750 1400

'850 1900 Time (s)

Figure 3L Comparison of core colicpsed liquid level to the colicpsed ilquid levels in the broken onc intact loop pump suctions from the RELAP5 calculation j

(1400 to 1900 s) 1

[

v 43

O 1.3 750 VOID FRACTION l - C'.AC0 LNG TDePERATURE l -703

-650 o

1 ll w

?lf ;'I ;)

'600 l

1

{

l 1.

~

~

u l

?

i III 2

~ 5s0 1

o i

3.

! il l

8 2

II l

-500 1

m/s -

e I

/1 0.5 I

h

/

9, f)s dl '#i

)

'j, jh' k 450

' I

.,%J+.

4,

.a.

s,_d I'

l a

350

!400 1450 1500 1550 1800

'650 1700 1750

'6C0

'850 1300 Time (s)

.Igure 32. Corrporison of the clodding tervoerature of the top of high power fuel pin to the void fraction et the too of the core from the TEL.APS :alcotton (1400 to i900 s)

-\\

O 44

(n 6.1.4 Break Flow Response The break flow response for the small break calculation is shown in Figure 33.

Following rupture, conditions in the break volume are single phase liquid, and the break flow decreases from its maximum just after rupture with decreasing primary system pressure. The slowdown in the primary system depressurization rate once flashing begins in the hot portions of the system (i.e., at about 40 seconds), causes the break flow to level off and remain fairly constant until about 190 seconds. By 190 seconds, the rate at which liquid is being supplied to the break volume by the broken loop pump falls below the break flow rate, and a flow reversal occurs in the piping between the break volume and the vessel. The flow reversal at this point is indicated in Figure 34 which shows the junction mass flow rate on the vessel side of the break' volume. The effect of the flow reversal is co carry cool ECC liquid back into the break volume, thus supplying the break with relatively low enthalpy fluid which causes a corresponding increase in the break flow rate. Figure 35 compares the break flow rate with the break volume equilibrium fluid temperature, (j

and shows the increase in break flow rate as the temperature decreases.

The break flow rate remains relatively high (under the influence of the ECC liquid being injected into the cold leg piping) until blowout of the loop seal begins around 500 seconds. At this time, conditions in the break volume change from primarily single phase liquid to a relatively high void fraction mixture as the liquid is swept away from the break into the vessel. The change to high void fraction conditions in the break volume causes a corresponding reduction in the break flow rate as shown in Figure 36 which compares the break volume void fraction with the break flow rate.

The break flow then remains relatively low until the initiation of accumulator injection (at about 858 seconds). The injection of accumulator liquid into the cold leg piping once again provides low enthalpy fluid to the break volume giving rise to an increase in break flow.

Figure 37 compares the break flow rate with the break volume fluid equilibrium temperature (for a period after accumulator injection begins), and again illustrates the increase in break flow rate with a corresponding decrease in fluid temperature, and vice versa. The cyclic nature of the break flow response during the accumulator injection period is due to the oscillatory v

45

i i

l O

l 2500:

l 2000 "

k 4

N Ef 1500 w

k b

k 1000 v

IN' ff e

4

'd if lI ji i

f 100 j

jpg y

b, O

200 400 600 800 1000 !:00 14C0 16C0 ?800 2000 Time (s)

Figure 31 Wass flow cut the breck from the RE' APS calculatlan.

O 46

I

?

2000 mm N

^

N E2 l

1 1

y 0

~

2 -1000 v

/

{

I

-2c00

'00 15 0 200 250 300 350 400 450 SCO Time (s)

Figure 34.

haces flow in the broken loop cold leg from the RELAPS calculotten (100 to 500 s).

700 2000

--- FLOW RATE FLulD TD4PERATURE

^

^ 600 t

e

500 f% [

3 i

?400 h4 -- 00 I

["\\,I%lllA '!

i

'" \\ 1i fu k

""_s J,l

.A.,. g '

yf 11

+-

0 300 1

} l1, jj

,8 C

3

\\

W \\$

U l

y e

C 200 f-o 10 0 O

100

?SC 200 250 300 350 400 450 500 Tim. (s)

Figure 35. Corrporison of break mass flow to the temperature of the fluid in the brook volume frorr. the RELAP5 calculation (100 to 500 s).

4 47

O IC00 1

k ROW RATE MO pj!

V010 F4ACT!CN j

?\\

l'\\!Y l '\\j \\ {\\ {\\ ;l \\ {\\.i, O 800

v '

\\

e

! \\j..J k

l.~

g

^

/..s/

i j i

i N 700 u

8 2 800 l~

l1

\\j 2

E l

'\\ ?'

l

~

2

~ 0.5 3

y SCO 400 V/

g 300 {-

p's ar 1

u a

l

'j i

4 j

2 200

,1 aco p

I 0

O 490 SGO aio 520 530 540 550 560 570

$50 590 Time (s)

Figuri 36. Compod scti of the break moss flow to the break volume vold *r ce tion from the PELAPS calcuistion (490 to ".,30 s) 560 1000

--- RCW RATE 540 FLUID TEMPERATURE

^

O n

j,i b

I 4,^.

s.

!' Aj ' \\\\,a %,l N

a

)

lisi

• 520

!'l ij; li i

E

=

de i 1 l

[

l

.c

'j r,

,hl'evv]*,/7'(y]

' l' ')

' I,[Ak, j v A

i[M) j g

~

480

-.- 0 ou CN{n g'

\\

r g

s l

,-=

c

"" 440 C

E 420 f

2 400

-1000 860 880 300 920 340 960 $80 1000 1020 1040 1063 '080 Time (s)

Figure 37. Coreporf son of the breck moss t'ow to t5e terrporoture of the fluid flowing to the Dreak from the RELAPS calculation (850 to 1080 s).

O 48

' low in the piping between the vessel and the break volume.

Figure 38 compares the break flow rate with the ficw rate on the vessel side of the break volume. The figure shows that when flow is back into the break volume (i.e., negative loop flow), the break flow tends to be high as cool ECC liquid is being carried toward the break.

Likewise when flow is out of the break volume toward the vessel (i.e., positive loop flow), the break flow tends to be low since the cool ECC liquid is then being carried away from the break.

Referring again to Figure 33, the break flow rate increases considerably during the periods between 1160 and 1325 seconds, and between 1450 seconds and the termination of the calculation. As discussed earlier, the broken loop pump suction piping during these periods is blocked by liquid (Figure 19), and the flow direction in the broken loop cold leg piping is directed primarily from the vessel toward the break. Figure 39 shows the mass flow rate on the vessel side of the break volume and indicates the relatively strong flow back into the break volume from the vessel. The net effect of the flow into the break volume is to supply the g

)

break with fluid that is as much as 100 degrees subcooled, as shown in Figure 40 which compares the break volume fluid temperature with the corresponding volume saturation temperature.

This large degree of subcooling in the break volume gives rise to the high break flow rates indicated in Figure 33. The break flow rate decreases from these high values only after blowout of the broken loop pump suction piping once again allows the cold ECC fluid to be carried away from the break (i.e., toward the vessel).

6.1.5 Similarity of RELAPS Calculation to Experimental Results The experimental basis for evaluating how representative the RELAP5 small break calculation results are relative to " expected system behavior" includes many small break experiments conducted in the LOFT and Semiscale test facilities.

Particular experiments which were similar in nature to the RELAP5 small break transient discussed herein include (but are not 8

9 necessarily limited to) LOFT Test L3-5 and Semiscale Tests S-SB-2 and S-UT-4.10 Each of these experiments simulated the rupture of a 4-inch ON pipe in the cold leg of a PWR, although boundary conditions (such as ECC 49

.~

l i

O 2000 4000 I

l h 2REAK FLOW EREAK VOLUME FLOW l g

g 2000 I

g l

-2000 E

I s

s 3

i ita 3!

E h;r*~-A.

',- ' N ' ! i'~

j b)

! l l t' 0

1s00 v.

-~~

f 3

Bo e

6 C 1000

--2000 g

g y

500

(

ff 1-4000 N YfY t N ff& *)

/

0

-6000 560 880 300 920 940 960 980 1000 1020 1040 1060 '050 Time (s)

Figure 38. Corrporison of the breck 9ow to the flow on *he vessel side of the beer.k volurne from the RELAPS calculation (860 to 1080 s).

1000 a

300 4 7

I I

l a 200 L

. lll

{

N 400l-l E

i' i

[.k[gh l

O

) l.

!?s i j l

i 3

l D p

1

}j 1 : 1 i

f f'

l g

1

-200,7 i

g

-400

' f' 1

4

.{II t

I u

. }

I:

4

.,1

-l Q,

J 2

qn l;

y

{,

f i

-600 l

.j (

i i

I f ij j

-a00 j

jj i

I i

i.

I e

_, coo 1000 1200 f400 1600

'800 2000 Time (s)

Figure 39.

ucas ficw on the vessel side of the breck volurre from the RELAPS calcuiation (1000 to 1900 s).

O SO

i I

i I

s i

j J

s u

e J

1 800

- FLul0 TEhePERATURE

-- SATLAATICN TEWTtATJRE

.l

^

W 500 -

j=;-

i'7iIT' ggg.

n v.y._z a

0

n.

1 1

e y

Aj 4

5 300 1

A-

.\\

\\ei 3*

j C

1 l

oo 800 1000 1200 14C0 iSCO iS00 Time (s)

Figure 40.

Comparison of the sreow volume fluid temoerature to the corressonalrg scturation terrperature from ?Ps t

1 1 ELAPS calculotton (300 to i300 s).

4 l

(

l l

51

injection rates and coro decay power) were not necessarily the same as those assumed for the RELAPS calculation.

Based on the similarities in overall system response between the small break tests conducted in the LOFT and Semiscale systems (and in particular the three tests mentioned above), and the RELAPS calculated system response, the RELAP5 results are considered to be representative of the system behavior that will occur in a small break LOCA. Generally, the system response in the experiments (including trends in data as well as particular phenomena which occurred during the transients) was similar to the response obtained in the RELAPS calculation for the RESAR-35 plant.

Voiding of the system progressed from the upper elevations downward with the gradual formation and then blowout of the loop seal. Depletion of liquid from the primary system via the break was not sufficient to cause a sustained uncovering of the core either prior to or following blowout of the loop seal, and no heatup of the fuel / heater rods was observed. The ECC injection rates from the accumulators were very low (because of the small differential pressure between the accumulators and primary system), and in some instances the accumlators exhibited an on/off behavior similar to that shown in the RELAPS calculation. Differences in the timing of events which occurred in the experiments, relative to the timing of the same event in the RELAP5 calculation, do occur. However, these differences in timing are attributable to scaling influences, as well as to the differences in boundary conditions assumed for the individual tests. The particular effect of cold ECC fluid on increasing the break flow rate (as occurs in the RELAPS calculation) was not readily evident in the experiments.

However, the ECC injection 11 cations in the experiments were located well away from the break location (as compared to at the break in the RELAPS calculation), and the injection rates were considerably less than that assumed for the RELAP5 calculation. Both of these factors would tend to reduce the amount of fluid subcooling at the break, thus reducing the cagnitude of any possible increase in break flow that may occur.

6.2 RELAPS/ Westinghouse Calculation Comparisons In this section, comparisons of results from the RELAP5 small break calculation and from a Westinghouse evaluation model calculation for a 52

i

\\

RESAR-35 plant are presented. These comparisons quantify conservatisms in the Westinghouse analysis required by 10 CFR 50.46 and Appendix K to 10 CFR 50.

The assumptions used in the RELAPS best estimate calculation and in the Westinghouse evaluation model calculation are discussed in Section 4 of this report.

The assumptions which provide the most significant degree of conservatism in the Westinghouse EM calculation relative to the RELAPS calculation are as follows.

First, the EM calculation assumes a significantly reduced ECC injection capability, both with respect to injection rates, and with respect to ECC fluid conditions.

Second, the EM calculation assumes a high core decay power. The net effect of the first assumption is to reduce the likelihood that the core will remain covered during the transient (and, as shown below, the core is uncovered for a period of about 500 seconds in the EM calculation). The net effects of the second assumption are to cause a higher rate of conversion of core coolant to steam (thus maintaining a higher transient system pressure, as well as assuring a faster depletion of the available primary system liquid inventory), and to increase the rate of fuel rod heatup and the magnitude of the peak cladding temperatures once the core becomes uncovered. Other assumptions listed in Table 1 of Section 4 add to the conservative nature of the Westinghouse calculation, but to a lesser degree. The comparisons presented in the remaining paragraphs of this section illustrate the conservative nature of the Westinghouse calculation, especially with respect to peak cladding temperatures obtained during the transient.

The primary system pressure response (pressurizer pressure) for the RELAP5 and Westinghouse calculations is compared in Figure 41. The considerably faster overall depressurization in the RELAPS calculation is attributed to the lower core transient decay power (see Figure 3), as well as to the higher ECC injection relative to the Westinghouse calculation.

These two factors, when combined, result in a generally more rapid cooling

~

of the primary system fluid in the RELAPS calculation, along with a corresponding faster reduction of the primary system pressure (saturation C\\

l%.)

53

l O

1 l

2500 lI RELAPS CALCULATICN l

-- EM CALCULATiCH i

CC00h 4

I e

l l' a

$ 1500 ld k

b.#Sa g

6 s

3 1000 r N-g eu s

A

$00 0

O 500 1200 iSCO 2400 Time (s)

Figure 41 Comparison of aressurizer pressures for the 9ELAP5 cno EW calculations.

o O

54

a

)

pressure).

(Figure 42 compares the total transient ECC mass injected V

into the system for both calculations, and illustrate the much greater amount injected for the RELAPS transient.) Although the primary system depressurizes to the accumulator setpoint pressure by about 800 seconds in the Westinghouse calculation (as compared to 858 seconds in the RELAP5 calculation), depressurization continues at a considerably faster rate in the RELAP5 calculation after this point. Thus, RHR pumped injection begins by 1760 seconds in the RELAP5 calculation, whereas the RHR setpoint pressure is not reached by the termination of the Westinghouse calculation at about 2500 seconds (although primary system mass inventory is increasing without RHR pumped injection).

The break flow rates for the RELAP5 and Westinghouse calculations are compared in Figure 43.

The difference in flow rates up to about 200 seconds is due to the difference in transient system pressure (Figure 41), with the higher break flow rate in the EM calculatior:

corresponding to the higher transient system pressure during this period.

p The increase in break flow at about 200 seconds in the RELAPS calculation Q

(as discussed in Section 6.1.4) is a result of the increase in subcooling in the break volume due to the reversal of flow in the broken loop cold leg. However, the increase in break flow in the Westinghouse calculation beginning at about 150 seconds is a result of increasing system pressure.

Both calculations show a rapid reduction in break flow as blowout of the loop seal (at 400 seconds in the Westinghouse calculation and 520 seconds in the RELAPS calculation) clears liquid from the break region.

The high break flow in the RELAPS calculation after accumulator injection begins is again a result of the increased subcooling in the break volume.

The Westinghouse calculation does not show a corresponding increase in break flow rate when accumulator flow begins because the broken loop accumulator is assumed to discharge directly to containment without interacting with primary system fluid.

i a.

ECC flow rates from all active components are first added, then integrated, to obtain the results presented in the figure.

\\

V 55

O S00000 RELAPS CALCULATION I

-- EW CALCULATICN l

600000 7j 400000 O

h2c0000

.. ~ ~',

• ']

1

~ ~~~~ ~ ~

O

-200C00 O

600 f200 iSCO 24C0 Time (s)

Figure 42.

Comparison of the total amount of ECC Inlected into the primary system for the RELAP5 cnd EM calculotrons.

2500 RELAPS CALCULATION

-- EW CALCULATICN 2000 n

N E

f 1500 5

k 2

j'g l

1000 ti ti

'h,/'1 1\\

I P

l hT* 4

k. -

i 0

0 600 f200 1800 2400 Time (s)

Figure 43.

Comparison of the break flow rates f or the RELAP5 and EW calculations.

Ol 56 1

The primary system transient mass inventeries fer the two calculations O

are compared in Figure 4A.

Considerably faster mass depletion occurs in the Westinghouse calculation following rupture because of the higher break flow rate and because of the lower ECC injection rates. The magnitude of the system mass inventory between about 340 and 850 seconds is sufficiently low in the Westinghouse calculation that the core becomes uncovered (see below), and a temperature excursion in the upper half of the core occurs.

The temperature excursion is terminated once accumulator injection begins, as the gradual increase in the primary system mass inventory causes the vessel mixture level to rise above the top of the core.

In the RELAPS calculation, the system mass inventory remains well above that in the Westinghouse calculation until after accumulator injection begins, and as discussed earlier the core uncovers only briefly during this period (i.e.,

just prior to loop seal blowout).

The primary system mass inventory in the RELAPS calculation drops below that in the Westinghouse calculation after accumulator injection begins as a result of the much higher break flow rate discussed above. However, because of differences in primary system mass distribution, only brief periods of uncovering of the core occur during the accumulator injection period in the RELAPS calculation, even though the V

total primary system mass inventory is less than the minimum inventory that caused uncovering of the core in the Westinghouse calculation. Both calculations show an increasing primary system mass inventory prior to the termination of the transient.

Figure 45 compares the mixture level in the core for the two calculations. RELAP5 does not calculate a mixture level. The mixture level presented in the figure for the RELAPS calculati.on is derived from core fuel pin temperatures.

The heat structure representing the fuel pin is assumed to be uncovered when a temperature increase occurs, and is assumed to be covered when temperatures follow the volume fluid temperatures. Again, the mixture level in the RELAP5 calculation drops below the top of the core only briefly during tha transient, and only the top third of the core is uncovered during these periods. However, the Westinghouse calculation shows a relatively prolonged period, starting at about 340 seconds, when the mixture level drops well below the top of the b

57

O 600000 RELAPS CALCULATION EW CALCULATICN S00000

'\\

^

6 400000

's,

.o O

\\

\\

se o 300000

\\

2

\\

\\.\\

200000

\\,

f..

%,,~., y 100000 O

S00 1200

'500 2400 Time (s)

Figure 44.

Comparison of primary system total mass for the RELAP5 and EW calculations.

14 i

12 1,

I I

10

,[j j

m C

. 'f l

v i

i I

8 ll c

o l

=

le

\\

i O

8

'd

's.,.

l a

s.

r w

4 2

RELAPS CALCULATION

-- EW CALCULATION 0

O 600 1200 1800 2400 Time (s)

Figure 45. Comparison of core mixture levels f or the RELAP5 and EW calculations.

O 58

/'"'T core with the minimum mixture level being at about 5 ft above the bottom of s-/

the core. The mixture level remains below the top of the core until about 850 seconds when accumulator injection causes the rapid level increase shown in the figure.

Figure 46 compares the cladding temperatures on the top 2 feet of the hot pin for the two calculations, and illustrates the differences in temperature response corresponding to the differences in core mixture levels. The peak cladding temperature in the EM calculation reaches about 1760*F, as compared to a peak cladding temperature in the RELAPS calculation of considerably less than the full power steady state temperature of about 670*F.

Figure 47 presents similar results to Figure 46, but for the 8 to 10 foot elevation above the bottom of the core. The hot pin heat transfer coefficients for the top 2 feet of the core are compared in Figure 48, and Figure 49 compares the hot spot fluid temperatures for the two calculations.

The differences in response due to core mixture level differences are again evident.

s l

.I

=

9 J

\\

59

O 2000 RELAPS CALCULATION p

j

-- EW CALCULATION a

w j

1. 1500 -

/

I

/

l

.I

\\

a t

/

f 1000

[

{

?

l*'

l I

c.

Y m

500 E

~

~~

O O

O 200 400 600 800 1000

'200 14C0 16CO

'800 2000 Time (s)

Figure 46.

Comparison of high poner fuel pin cladding termoeratures for the RELAPS and EW calculations of the *O t o 12 f oo t e levo tt en.

1400 RELAP5 CALCULATION l

/ N,

-- EW CALCULAT10H l

p 1200

/ l v

/

i o

t

/

i l

1000 l

C l

l i,

e f

f800-

/

g I

pl

/

e o

_f 500 j

)

e

~

7 o

2 400 y

0 200 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Time (s)

Figure 47.

Comparison of t'io hign power fuel pin cladding i

temperatures for the RELAPS and EM calculations at the 8 to *O foot eleva tion.

.l O

60

18 e

i i

i

~

r=- ~-:

- i l

(

'[ l

(. (- _ - f _ _

c

• L

-- J a,

1.,_

_ j_

,,, L, j_

-nOb I,,,

V E

RELAPS CALCULATION 3

$1

-- EW CALCULATION l

oc yb

~

l

~

A n :s c-1 cm Ig r

i s

I t

e I rs._'

1 0

\\ /

f g

q 1d 300.0 400.0 S00.0 600.0 700.0 800.0 900.0 Time (s)

Figure 48.

Comparison of the boot transfer coefficient of the 10 to 12 f oot elevation on the hIgh power fuel pin for the RELAPS and EW calculatione.

1800

(

RELAP5 CALCULATION

-- EW CALCULATION 1500

,a'.. \\

e' l

I

/

/-

1200

//

/

I 900

,/

\\

i,\\

i.

/

/ %

~a '.'

l 3,,

300 400 500 600 700 800 900 Time (s)

Figure 49.

Corrporison of the hot spot fluid temperatures for the RELAP5 and EW calculations.

J 61

7.

CONCLUSIONS The analysis of the RELAPS results from the RESAR-3S small break calculation, and the comparison of the RELAP5 results with the results from a Westinghouse evaluation model calculation for a RESAR-3S plant have led to the following conclusions:

1.

Based on the similarity of the RELAP5 results with available small break experimental data, the RELAPS results are considered representative of the response expected for the limiting small break transient with "best estimate" assumptions.

System response to the small break is characterized by a continuous primary side depressurization with only brief periods of dryout of the top third of the core. Maximum cladding temperature increases during the periods of dryout are limited to less than 30 F, and cladding temperatures remain less than the full power steady state temperatures throughout the transient.

2.

The requirements for evaluation model small break calculation specified by 10 CFR 50.46 and Appendix K to 10 CFR 50 result in significant conservatisms in calculated system response relative to "best estimate" calculations.

Comparisons of the EM and best estimate calculations show a considerably slower overall system depressurization rate in the EM calculation, and a prolonged period during which the top half of the core is uncovered.

Peak cladding temperatures during the period of core dryout in the EM calculation reach about 1760*F, again as compared to peak cladding temperatures in the best estimate calculation of less than the initial temperature of about 670 F.

The differences in system reponse between the two calculations are attributed primarily to the significantly reduced ECC injection capability and to the higher core decay power assumed in the EM calculation.

O 62

[

8.

REFERENCES 1.

V. H. Ranson et al., RELAP5/M001 Code Manual, NUREG/CR-1826, Idaho National Engineering Laboratory, Novemoer 1980.

2.

Westinghouse Nuclear Energy Systems, Reference Safety Analysis Report, July 1975.

3.

V. H. Ransom et al., "RELAPS System Modeling Capability for Small Break LOCA," ANS Soecialists Meeting on Small Break Loss-of-Coolant Accident Analyses in LWRs, Monterey, California, August 25-27, 1981, pp. 3-13 to 3-25, 4.

H. S. Ozmelek and D. A. Driscoll," Comparison of RELAPS and RELAP4 LOFT Small Break Experimental Safety Analysis Models," ANS Specialists Meeting on Small _Sceak Loss-of-Coolant Accident Analyses in LWRs, Monterey, California, August 25-27, 1981, pp. 3-135 to 3-166.

5.

H. H. Kuo, H. M. Chow, and V. H. Ransom, "RELAPS Horizontal Stratified Flow Model with Application to a Wyle LOFT Nozzle Calibration Experiment," ANS Specialists Meeting on Small Break Loss-of-Coolant Accident Analyses in LWRs, Monterey, California, August 25-27, 1981, pp. 6-91 to 6-110.

6.

C. B. Davis, S. M. Modro, and T. H. Chen, "RELAPS Calculations of the gx Effect of Primary Coolant Pump Operation During LOFT and Semiscale s

Small Break Experiments," ANS Soecialists Meeting on Small Break Loss-of-Coolant Accident Analyses in LWRs, Monterey, California, August 25-27, 1981, pp. 6-111 to 6-127.

7.

M. T. Leonard, Posttest RELAPS Simulations of the Semiscale S-UT Series Experiment, Idaho National Engineering Laboratory, October 1981.

8.

L. T. L. Dad and J. M. Carpenter, Experiment Data Reoort for LOFT Nuclear Small Break Experiment L3-5/L3-5A, NUREG/CR-1695, EG&G Idaho, Inc. Novemoer 1980.

l 9.

J. M. Cozzuol, Quick Look Reoort for Semiscale Mod-3 Small Break Test l

S-SB-2, EGG-SEMI-5073, EG&G Idaho, Inc., December 1979.

f 10.

D. J. Shimeck, J. E. Blakely, B. W. Murri, Quick Look Report for l

Semiscale Mod-2A Test S-UT-4 EGG-SEMI-5429, EG&G Idaho, Inc. April l

1981.

l

('~')

r

\\_ /

63 i

.l I

i 1

I 5

1, h

i 1

1 i

J l

I i

t 1

APPENDIX A RELAPS UPDATES f

i i:

1 J

t a

l l

i f

j l

I I

I l

i i

f a

i n

l J

9 O

J b

Il A-1 1

APPENDIX A RELAPS UPDATES Table A-1 contains the updates to RELAP5/M001 that were used up to 1522 s.

Table A-2 contains the updates used from 1522 s to the end of the calculation. Comments in the update files explain the purpose of each update.

O S

9 O

A-2

TABLE A-1.

RELAPS UPDATES USED FROM 0 TO 1522 S.

• COMPILE GEFINE,SEGdIR

• /

-*1-00 ft 0 RS-V EbC Cti-f-T E R tt-I N-t+0 MittT U M-F L-ttX- ?Q u arfi O N

• /

• IDENT SRFIX

• 0ELETE,VcxPLT.142,VEXPLT.143 C uttV F=CONWF+A 85 (VELF t t.-H *0IFVF( k4 ) *S C tagM

• AR AT t-! +1-) **E quNVG=C ONVG+ ads t VELG(L ))*01FVG(L4) *S;R A.H* AR A T (I +1)* *2

• uiL E T, V cx PL T. l*o, V E xPL T.149 CONVF eCONvF-ABS ( VE LF (K ))*0!FV F( 44 ) *S:4 & Ci* AR A T (I) ** 2 CONV G =C ONVG-4 es t-VELG ( K 1-)*0 Z F-V G( K4 ) *S C 4 AC H + AR A T (-I-) * *2

• /

• IDENT RJWR017

• /

-*/

A 00-MA S S--E R R OR-A tt o-f 0 T A t-MAS S-T O--S tA tt-t?CU E S-T-tA R I A8t E1.

• 0 VndQS.4 INTEGER T1(4),72(20),T3(10),T4(5).T5(5),T6(4),77

• 0 VkEJ0.4 O A-T A-T tt *T-fME *r* CPU T f Mi *s "E M A S S "s *T *t A S S */

• u SCNREQ.76 15 00 20 I = 1,4

• 0 SCNREQ.do,d2 21 1 F-ti T MI ; eT.-01-G O-TO-19 uG TO (24,22,25,26),.1 19 GO TO (301,501,17,17), I 26 PCKC00 = SHIFT (.NOT. MASK (42).ANO.(LOCF(TIMEHY)-LOCF(FA(31)),19)

~*I-5Ctt&Eov+G SHIFT (.N0T.M ASK (42). ANO. (L OCF-( E M ASS)-LOCF( F A(0 ) ) ),13) 25 PLxCc3

=

GO TO 16 S HI FT(.NOT.M ASK ( 42). A10. ( LOCF ( TM AS S )-LOCF( F A( 011 ),19) 26 PCKCUD

=

15--TP t TT fP irSQ.-07-G0-70-1000 17 CChV = CUNNF Ti(I)

LA8L(1)

=

UNMS(L)

LABLt3) =

Gtt-T O-90 G

~

• 0 SCNGEQ.42 INTEGER UNHT(4),UNHC(4),UNMS(2)

• I Sc hR t:Q.2 4 G Kt A-t!NMTt*t1t01*T*tt81*i

• O WRPLIO.61,66 J

=J

+1 K = K + 1 00 -2 fr-1

=-r, 4 IdVF(J)

Tl(I)

=

18UF (K )

0

=

J

=J

+ 1 26 CONTINUE

• 0 PL TWg t. 29 LF A t J+ 3 )

EMASS

=

tFA t-Jt4 )TTMA13 J+5 J =

j ENSORES THAT ALL RESTART VARIABLES 4RE DEFINED

• O RJhROJ9.33,5d

• /

• / 00h0RS MASS TERMS IN MOMENTUM FLUX EQUATI31 b RARROl7.2,3 130 CONVF = CONVF*ReVFJ CONVG=CONVG*RGVGJ V)

A-3

TABLE A-2.

RELAPS UPDATES USED FROM 1522 S TO END OF CALCULATION.

ariti U E F i iE,3 ETN 4

• / 30NORS VELOCITT TERM I ?4 1315:4 T UM cLUV EQUATION

=/____

____3<rt4

- t v e rn

• 0 dL E T E, VE X P L T.14 2, VE X 7 L T.14 3 CJNVF=C3:41F + A33 (VE LF ( L I) =3 !FV C(L 4 ) *S CR AC 4* AR A T (I +1) s* 2 CJNVG=C J1VG

• A33 ( V E L GI L ))

• 3 IF / G(L 41 =iC R t:9*ARAT(I+11=*2 Mit~tT1E,1 c G u i. L + d,TTM i. i s h C JHV A =C11VF-43)( /E LF ( < )) = 3IF / F( < 4 )*SC R t C9 a AR AT (I) ** 2 C1NV3 =C'J 413-A3 3 ( / E LG l( ) )

• 0 IF1G( < 4 ) *i:R A Cit iR A T ( I) * *2

• /

'I

.146:3 ALL 4thi4dl VAdiACLt) vtPLNtU

• 0 RJwRCC9.33,58

• /

• / OCNORS MASS TERMS IN MOMENTUM FLUX EQUATION au r%RAQli.2,3 CONVF*pFVFJ 13C CCNVF =

C CNV G =CONVG *RG VGJ 3 */

+/ UFK NAS) cMMUM P L A c)

• IDE NT OPKR 018

• */

i */

A00_ MASS ERROR AND TOTAL MASS TO SCAN #EQUEST VA2 TABLES.

, *v vrew).*

IN TEG E R T 1( 4 ), T 2( 20 ), T3 ( 10 ),74 ( 5 ), T5 ( 5 ),76 ( 4 ), T7

• 0 V RE 30.4 0ATA T1/"T I mea, ac PUTIMi a.acM ASS a,aT M A SS a /

' =u sc;44:g.to 15 OG 20 I = 1,4 2

• 0 SCNREC.80,J2 21 IF (ITYpI

.LT.

01 GO TJ 19 4

QU IU 44*,44,45,23)e 1 19 GO TO (501,501,17 17), I 26 P CMC 00 = S HIF T (.NOT.M AS 4( 42). ANO.(L9CC( TIME 4Y )-LOCF( F A(0 ))),18 )

- *I S CN RE C.90 3

da ecscuu = > d t r i t. 't u i.142 s t + 4 p. a'a u. L L ac e t : i a a ) ; -L ac e ( F a t a ; 31,1 s 1 30 TO 16 25 P C :(C O O = SHIF T (.NOT.1 AS( ( 4 2).410. (L 7CC I TM AS S )-LOC F( F A (01 ) ),13 )

15 IF (ITY7.53

3) GO TO 1330 A(

LUNV = G 'Jt N P LABL(1)

Ti(I)

=

UNMS(L)

LASL(3)

=

3 GO TO 930 m+u 3 c:4 < = w. 4 2 IN TE GE3 UNH T( 4 ), UNHC ( 4), U11S (2 )

1

' *I SCNREC.54 0ATA UNMS/a(KG)n,n((g}w/

- *u wartit.oA,co J + 1 J

=

K = K + 1 1

OC 26 I = 1,4 i

id vfl J )

= 11t1J a

ISUF(K)

=0 "l

J =

J + 1

-a K = K + 1 4C bbNi1NUc

• 0 P LTWR T.29 LFA(J+3)

EMASS

=

LFA(J+4)

TMASS

=

J = J +5

• /

PASS ERROR IN STATE

• 0 HXCROC8.2199,HXCR010.8 CC... DEBUG PRINT CF MASS ERROR C

IF (HELP.NE.

0)

WRITE (OUTPUT, 3001) T!1Eif, 3T, NC1UNTe 4EL', SUCCES, FAIL A-4

TABLE A-2.

(continued)

TMASS 0.0

=

T 0 5."4 S

=

0.0 d4nMMX

-1 0

=

OEMAXS

=

y.0 20" ASS 0.0

=

GLOBAL 0.0

=

J v0 L-5 0

=

LT ER

=

0 JTER 0

=

C

-O tH5 00-I--I y r-IV E,-I-V S C IF (VCTRL(I)

.LT. c) GO TO 1503 FRAHG RH0(!)

RHOM(I)

=

OcMASS V(I)

• ERRHO

=

9 T f* A >S

=- V ( I+-+- R H0 01-)

C TOEMAS T06 MAS + DEMASS

=

TMAS3 TMASS + DTMASS

=

Gt08AL

- G L OB A t: +-DettA S S-*--DEM A S S SUMASS SUMASS + DTMASS

• DT9 ASS

=

JVOLS JVJLS + 1

=

Ett ib" eER R HO-/-RHOF ( I-)

C EARMX ERRHU / RH0(I)

=

C A F-FA 8 S-( Eutt 1 4 t-fr-2. 0 E-3 t-GO-70-1-SM VCTRL(I)

VCTRL(I)

.OR.

SHIFT (MASK (1), 13)

=

1501 IF (ABS (ERRM)

.LE.

ERRMMX) GO T3 1532 ITER I

=

ERnMMX

- A&S ( E RR M) 1502 IF ( ABS (OdM ASS)

.Lt. DEMAAS) GO TO 1535 0

JTER

= 1

\\

DEMAX ABS (DEMASS)

=

\\

iMS-IF-tH CP ;ftEs-0)

N- ^

• wRITd (OUTP LT, dolo)

I, VOLN0(I). V(

),

RH0(I), RHOM(I), ERR 1, ERRay, DEMA S 15C0 CONTINUE

_-C dMASS

=

EMASS + TOEMAS SQRT(SUMASS / FLOAT (JV3LS))

eMSVRO

=

SUMASS 1.0 / SUMASS

=

3 C-AtER-

- 50 *T-F3 6 &&ltr- +--F t0kT H W3t. S t-/-AfttX J Fk,--H VOL 3 itH WERRMX

= DEMAXS / RMSVRO 4MSFRR

= SQRTjGLOBAL

• SUMASS) u JLOBAL 3CALcA

• RMSERR

=

t F- ( V E R R tt X7t E.-t R R MM X-)-G O-TO-1510 IF (JTER.NE. 0) VCTRL(JTER) = VCTRL(JTER)

.OR.

SHIFT (MAS ((ll, 13)

GO TO 1511 u

3:

15A4 IF (ITER.NE. 0) VCTRL(ITER)

VCTRL(ITER).02 SHIFT (NASK(1), 13)

=

.S1 5R1tft U

=-A ftA ri-t E P R MttX s--VE R R *t X e-G t-J t4 t1

,.7 AF (dRRMAX.GE. 2.uE-3) SUCCES = 1 if (HELP

.NE.

J)

+-wtIfi-(CU T eu h-8030)---T 0 E M A S s-T M A S S,-5 4 4 S ST-E RR MMrs-V0t N0 tt T E tb VERRMX, VOLN0(JTER), GL3BALs RMSVR3, RNSERRs SJCCES, sRRMAX

s C

-*0-H r CR 00eT2tl6 10C0 IF ( HtL P.EQ. 0) GO TO 1650 WRITE (OUTPUT,8001) TIMEHf, 07, iCOUNT, 4ELP, SUCCES, FAIL IF (CHECK

.EQ.

O) WRITE (OUTPUT,9032) CHECK

~

TF i C Hc CR TIN. ii WRtT2-t00T PtJTTt0031-0MW QALL STRACE WALL HEL P fR ("ST ATEM )

• I HXCR00d.2217 CC *** FORMATS A-5

1 TABLE A-2.

(continued) ou Jr90P.04 j

Bbd CCNTINUE i

IUIJS

/0IDFJ(il + V010 GJ ( I )

=

IF (HELP.EQ. Of GO TO 12 VOIOSK VOLDF(K) + VOIOG(4)

=

wR-I-T+--( 0 tit 70T r90 22 )--aKVO L a,- VO I D F( < Fr-RHO F ( r( Pr-UF ( K-),

V010G((). 4 HOG (4), UG(K),

1 CCA4, VOIOSK l

VCIOSL

= VOIDF(L) + VOIOGtL) d RE T E-(00T PU T,9024)-*L VO L *, -V O L O F ( L t,- R HOF t er, -UF ( ti r VOIOG(L), RHOG(L), UG(L),

CCAL, V3IDSL F

= VELFJ(I)

VG

=-YsLGJ01, 4 RITE (JUTPUT, 9021) "JNEwa s VF, VOI3FJ(I), RHOFJ(I), JFJ(!),

VG, /JIDGJ(I), R40GJtI),

'J G J ( I ),

A J(Il e VOIDS 1:

CDWTInUE IF ((VELGJ(I)

• VELFJtI)). LE. 0. 3) 30 TO $4d oI JPROP.114 IF (HELP.EQ. Of GO TO 900 VOEdi

=-V OI D F-J 01-1-+- V O ID G J t I l-4 RITE (GUTPUT, 9023)

"STR",

VOIDFJ(I), VOIJGJ(I), /0 IDS 80 JrROP.119 IF (VOIOS

.LE. 1.0) GO TO 903

-+0-t P Rt)Pvild rl7 5---

IF (McLP.NE. 0) WRITE (OUTPUT,9323) " NORMA, VOIDFJ(I),

VOIDGJ(I), VOIOS

• I JPROP.177

-C C *** JEBbG FOMMATS C 90C0 FORMA T(a0", 134(a**),/,8 JPROP DIAG 1OSTIC #RINTOUT, TIME 4Y

=a, etP

• G 15. 77*r-0 T,r*rlftfilST 77 *i-1CO U N T,*rtt0,-*r-M E t? =*rIt

= ", SUCCES =a,I3,as FAIL

=",L2) 9010 FuaMAT("0JUNCTIJN 00NORED PROPERTIES, ICHOKE

=a,I3,*,

<BAO

=",Ilo,

• /," a,144(a=a),/,

- *r2X r*f *r

• X r -J U N N G FI-t ar8Xr Ma a

a r&X F *V0bMOtMc)-*r9X r t*1&Y r V0tO f bt*-

a a

• ,/,

a,6X, aVEL FJ - V ELJ a,4X, aVOIDFJ a. 7 %,84 HOF J a,9 X, aVFJ a,6X,

=a

• aVELGJ - VcLJas4Xs aVOI DGJa,7X,"RHGGJ a,9 X, auGJ a,10X,"CC AJ a,3 X,

*VtittiS *r/T*-*rt3 4 t *** M 9020 F OR M A T ( A1, I o,111, Ilo, I 15, I11, I15 )

9021 FORPAT(a a, AS,1P1GG13 5 )

9022 FuaNAT(a asAa,13X,1P3G13.5,13X,1P5G13.5) 9;c3-F il? MK Tt*-ay A :r,13 X r1 P 1G 13. 5, 2 ( 39 t rt P-1 G 13 v 9 Pi 903C FORMAT (a CHOKcD.. 5 KIP JUNC TICN 00NJRINGa) c/

UPDATES 70 NE00 PREScQ...VFINL A10/ JR CUT THE TIME S TE) 4N0 RE8 EAT 0/

(auCCES = 3) IF VELOCITIES FLIP-FLOP. (SJ3R00 TINES S YSSOL, VFI1L,

.~ ei G IS-T S 7, r.f0R07-TRANi 00 SYS50L.80 P(I)

P0(I) + SOURCP(J)

=

80 VFINL.2 Su8 ROUFINi-VFINt( IC YCLE r-J R E 00)

• A VFINL.11 REAL VELFJX(1), V E L GJ X (1) d4UIVALcNCE (VELFJX(1), VELJ(1)), (VEL 3JX(1), VELJ0(1))

R:At-ERRH3tt)

EJUIVALENCE ( E RRHJ (1), VNEd(1))

LUGICAL ICYCLE

. 01 VF INL.12 0

I N I Fi-*t-i-fi-ft*tS-iR*t]R =-0VG ICYCLc

. FALSE.

=

Ju 20 I = I V, IVE, IV5KP ERRH0(I)

= 0.0 29 C ON TI nu!

IF (J nE 00.G T.

1) GO TO 33 A-6

T TABLE A-2.

(continued)

• 0 RJWR009.02 Ir ( V3105.LE.

1.0 )

GO TO 599

• 0 WJ.ROO9.63,115

• /

J PROP.. 00 NUMERIC AL AVER AGE OF JUNCTION PROP RTIES AT J VELOCITY,

• /

INCLUDE REDONun LOGIC F04 VEL 3 CITY FLIP-FLOP ( I:40<? = 2e

-; i MOD I F IES-1M RE G... J UM-OtT4-3t0 t)-rNO-05tf T E-PtJ41-t33IPPf]F

• /

V010FJ + VOIOGJ = 0.0.

• I JPROP.11 JEAL /ELFJX(1), VELGJX(1) c o ul v a L-E N Crt v it F J x-ttt,-V E L-J tit) e-t'/ E t:G ittri e-VEtJ 0 ttri INTEGER OUTPUT DATA QUTPUT/6LOUTPUT/

--C

~

LINCTL

=""

AF (HELP.EQ. U) GO TO 10 WRITE (durPUT, 9000) TIMEHY, DT, NC3UNTe HELPe SUCCES, FAIL C4LL STRACE WRIld (GUTPUT, 9010)ICHOKE 10 CONTANUE C

--*0 Jee0P.1:r(16

• I JPROP.21 INAME

= "JOLO" VF

= VELFJ0(I)

VEtGJ0Tri VG VOIDFJ(I) + VOIDGJ (I)

VOIDS

=

IF (ICHJKE

.EQ. 0) GO TO 1 IF (ICHJKE

.EQ.

1) GO TO 2 INAME e-*JEXPN VF VELFJX(I)

=

VG

= VELGJX(I)

IF (SHIFT (IMREG(1), 43)

.GE. 0) GO T3 900 O:

AF ((17Tr7). LT s 7).AND.

($HIFT(JC(I+1), 4)

.LT. 0)) INAME

• SKIP"

=

GO TO 1 2

IF ((IJ2(I)

.LT. 0). Atl0.

CSH TF1 i J L. TIT 1), 417Ci. ~i i i fatt r essxTP-1 IF (HELP

.EQ. 0) GO TO 3 WRITE (GUTPUT, 9020) LINCTL, I, JUNN0(Ils 4,

VOLN0(K),

L, VOLN0(L)

LIMCTL eQ"

=

WRIIE (OUTPUT, 9021) INAME, VFe VOI3FJ(I), RHOFJ(I), UFJ(I),

VGe V0103JtI), RHOGJ(I), UGJ(I),

CCAJ(I), VOIDS 3

IF71MKME.EQ. " 5itrF

• i GG TD 90G

• O sPROP.26 IF (VELFJ (II) 100, 200, 300

• D JPROP.33,42

~

V DFJtI) 3.5 * (VOIDF(K)

• RHOF(<3 + VOIDF(L)

• RHOF(L))

=

J.5 * (AMAX1(1.0E-15. VOIDF(43)

• RHOF(K)

• UF(<)

UFJ(I)

=

+

AMAX1(1.3E-15. VOIDF(L))

• RHOF(L)

• UE(L))

VOt0F7tt)-- MGIDF7ttr 7 RHOFItti UFJ(I) / (RHOFJ(I)

• AMAX1(1.3E-15, VOIDFJtII))

uFJ(I)

=

• O JPROP.56 IF (VEL 3J(II) 500, 600, 700

- *0-7P ROP'. 56777

~

• O JPROP.79,80 V010GJ(1)

= J.5 * (VOIDG(K)

• RHJG(<) + VJIDG(L)

• RHOG(L))

0.5 * ( AM A X1(1.0E -15.

VOIOG(<ll

• RHOG(K)

• UG(()

UGJ (I)

=

Airrrt1 TOE <r5mmtt-i i & R40Gili JGri i i

+

v CCAJtI) 0.5 * (AMAX1(1.0E-15. V0 0 G ( K ) )

• RH0G(K)

• CCAK

=

+

AMAX1(1.0E-15. V0;,0G (L ) )

• RHOG(L)

• CCAL)

WOIOGJ(I)

VOIOGJ(I) / RHOGJ(I)

=

1.J / (RHOGJtI)

• A1AX1(1.0E-15, VOIOGJt!)))

R R*40G J

=

UGJ(I)

UGJ(I)

• RRHOGJ

=

CCAJtI)

CC AJ (I)

• RR HOGJ O

=

A-7

TABLE A-2.

(continued)

dCwl FGRMAT("]

,1r iG 15. 7,"",,134("*"),/," MASS E840R 3[AG10STI PRINTOUT, TI ME4Y ="

OT ="slPlG15.7,",

1CTJNT

=",[1],

=

3 X, " V ",11 X, ""k HO ",10X, "k H OM", S X, "0 R HO /= ", L 2, /,

• 0 5, A 3 2 (

", suCCES =

,13,",

FAIL I",5X,"V3L10",

RHOF",3X,"0 RHO / R40",4r,

• V-*-O n H D n t,

=

r "s102(*="))

___ 8002 FORMAf(" STATE... RESET TO OLD TIME _3ROPERTIES,_t CHECA

=",!3,"

I")

~

6003 FORMAT (" STATE... FINAL SOLUTION PRO 3ERTIES, ( CHECK

=",I3," )")

C 801C FORMAT ("

",Io,Ila,1P6G13.5)

___ c_

8030 F O R." A T ( "sfSic}M

", J2("=")s/,2X," TOTAL 3E4AS3",63(*7/,

"),1PlG13.5,/,

• a0 TOTAL MASS (TMAS3) =8,ldlG13.5 CUMULATIVE MASS ERROR (EMASS) =",1P1313.5,/,

4AIfMUM-DR HO--/-R90F w*ritlG r3T5~,

• " ( VOLNG

=",I10,"

!",/,

M A X I.1U M v

• ORHO / RMS(MASS) =",lPlG13.3,

• a

( VOLNG

=",Ild,"

)",/,

• -a-G t:0d AL-s RROR-( 0. 95-C ONFI D ENC E P=*vl*1G 13T 5,-

• a ( RMS(AASS)=","3,11X,"(plPlG13.5," GLOBAL RMS(dR40R)

=

=",,lPlG13.5,"

)",/,

• 25X, SUCCES

• " ", "10 Z ( " = " ) )

I ERAMAX ="slP1G13.5,"

}"

/,

- */

UFC A TES TO REMOVE ENERGY CORRECTION mHEN QUALi, QUAL A TRJ4C ATE 3

• 0 HXCAQJ8.32 QUALA(1)

QUALA0(I) + DELXA

=

-*0-H A G R00 tr, +2 r71 C

C TEST FOR QUAL 3, QUALA TRUNCATION (KEEP U(I) SOLUTION)

C

!F-ttt04AettI-PrGT.,-Or0005t rA401--t QU At5 tt t

.t T s-liot1-)

.AD.

~

({QUALA(I)

.GT. -0.00C5).410. (QUALA(I). L T.

1.01))

.ANJ.

((QUALA(I) - OUALS(Ill

.L T.

3.3J13)

.ANO.

(RHOM(I).GT. 0 0)) GO TO 201

<C T d t-tFi

-VC T R L-Pf t. O R.-SHI F T ( RS t ( 1-l e-t2-t-5bCCd5

= 1 2G1 GUALCX

= 0.5

• AHAX1(0.0i (QUALA(I) - QUALS(III)

QUALA(I)

= AMIN 1(1.0, AMAX1(J.0, (20ALA(I)

QUALCY)))

UtJAtStri

= AMIMrt UTC,- AMrXtt0F0,-P3UAt$ (MTCOACCtiti DOTH(I) 00TM(I)

=

+ DMOA(I) * ((QUALA(I)

QUALA0(I)) - DELX4) u

+ JMOX(I) * (QUALS(I)

CUALSO(Ill

-"t IJPROF...to NUME RIC At-AVER AGE-OF JUNCT104-PR-JPER TIES APG VEi.OltTY-

• 0 IJPROP.26 N

IF (VELFJ(I)) 100,200,300 00 IJPROP.31,41

~

PJG COM-ftNUE RHOFJ(I) 0.5 * (RHOF(K) + RHOF(L))

=

VOIDFJti)

= 05* ( VOI DF (K )

• RHOF(<1 + VOIDF(L)

• R40F(LI)

UFJ(I)

= J.i * (AMAX1(1.3E-15, VOIDF(K))

• RHOF(K)

• If F ( < )

e AnxXt(1.05--157 -V3I7C (ti r r RH3F(ts & UC (tri VOIDFJ(I)

VOIDFJ(I) / RHOFJ(I)

=

UFJ(I)

UFJtI) / (RHOFJ(I)

• AM4XI(1.JE-15, VOIDF)(I)))

=

• 0 I J P R3 9. 5 9,76 oc0 CONTrNUE MHOGJ(I) 0.5 * (RHOG(4) + RHOG(L))

=

VOIDGJ(1)

= 03 * ( VOI DG ( K )

• RHJG(<) + VOIDG(L)

• RHOG(L))

UGJtI) 0.5 * (AMAX1(1.0E-15, VOIDG(K))

• RHOG(K)

• UG(<)

=

AWA Il-fts0 E-TST t 307G tr7) &

0.5 * (AMAX1(1.0E-15,'VJIOG(43)

< 80GttF#~U$ U ),

v CC4J(I)

=

• RHGG(K)

• ggA4

+

ANAX1(1.0E-15, V010G(L))

• R iO G( L ) *

..AL)

VOIOGJtI)

VOIDGJ(I) / RHOGJ(I)

=

A bt0G J

= 1.0 / (RHOGJ(I)

• AMAX1(1.0E-15, VOIOGJ(I)))

UGJfI)

UGJ(I)

• RRHOGJ

=

PCAJfI1

= CCAJ(T1

• RRHOGJ A-8

TABLE A-2.

(continued)

C FIA3T TIME TH4 dQ4H.!'.~5 E/ E lX 3(IIIT ' 1E d TIME VELJCITIES

~

00 al I = IJe IJE, IJSKP VELFJtI)

't i L F a % ( I )

=

VELGJ(I) vcLGJX(1)

=

31 CON TINUE GO TO 40 1.

S Em 0 N O-T-I n t-THR 0tsG rt s. -IN FTf A L-Hi--iR R94-I Y-Ftd X-T i R t3 IJe J s JSKP 30 DO 34 I

=

.NOT. AA A 43

. AND. $HIFT( IJ 1(Il e 19)

K =

ER440(K) + MFLOWJ CI)

IF ( VC TRL(K)

.GE. 0 ) ERRH0(K)

=

v3 ). AN 0r-SHIF T tIJ 2 ( Ite-18 )

n-

.N-JTr MASKj'E. 0)

MFL OWJ (I)

ERRH0(K)

IF (vCTRL(K)..

ERRH0(K) =

32 C0h TIN UE

• D VFINL.14

• 0-J-

--ftVF

• I VFINL.21

.NOT. SHIFT (MAS ((lle 17)

.ANO.

IM R EG( I)

IMREG(I)

=

• I VF INL.3 7 1 r-tJR c 00. G Tv-M-G0-TO-1-3 G C

s FIRST TIME THROUGH C

IF-(+VEkF-fkFF)-*-VELFJ FM stTrM).0R.

((VtLFJX(I)

.EQ.

0.0)

.AND.

(VELFJ(I)

.d E. 0.0)).OR.I

(/ELGJX(I)

• VELGJ(I)

.LT.

0.3)

.OR.

((VELGJX(I).EQ. 0.0).AND. (4ELGJ(I)

.NE.

0.011) GO TO 131 GO-TO-tvu 130 CGNTINUE C

SECONO TIME THROUGH

--C IF ((VQLFJKtIl

• VELFJ(I).Gs. 3. ))

.GR.

(VELGJX(I)

• VELGJ(I)

.Gc.

0.))) GO TO 40 SHIF T(M ASK (1), 17)

.34.

IMR EG( )

131 IFREG(I)

=

4G I t TCt-i r-I C YCtE--a 0 R. t$HIF PCIMEttthMI).LTWi 6

• D VFINL.40 g-C REDONOR AND RES T FOR VELOCITY FLIPFLOP IF (. nut. I YCLE) GO TO 3010 IF-t-JReOtt sG TrM-GO-TO-303G C

FIdST TIME THROUGH...REDONOR FLIPFLOP JUNCTIONS CALL JPR0P(2)

C ESTIMATc THE EHROR IN FLUX TERMS DUE TJ FLIPFLOP 1;frMEr-MS KF Oc suei I DMFLud = AJUN(1) * (VOIDFJ(I)

• 440FJ(I)

• VELFJ(I)

+

VOIDGJ(I)

• RHOGJtI)

• VELGJtII)

- MFLOWJtil

.M Tr MA3K043-i. AMDr$HIF it-I-Ittth-t8-i R

ERRdO(K) - OMFLOW IF (VC RL(K)

.GE. 0) ERRHG(K)

=

l K =.NOT. MASK (*3)

.AND. SHIFT (IJ2(I), 18)

ERRH0(K) + DNFLOW IF (VCTRL(K)

.GE. 01 ERRH0(()

=

t r

I YJ01

&CNTINUE C

kcSET cXPLICIT VELOCITIESs SAVE FINAL VEL 3 CITIES IJ, IJEs IJSKP 00 3020 I 2

=

VELFJtI)

SAVEF

=

I SAVEG

^t ELGJtI )

VELFJX(I) l VELFJtI)

=

VELGJX(I)

VELGJ(I)

=

3AVEF VELFJX(I)

=

VE tGJ Xtt )-

-5 AVEG 3040 CON T INUE GO TO 3J40 C

SECJND TIME THROUGH... ESTIMATE ERROR IN FLUX TERMS DUE TO FLIPFLOP 3034-00-39,1

- - t n-M ET-1;tSK1-K =.NOT. MASK (43).AND. SHIFT (IJ1(Ile 18)

ERR 40(K) - MFLOWJ(I)

IF (VCTRL(K)

.Gc. 0) ERRH0(K)

=

x =

.NCT. MAS &(43).AND. SHIFT (IJ2(I), 18)

EARi3(4) + 1FLOWJtI)

IF

(< CTRL (K)

.GE. 0) ERRH0(K)

=

304 CONTINbE C

ESTIMAT: THE MASS ERROR 0

3U40 IMAXF

=

'~N EMAXF

= 0.0

\\

IMAXFG =0 EMAXFG = 0.0 GLOdAL = 0.0 3UnK3S r-070 JVOLa

=0 A-9

TABLE A-2.

(continued) 00 3031 I

= IVe IyE, IVSAP

~

__.j ffgt f yggj',k3 9 E --

OEMASS

= A3S(ERRHQ(I))

• Of DEMASS / V(I)

ERRH0(I)

=

G60dAL

= GLOBAL + OEMASS

• OEM&SS r:1 AyS M t i t-rR H0 (-! )

SUMAS5

$UMASS + XMAS $

• X1 ASS

=

E R A.1

.Le( A RH0 (I )

EMAXF) GO TO(I)

/ RHOF

=

IF (ERRM 3032 I: TAX F =-i EMAXF = ERRM 3032 IF (DEMAs3.LE. EMAXFG) GO TO 3031 I

AMAAFG

=

EMAXFG = VEMASS 3031 CON TINUE SUMASS = 1.0 / SUMASS SCALER 3.8416

• FLOAT (JVOLS) / A1AX3(1, (JVOLS - 1))

=

EM a rt- = ~A.TA~rIC EM A X F~,

~

EMAXFG

• SQRT(FLOAT (JVOLS)

• SUMASS),

SQRT(SCALER

• GLO9AL

• SUMASSI)

IF (EMAXX.LT.

4.0E-3) ICYCLE =. FALSE.

IF-t-( I CYCtE T TOk.--( J R E 00 TG T.-17 t-*10-T 0-3 010

~ C FIRST TIME THROUGH... SMALL ERROR. RESET 0010RINGe FINAL VELO:ITIES CALL JPROP(2) 00 3011 I = IJe IJE, IJSKP VEL 7JL1) 87EtF7ICTI--

vELGJ(I)

= VELGJX(I) 3011 CONTINUE 3010 CONTINUE

+0-0T!TEFT2 2U8kOUTINE O TSTEP(O TLIMT)

• u HYD40.2 SUBROUTINE HYORG(O TLIMT,NRE00)

T*-I MY0rd.TJ LOGICAL ICYCLE

• I HYOR0.51 JnE00

=0

---et-MYORWs32 100 CONTINUE JRt00

= JRE00 + 1

• 0 nYORO.71 C Att vFINtt-TCYCtli, JRE00i IF (.NOT. ICYCLE) GO TO 110 AF (JREDO.EO.

1) GO TO 130 SUCCES

=3 i

f r-00&*--0 T-- s tT s-0 T tI MT)-GO-T3-t11 1

NRE0J = NRE00 + 1 GO TO 1000 lib CONilhuE FT~ H YOROT33 IF (SLCCES.E0. 3) SUCCES = 0

• I TRAN.12 0ATA NRE00/ 0 /

I0 TRAN.17 5

CALL OTSTEP(OTLINT)

• 0 TRAN.44 11 CALL HYOR0(DTLIMTeNRE00)

• I TRAM.60 WRITE (OUTPUT,2005) NRE00 200D F0HNAT (a0***** NUMBER OF TIME STEP RE* EATS ode TO VELOCITY FLIP-F

• LOP ",Il0)

• /

FIX MOVER TO RES ET STATE IF SU :ES

=3

• 0 MOVdd.11

• /

CORRECTIONS TO WALL FRICTION, FORM LOSS AND OISSIPATION I4 VEXPLT

---*0--ttX OR u-15.tir15 AVRF VOIDFA

• RHOFA

=

AVRG VOIDGA

• RHOGA

=

A-10

s TABLE A-2.

(continued)

- -- R A V R T

• - 1. u- / - ( RHO F A -

• A.4 A t 1( 1.1 -15, V O O P A l-1 1.0 /- ( RHOGA

• AMAX1(1.3

-L2,

/0 OGA))

4AVRG

=

C RFVFJ

= 0.5

• VOI3FJ(I)

• R40FJ(I)

= 4gyRp 46V4J-

~J. 3-*-V O I OG J ( 1-) *-RM3 G J E-I-)- *-4 4 V R G

• 0 V:XPLT.131,132

--C kI-Q L43-F*f C-FI G H FWALA(K)

• H;K * /0I3F(4)

• RH3F(K)

• RAVRF FRICFA

=

C FkIC FL e F w A L F ( !; )-+- 0 X L-*--V O I G Ff L )- *- R H O F ( L-1-- 6-4 A V a "

CC ----- V A/ 0R F R IC TI O N C

FR-f CG K--

"FwAt:GtM-)-*-DX K-*-V010G ( +)-*-R HO G ( W)- +-R A V tG FRICGL

= FWALG(L)

• DXL

• VOIOG(L)

• RHOG(L)

• RAVRG C

JUNCT IO4-FkIC T ION

-C FRICFK + FRICFL FRICFJ

=

FRICGJ

= FRICGK + FRICGL

• 0 rrtCR7t3T2a (FJFG + FIJ(I)

• 3X) * (RAVRG + R A VRF )

FJFG

=

• 0 VEXPLT.213

• O HXCd013.32

--*v-ftkPtT v%ti

• 0 HXCRJ13.33 AMAX1(0.0, (VPG{N 4tVRG))

VPGNA 4 *

=

(VPGe

• 4AVRFI)

- AMIN 1(0.0,

• t)-VE1ei-TT&&iT2t2

• O HXCAJ13.3d,39

• 0 VLx/LT.286,291

• 0 nxCAQA3.40

--*D-V E-A PL-i e 4 9 3 HLOSSF

• AVRF

• AJOT

• VEL *J0(I)

HLSF

=

HLOSSG

• AVRG

• AJOT

• VEL GJ0(I)

HLSG

=

DISPK HLSF

• AMIN 1(0.0, VELFJ0(It)

=

O e <i L 3 G-*-A ttiltt( Os 0,-V E t-G J 0 0ItP ---

OISPL HLSF

• AM A X1(0.3, Vk-L5 J3 f t ) )

AMAX1(0.3, V

=

t3J0(I))

+ HLSG *

• 0ELETE,VEXPLf.3w2,VEXPLT.303

.--+&-v etP tTT312

• 0 DMKRGO8.7

'O VEXPLT.314,315

• U DhKRa08.o

--*tt-V t-X P t-TT3 t7 DIS PK + AVOL(K)

• ARAT(I)

• AVK

• OT

~

DISP <

=

• (FRICFA

• AVRF

• VELF10(I)

• VcLFJG(I)

+

FRICGK

• AVRG

• VELGJ3(Il

• V!LGJ0(II)

'AVLt* OT DISPL

= OISPL + AVOL(L)

• ARAT(I+1) *

• (FRICFL

• AVRF

• VELFJ0(Il

• VELFJ0(I)

+

FRICbL

• AVRG

• VELGJ0(I) *. V f L'3 J 0 ( I ) )

• /

PuT PV WORK SEMI-IMPLICITLY IN E3FINL

• 0 HXCR008.6,7 SOURCF(K) 20VRCE(K)

=

00(KX)) + P0(KX) / RHOFJ(I))

• C3HMC

+ ((UFJ(I)

• O HXCR006.13,l',

P0(KX) / RHOGJtI))

• C1NMG

+ ((UGJ(I) - UO(KXI) +

SOURCE (L) 500RCE(L)

=

UO(LX)) + 90(Lr) / RHOFJ(II)

• CONMF

- ((UFJ(I)

- i t UG7 ft)

-UC rt;X t)- *--PG (t-r1-/-NOG-J trr1-*-O'3*tM S s

• /

PUT PV WORK SEMI-IMPLICITLY IN PRESEQ

• 0 PRc5EO. 01,102 TLF PCNF(L-5)

• T1F

=

x,

- *-PC iF( L-$ 1- *- ( TE C-- *- T M F - * - P0( L-T y t-/-tH05t tri i A-11

O TABLE A-2.

(continued)

C TLG P C.*G ( L-5 )

• T*3

=

+ PCEG(L-5) * (TE3 + TMS

• P0 ( L T'/ ) / RHOGJtII)

PC M A ( L-5 )-- *- T M A

=

+

• O PRESEQ.136,137 PCNF(K-5)

• TMF i

IKF

=

g PCdF(K-5) * (TE: + TMC

• P0(<TV) / 4HOFJ(I))

+

TKG P C.9 G (.<-5 )

• TMG

=

+ PC4G(<-5) * (TEG + T.3 3

• P0(ATvl / RHOGJ(II)

=

+ PC.M A ( K-5 )

• T.1 A

- ci Ott RTE-THi-6tPLIC IT-PV-WORK-FROF V E tPt T

• 0 VE XrL T. 318 s 3 du C

A00 OISSIPATION TERMS TO THE S0)RCE 754M 2

C* W xttTT323 SOURCE (KX)

SOURCd(KX) + DISPK

=

• 0 VEXPLT.320 SLORCE(LX) = SOURCE (LX) + DISPL

-- *-/

Get rT E-FHS--5APtIC I T-PV-40RK-FR OM-G O UM

• 0 ONKdO17.299,250 ea een em eee+eym em

.. w 9

i e

O A-12

APPENDIX B QUALITY ASSURANCE PROCEDURE FOR DEVELOPMENT 0." THE RELAPS RESAR-35 SMALL BREAK MODEL I

l l

j l

l i

I B-1

APPENDIX B QUALITY ASSURANCE PROCEDURE FOR DEVELOPMENT OF THE RELAP5 RESAR-3S SMALL BREAK MODEL The following is a quality assurance procedure that was developed, and followed, to assure the accuracy of the RELAPS RESAR-3S small break model.

1.

System Nodalization Diagram--Based on FSAR information and knowledge of the transient to be run, a complete system nodalization diagram is constructed. All components and subsystems required for the calculation are included in the nodalization diagram. This process allows a straightforward determination of the type of data /information required to compile a plant data base (Step 2).

2.

Plant Data Base--A plant data base is compiled to include all the data /information required to develop the plant model. The contents of the data base are of the form of actual plant drawings, technical specifications, operating manuals, FSARs, etc. (or copies of the same), and are limited to first hand sources (if possible). This step allows checking of all data /information back to an original source, rather than relying on second hand information.

The data base also includes a table of contents that uniquely specifies all material contained therein. The table of contents lists all drawings by drawing number and revision number (if any), and all other sources of data /information by title, date, and revision number (if any).

The table of contents is sufficiently detailed to allow duplication of the plant data base by an independent party if required.

3.

Calculation Worksheets--A set of worksheets, which completely document all the calculations required to develop 4/.e input model, is compiled. Data used in a calculation are referenced to a drawing or other source of data listed in the plant data base (Step 2).

Each calculation is written out in sufficient detail B-2

to allow easy checking, and any assumption required in the calculation or any "special method" required to derive a given quantity are clearly indicated.

If a calculation is a revision of a previous calculation, it is so stated on the worksheet, and the reason for the change is included. Both the initial calculation worksheet and the revised calculation worksheet are kept as part of the worksheet package.

4.

Input Deck--The input deck is developed directly from the worksheets compiled in Step 3.

Once the above steps have been completed, the checkout of the system model proceeds as follows:

1.

All data used in the calculation worksheets are checked and varified against the references in the p' ant data base.

2.

All calculations are checked for accuracy and completeness.

ku 3.

Input deck values are checked against the values developed in the worksheets.

Notification that the calculation worksheets has been checked for accuracy is included on each worksheet by affixing the reviewers name and date (i.e., CHECKED BY

, DATE

,). The " checked" status on the worksheet means both the calculations and initial data have been 1

checked. Notification that the input deck has been checked for accuracy is included at the start of the input deck, along with the warning that no changes are to be made which would alter the plant model portion of the input, without first providing the appropriate calculation worksheet and input from revisions, and going through the checkout procedure (listed above) for each revision. By following this procedure, the continued accuracy of the input deck is assured.

O B-3

a,__w a

-.&a

-a-.

6*A

=ma

.sA,-

J

-as%.-

4 -,.

s n

a w-4 0

APPENDIX C FUEL STORED ENERGY CALCULATION E. T. Laats l

i i

I e

e i

C-1

APPENDIX C FUEL STORED ENERGY CALCULATION a

The FRAPCON-2 steady state fuel rod behavior code was used to estimate the initial conditions of the RESAR-35 fuel rods prior to the LOCA events analyzed in this study. First, the hot rod was modeled to operate at constant full power (29.9 kW/m peak power on the hot rod) to determine when during the rod lifetime that maximum centerline temperature and stored energy occurred. That time was found to be 10 days after initial startup, when fuel centerline temperature was about 25 K higher than at beginning of life. Then, the radial temperature profile and stored energy were determined for a typical hot bundle rod, a core average rod, and a rod operating at 90*.' of core average power.

Presented in this Appendix are a brief description of the FRAPCON-2 code, the input to the FRAPCON-2 code used for this analysis, and the results obtained.

1.

FRAPCON-2 DESCRIPTION l

The FRAPCON-2 code calculates steady state thermal and mechanical behavior of light water reactor fuel rods under long-term irradiation conditions.

FRAPCON-2 is a modular code containing isolated subcodes that model fuel temperatures, considering fuel cracking and relocation; fuel and cladding deformation, including elastic and plastic cladding deformation and creep; and rod internal pressure, including fission gas release effects.

Fuel, cladding, and internal gas properties are modeled by a materials properties subcode, MATPRO-11.2 FRAPCON-2 also includes the FRAIL-5 subcode that determines the probability of fuel rod failure.

Input to FRAPCON-2 includes axial nodalization and fuel rod design parameters, which are to be supplied by the user. The rod operating a.

Idaho National Engineering Laboratory Configuration Control No. H0198828.

C-2

/

history, which specifies the system coolant conditions, axial power distributions, and time dependent rod average power, must also be given.

A detailed description of FRAPCON-2 is available in References C-1 and C-2.

2.

FRAPCON-2 INPUT The FRAPCON-2 input deck for the hot rod (with Westinghouse proprietary information deleted) is listed on Table C-1.

The required input to model the rod and coolant channel geometry represent the RESAR-35 17 x 17 rod and bundle configurations. The FRAPCON-2 model options selected were the PELET deformation model and the FASTGRASS fission gas release model. These selections are based on the recommendations in References C-3 and C-4.

The corewide power distributions used in this study represented values reported in the RESAR-35 Safety Analysis Report. The rod axial power

.)

y distribution attained a peak-to-average ratio of 1.19 and a corewide radial peak-to-average ratio of 1.41.

Thus, the peaking factor at the hot location of the core hot rod was 1.41 x 1.19, or 1.678.

For the average rod in the core hot assembly, the radial peak-to-average ratio was assumed to be 1.20, rather than 1.41 as used for the hot rod. The radial power distribution within the fuel pellets was calculated within the FRAPCON-2 code. That power distribution is illustrated in Figure C-1.

To determine the time during operation when maximum stored energy occurred, the power history of the hot rod was divided into two parts.

I First, the rod was ramped to full power (9.13 kW/ft or 29.9 kW/m at the peak power elevation) at the rate of 3 kW/hr. Then, constant full power operation was maintained for 1000 hrs.

It was noted from this calculation that maximum stored energy of the hot rod occurred at 10 days after 4

startup. Then, the three other calculations were performed to represent an average rod in the hot assembly, a core average rod, and a rod operating at 90*.' of core average power. (These three calculations were needed as input to subsequent thermal-hydraulic calculations.) Each of the three calculations was also subjected to the 3 kW/hr startup ramp and subsequent C-3

O TABLE C-1.

FRAPCON-2 INPUT DECK.

e r c:. A 92 777 c1f

- I C,.it, I D =. E T1, 5a c = i 2 c, C H G =.

, e i s =.Tx 9.

R F L,-C M = 3 7 7 0 0 0, E C 12.

ATT AC H, FR PCN 2s F R API, ID = BNWVlM3.

FRPCH2.

O HOT ROD RESAR-35 CALCULATION 5FRPCN IM 20,NA 12,NR 11,NF Si NC

=

4,MECHAH

=li H G AS R=-2s

=

=

=

=

CON COMP 0., CPL

=0.16457,0CI

.008357,CC3

=. 01177,0 EN 0.0095,

=

0E 9 5.0 0, DI SHSC= m D P

.008192i

=

=

s D S P.G_._= N f D S P GW._= N s.E N RC H =

. 2.. c s F.G P A V__=_

m, FLUI

= 3. 9 513r H C IS H =. N iHPLT

=0.01346,1CM

+,

=

IDxGAS=

ls I Pt ANT =

lilo 0,JCLPR 1,

=

=

lliJSI JN 1s NCP T

=

0 N

=

Or.RCUGHC= m sROUGHF= W,_f.bP

=

=

0, N U N I.T.5 =

.E23r T t-

3. 6 5.a3,

=

VS

=

CLUVXSa Ms C R EP HR =

lO.i GRMSIZ=

1G.,

NPRINT=

1,PPMH20=

^C.iPPMNZ =

15.iRSNTR 93.2,

=

T.S.IN T =Ms GO(1) 3532.,

=

TV(1) 565.,

=

P2 (1 )

1. 5 3 + 7 s

=

0 F_( l_)

0 2.7, 0. 9.7.s.1_.13,.1.1.9 1. 1.9.r.l.. ISs.1 18 s.1..L4 s.1.10s.O. 91s.0. 22s

=

0.

0. 3 6 6,0. 7 3 2,1 0 9I3 s 1. 4 6 3 r l. 8 2 9,2 195,2 5 6,2 9 3, X(11

=

3.27,3.6585, OM PY (1) =

0 1,3.0,6.0

.21.. A ls.2h. 0 s.# 9, C,12. 0,15. 0,16. 0 6,17. S 4,15. 0, 21 0, a*25 16s-T I M E ( 1) = 0 01, c.0 4 2i o.0 e 3 r G.12 5 s o.16 7,0. 2 0 S i o. 22 3 r c. 24 3, 0.25,0.292,0.297sC.333,0.349,1.,5.ilo.r15.i20.r25.s3C.s

~

s O

C-4

1 i

i l

I i

i LM i

i a

i i

i i

i I

i i

t.03 f

1.02 i

O 3

l 0

b LO1 i

6 i

e 8, wo o

ao i

=

1

~l ose e

C i

s l

x 028 i

e o

t 1

I OD7 i

t osa og 0.0 0.1 02 0.3 0.4 0.5 0.S 0.7 03 02 10 Fractional radius k

Figure C-1.

Radial power distribution in the fuel pellets.

j I

i C-5

constant power operation to 10 days. The radial temperature distribution noted at the end of the 10 day irradiation, was subsequently used to initialize the thermal-hydraulic calculations.

3.

RESULTS The resula.s of the four FRAPCON-2 calculations (hot rod, hot assembly, core average, 90% of core average) are summarized in Figure C-2.

Shown are four curves representing the radial temperature profile for each case, at the rod hot spot.

These profiles were obtained at 10 days after startup.

The fuel centerline temperature of each curve shown in Figure C-2, is plotted in Figure C-3 against local power.

Since maximum stored energy occurred at 10 days, no significant effects of long term irradiation were noted, such as fission gas release, cladding creepdown, and fuel densification. Thus, the boundary conditions and general state of the fuel rods, as subsequently modeled by the thermal-hydraulic codes, reflect fresh fuel rods. The only exception is decay heat, which was assumed to be 91% of the heat generated if the ANS 73 model was used.

O C-6

i 1

I i

1

?

a i

1 i

I j

1800 8

i, 9

rioo i

l 1800 Hot rod Hot bundle

-. - Core average rod l

1500 907. of core average n

i x

- 1400 1

e

~

a a

,~~,

Y 1300 N

I; s

z I

c N

tm c2 N-s N

$ 1100

\\

\\

x

-x tooo x,,

x l

\\, ',

l 900

\\ 's, s

\\k.,

eoo X::...

J N,O 0.00 0.05 c10 0.15 020 025 0.30 0.35 0.40 0.45 0.50 Radius (cm) e Figure C-2.

Radial temperature profiles frcm FRAPCON-2 calculations.

C-7 l

,,-m, g

-.w

.e,,ca.

---

,,re=

= ~. - - - -

- - - - ~ - - - - - - -

O ND O

HOT ROD A

HOT BUNDLE UN O

CORE AVERAGE ROD x

90% OF CORE AVERAGE 2

3 1800.0 2

c y:

1500.0

?.:

y 1400D a

5 O

~

w00.0 z

O E

h 1200.0 -

+

.=

xe2 1100.0 i

1000.0 14 0 itto 18.0 20.0 220 24.0 26.0 28.0 30.0 32.0 34.0 38.0 Power ( kW/m)

Figure C-3.

Fuel centerline temperatures from FRAPCON-2 calculations a

vs. local power.

O C-8

REFERENCES C-1. G. A. Berna et al., FRAPCON-2: A Computer Code for the Calculation of Steady State Thermal-Mechanical Benavior of Oxide Fuel Rods, NUREG/CR-1845, December 1980.

C-2. D. L. Hagrman et al., MATPRO-Version 11 (Revision 1): A Handbook of a

Materials Properties for Use in the Analysis of Light Water Reactor Fuel Behavior, NUREG/CR-0497, TREE-1280, Rev.1, February 1980.

C-3. E. T. Laats et al., Independent Assessment of the Steady State Fuel Rod Analysis Code FRAPCON-2, EGG-CAAP-5335, January 1981.

C-4. G. A. Berna et al., FRACPON-2 Development Assessment, NUREG/CR-1949, PNL-3849, July 1981.

O 4

C-9