ML20151Y916

Research Info Ltr 155 Re full-scale Fluid Mixing Test Results in Support of PTS Resolution.W/O Stated Encl

ML20151Y916
Person / Time
Issue date:	08/24/1988
From:	Beckjord E NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	Murley T Office of Nuclear Reactor Regulation
References
RIL-155, NUDOCS 8808290087
Download: ML20151Y916 (5)

Text

v - T 9-

• gnnicy

[o

'g UNITED STATES

[(

g g

- t NUCLEAR REGULATORY COMMISSION WASHING TON, D. C. 20555

\...../ -

AUG 2t1998 MEMORANDUM FOR: Thomas E. Hurley, Director Office of Nuclear Reactor Regulation FROM: Eric S. Beckjord, Director Office of Nuclear Regulatory Aesearch

SUBJECT:

RESEARCH INF0FEATION LETTER NO.155 , FULL SCALE FLUID MIXING TEST RFSULTS IN SUPPURT OF PRESSURIZED THERMAL SHOCK RESCLUTION This Research Information Letter transmits the results of fluid mixing tests conducted in the full scale Upper Plenum Test Facility (UPTF) located in Mannheim, Federal Republic of Germany. Several experiments were conducted to study mixing between cold Emergency Core Coolant (ECC) and hot primary coolant in the UPTF as part of the 20/30 International Loss-of-Coolant Accident (LOCA)

Research Program. Test reports prepared by Siemens AG are enclosed as Attachments Al and A2. The UPTF tests and other small scale fluid mixing tests were analyzed by Professor T. Theofanous, University of California, Santa Barbara, CA. The analysis results were reported in several journal articles and reports. Four articles are attached (Attachment B1 through B4) for more detailed information.

1. Regulatory Issue Reactor vessels are subject to embrittlement from fast neutron fluence.

Should an embrittled vessel be subjected to high thermal stress in combinatior with high pressure, an existing flaw in the vessel wall may develop 1%o a crack from the combination of high tensile stress and low fracture toughness. This concern, known as Pressurized Thermal Shock (PTS), was investigated under Task A-49.

There are two predominant causes of high thermal stress due to overcooling of the vessel wall; steam line rupture and the High Pressure Injection (HPI) of cold Emergency Core Coolant (ECC) into possibly stagnant cold legs following a small break LOCA. In this Research Information Letter (RIL) we address only the latter. Since the most susceptible part of the vessel is the welded part in the belt line region of the vessel, our attention is directed primarily to the degree of thermal mixing in the downcomer below the cold legs.

The magnitude of thermal stress is determined by the degree of mixing between the cold HPI fluid and the hot primary coolant. The most severe 7.:se would be for complete stratification to occur between the HPl fluid and the primary coolant. Realistic mixing must be determined to avoid artificallt conservative regulatory actions.

8800290007 000024 N002%$g7 PDR c _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

e 1

Thomas Murley g NRR requested RES assistance in evaluating fluid mixing as well as other issues as part of Task A-49 (see Attachment C). RES provided sufficient ,

fluid mixing information from a series of small scale tests to help '

resolve A-49. However, to confirm the uncertainty in extrapolating small scale test results to a full scale PWR, representative tests in a full scale test facility were arranged. Such experiments were performed in UPTF and are reported here. ,

2. Conclusion Experiments conducted at test facilities ranging from 1/5 to full scale show that HP1 fluid mixes well with primary coolant. Cold ECC warms quickly from mixing occurring primarily in three regions; (1) HPI location; (2) entrance of the cold leg to the vessel; and (3) downcomer region below the cold leg. Primary coolant flows continuously to the i mixing regions frum the lower downcomer and the lower plenum by natural circulation due to density differences between the cold HPI fluid and the hot primary coolant. The two fluid streams are separated from each other because of a density difference and develop a countercurrent flow pattern.

Therefore, the hot primary coolant in the lower part of the downcomer and the lower plenum is able to re6ch the mixing locations in a continuous manner. This process results in two beneficial effects; the reduced temperature gradient between hot and cold stream and the reduce'd cool-down rate of the vessel wall. This was illustrated by a UPTF test employing typical PWR HPI (10 kg/sec). The temperature difference between vessel wall and the midplane of the downcomer at the elevation corresponding to the top of the core was less than 10'K, and the cooldown rate was less '

than35'K/hr(63*F/hr). This compares to the maximum cooldown rate of '

56*K/hr (100'F/hr) typically specified in technical speqifications. The full scale UPTF experiments covered an HP1 rate of 5 to 70 kg/sec. In all cases, good mixing occurred. As expected, the temperature difference ,

between the wall and the downcomer center increases along with the  ;

cooldown rate as the injection rate increases, i

I The special purpose computer codes REMIX and NEWHIX, developed by Professor Theofanous, predicted the temperature response in the cold leg

, and downcomer reasonably well. REHlX is used for Westinghouse and Combustion Engineering designs while NEWMIX is used for Babcock and Wilcox designs which employ HP! with high Froude numbers. System codes such as 1'

TRAC and RELAP are not applicable to liquid-liquid fluid mixing analysis because they are not designed to handle two liquid streams, only a vapor and a liquid field (two field equations).

1 3. Regulatory implications full scale UPTF test results are consistent with earlier small scale test results used in resolving Task A-49, and therefore support the resolution previously made under Task A-49.

I i

~

Thomas Murley AUG 2 419g3

4. Restriction on Application The UPTF is a low pressure facility so the initial primary coolant temperature could not be raised to the usual PWR operating temperature, 550'F(228'C). The initial temperature used in the UPTF experirrents was 374*F (190*C). However, the thermal stress and the couldown rate of a typical PWR downcomer wall can be calculated with the codes REli!X and NEWMIX.

5. Unresolved Questions There is no unresolved question concerning the fluid mixing phenomenon <

related to the PTS concern. The computer codes REMIX and NEWillX have been verified with the full scale UPTF test results. Therefore, these codes can be used to estimate the degree of thermal stratification between the HPI fluid and the primary coolant for a given transient.

n Eric S. Beckjord, Director Office of Nuclear Regulatory Research Attachments:

A1. 20/3D Program Upper Plenum Test Facility Test t o.1 Fluid-Fluid f11xing Test Quick Look Report A2. 20/3D Program Upper Plenum Test Facility Test No.1 Fluid-Fluid Mixing Test Experimental Data Report Bl. Scaling of Thermal liixing Phenomena B2. Ilixing of Phenomena of Interett to SBLOCA's B3. Decay of Buoyancy Driven

Stratified Layers With Applicants to PTS

Reactor Predictions B4. PWR Downcomer Fluid Temper-ature Transients Due to High Pressure Injection at Stagnated Low Flow C. Task Action Plan, Pres-

, surized Thermal Shock (TaskA-49) l i

Thomas Murley . AUG 2 4 msg

4. Restriction on Application The UPTF is a low pressure facility so the initial primary coolant temperature could not be raised to the usual PWR operating temperature, 550'F 228'C . The initial temperature used in the UPTF experirents was 374'F 190'C . However, the thermal stress and the cooldown rate of a typical PWR downcomer wall can be calculated with the codes RElllX and NEWi1IX.

5. Unresolved Questions There is no unresolved question concerning the fluid mixing phenomenon related to the PTS concern. The computer codes REMIX and NEWMIX have

• been verified with the full scale UPTF test results. Therefore, these codes can be used to estimate the degree of thermal stratification between the HPI fluid and the primary coolant for a given transient.

OrigJ nal 5fgs,4 g D e P,s eds f S. Beckjord, Director Office of Nuclear Regulatory Research Attachments:

A1. 20/30 Program Upper Plenum Test Facility Test No. 1 Fluid-Fluid Mixing Test Quick Look Report A2. 20/3D Program Upper Plenum Test facility Test  !-

No.1 Fluid-Fluid flixing - - ,,D 2 '

Test Experimental Data Report ,' w. n u ::o. k- 2.000.02.

Bl. Scaling of Thermal Mixing R. [,f{-

Phenomena B2. Mixing of Phenomena of '.

~~

Interest to SBLOCA's l

.o . 33.

B3. Decay of Buoyancy Driven -

Stratified Layers With

- ~ ~ ~

Applicants to PTS: -

s.

~ " - -

Reactor Predictions ,

w : o. ~ " ~ ~ ~ ~ ~

l B4. PWR Downcomer Fluid Temper- C- - Mc.gg ature Transients Due to a'"n: tm High Pressure Injection ' t E :, 'wa _ng at Stagnated Low flow =

C. Task Action Plan, Pres-surized Thermal Shock l (TaskA-49)

Distribution: Circ; chron; drps chron; RPSB r/f; GRhee r/f: GRhee; NZuber; DBetsette; LShotkin; JMurphy; BSheren; Dross; EBeckjord SEE PREVIOUS CONCURRENCE

• p / '

^

RPSB RPSB RPSB RPSB h . D:RE D:RE5ib I GRhee*/cb NZuber* DBessette* LShotkin phy EBeckjor9 8/9/88 8/10/88 8/10/88tr//g68 /g8 er

/88 f D >8 $/ir/88 u o  :

Thomas Murley '

AUG 2 41999

4. Restriction on Application /

The UPTF is a low pressure facility so the initial primary coolant v temperature could not be raised to the usual PWR operating temperature, 550*F 228'C , The initial temperature used in the UPTF experiments was

'374*F 190'C . However, the thermal stress _and the cooldown rate of a typical PWR downcomer wall can be calculated with the codes REMIX and NEWHIX. ,

5. Unresolved Questions ,

There is no unresolved question concerning the fluid mixing phenomenon related to the PTS concern. The computer codes REMIX and NEWMIX have been verified with the full scale UPTF test results. Therefore, these codes can be used to estimate the degree of thermal stratification between the HPI fluid and the primary coolant for a given transient.

E. S. Beckjord, Director Office of Nuclear Regulatory Research

Enclosures:

A1. 2D/3D Program Upper Plenum Test Facility Test No.1 Fluid-Fluid Mixing Test Quick Look Report A2. 2D/3D Program Upper Plenum Test Facility Test No.1 Fluid-Fluid Mixing Test Experimental Data Report Bl. Scaling of Thermal Mixing Phenomena B2. Mixing of Phenomena of Interest to SBLOCA's B3, Decay of Buoyancy Driven Stratified Layers With Applicants to PTS: Reactor Predictions B4. PWR Downcomer fluid Temperature Transients Due to High Pressure Injection at Stagnated Low Flow C. Task Action Plan, Pressurized Thermal Shoc,k (Yask A-49)

\

Distribution: Cire; chron; drps chron; RPSB r/f; GRhee; DDessette; LShotkin; JMurphy; BSheron; Dross; EBeckjord f L') /v pg L5 RPSB~ s RPSB y f RPSU RPSB DD:DSR D:DSR DD:RES D:RES GRhee/ NZuber DBessette LShotkin JHurphy BSheron Moss EBeckjord 8'/ct /88 i O/08 Y//v88 tr//d/88 / /88 / /88 / /88 / /88 w - - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ - _ _ _ - _ _ - _