RBG-21771, Forwards Supplementary Info & Clarifications to Assist in Evaluation of Adequacy of Hydrogen Control Sys,Including Suppl to 850701 Preliminary Equipment Survivability Rept & Sequence of Events for stuck-open Relief Valve

Forwards Supplementary Info & Clarifications to Assist in Evaluation of Adequacy of Hydrogen Control Sys,Including Suppl to 850701 Preliminary Equipment Survivability Rept & Sequence of Events for stuck-open Relief Valve

ML20134G813
Person / Time
Site:	River Bend
Issue date:	08/05/1985
From:	Booker J GULF STATES UTILITIES CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
RBG-21771, NUDOCS 8508280101
Download: ML20134G813 (26)

Text

,_ . _ _ _

.*. 4 e

GULF STATES UTELETIES COMPANY POSTOFFICEBOX2959 . BEAUMONT. TEXAS 77704 AREACODE409 838 8631 August 5, 1985 RBG - 21771 File No. G9.5 Mr. Harold R. Denton Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Mr. Denton:

River Bend Station - Unit 1 Docket No. 50-458 As requested by the NRC staff, Gulf States Utilities (GSU) is providing supplementary information and clarifications to' assist in your evaluation of the adequacy of the Hydrogen Control System at the River Bend Station. This l

supplementary information consists of three attachments. Attachment 1 provides

! a supplement to the preliminary equipment survivability report previously submitted on July 1, 1985. Attachment 2 provides a sequence of events for the stuck-open relief valve and drywell break base cases. Attachment 3 provides a supplement to the report on hydrogen deflagration pressure effects on equipment submitted on July 5, 1985.

Sincerely,

. f.

• 7 J. E. Booker l Manager - Engineering

[ Licensing & Nuclear Fuels River Bend Nuclear Group EJZ/ s r

8508280101 850005 fh PDR ADOCK 05000458 i j A PDR

Attachment I RIVER BEND STATICN PRELIMINARY EQUIPMENP SURVIVABILITY REPORT

^

SUPPLEMENP FOR DEE7JGRATION BURNING IN TfE INPERMEDIATE VOLINE

~. .

r TABLE OF 00tmWIS l

1 Secticri Title Page 1

1. 0'. INTBODUCTION/SIANARY 1-1 l 2.0 MODEL DESCRIPTION 1-1 3.0 REi ITS/CONCWSIONS 1-4 REFERENCES 1-5 l

i s

I l

i i.

i l

L.

?

1.0 INTRODUCTION

/SU M ARY t

In the previously subnitted CLASIX-3 analysis of the pressure and tanperature response of the River Bend Station follwing a hydrogen generation event (Reference 1) , no hydrogen burns were predicted to occur in the containment intermediate volume (above the HCU floor) during the hydrogen release period. However, a single burn was forced to occur concurrently in the wetwell, intermediate volume, and upper containment volune in order to deplete the ranaining hydrogen inventory after the hydrogen release was canpleted. The thermal profiles predicted by CLASIX-3 for the wetwell and intermediate volune were used to determine equipnent response to deflagration burns. We thermal environnent predicted for the intermediate volume was used to assess the survivability of equipnent located in this volume.

The predicted intermediate volume thermal envirorrnent predicted by CLASIX-3 is consistent with the design of the hydrogen ignition system which is based on maximizing hydrogen burning at lw concentration which will occur prbnarily belm the HCU floor. However, it may be possible to postulate that a small nunber of burns will occur above the HCU floor. Therefore, the objective of this study is to determine the thermal response of various pieces of essential equignent located in the intermediate volume, assuming that it is exposed directly to the burn envirornnent predicted for the wetwell region belm the HCU floor.

B is report considers the Target Rock solenoid and small Reliance motor previously analyzed (Reference 2) and shown to survive the intermediate volume thermal environment. In addition, two pieces of essential equipnent not previously analyzed have been included. The additional models included are a Rosemount pressure transmitter and a typical power cable.

The HEATING 6 thermal analysis code models of this equipnent are described in Section 2.0. For each piece of equipnent, the most thermally sensitive nonmetallic canponents and associated thermal limits were identified. The number of wetwell hydrogen burn spikes which each piece of equipnent could sustain before sensitive canponent tanperatures exceeded their limits was determined. The results of the study are sunnarized in Section 3.0.

2.0 MODEL DESCRIPTION In this study, the tanperature function applied to the outside of the equipnent is the CLASIX-3 predicted wetwell tanperature shwn on Figure

1. As indicated on Figure 1, a series of burns occur in rapid succession between 3,000 and 3,500 sec after accident initiation. These burns are the direct result of the high hydrogen release rate of up to 5.125 lbn/sec used in the CLASIX-3 analysis. Hydrogen release rates of i

1-1 l

this magnitude will not result in deflagration burning in the intermediate volume region. For this high hydrogen release rate, the hydrogen concentration in the wetwell volume will reach an ignitible concentration long before significant hydrogen accumulation occurs in the intermediate volume. This will result in a single deflagration in the wetwell volume followed by steady diffusion burning anchored at the suppression pool. All quarter scale test results to date have confirmed this sequence of events. Therefore, the series of burns corresponding to the period of high hydrogen release rate were not included in this evaluation. The thermal envirenment imposed on equipnent in this study is the series of hydrogen burns beginning at 5,000 sec. This series results fran the extended hydrogen release at the rate of 0.1 lbn/sec starting at 3,645 sec. It should be noted that the thermal environment being imposed on the equipnent in the intermediate node is based upon the release of hydrogen into the small wetwell volume with burns occurring when the hydrogen concentration reaches 8 volume percent. If the release of hydrogen was directly into the larger intermediate volume, the time required for the hydrogen concentration to reach 8 volume percent would be longer, and the resulting thennal profile would consist of fewer total burns with a lower frequency of occurrence. This will allow more time for the equipnent in this volume to cool between burns. In addition, the use of the wetwell thermal environment to evaluate equipnent in the intenmdiate volume does not allow credit for cooling provided by the containment unit coolers. The CLNSIX-3 analysis assumed that cooling was only provided to the intermediate volume.

Therefore using the wetwell thermal environment for evaluating equipnent in the internodiate volume effectively disalloas credit for the containment unit coolers.

The IIFATING6 thermal models of the Target Rock solenoid and the small Reliance motor used on the Limitorque valve operator have been prcniously subnitted (Reference 2) . Figures 2 through 5 (repeated frcm Reference 2) provide detail 9 cf the equipent and corresponding HFATING6 models.

'Iwo additional models have been developed to represent other ituas located in the intennediate volune. A Rosanount Model'1152 level transmitter, as shown in Figure 6, has bcen modeled in two parts. "ihe thennal model of the electronic housing is shown on Figure 7. The epcncy glass laminate electronic circuit board is considered a sensitive couponent inside the nickel / iron alloy, cylindrical housing. The header board lead wires which contact the inside surface of the electronic housing may also be susceptible to thennal degradation but are not explicitly modeled in this study. The model of the sensing module is shown on Figure 8. The limiting canponent for the sensing module is the silicone oil on either side of the steel diaphragm. A 600-V, three-conductor Okonite power cable has also been modeled in this study.

The critical cable canponent is considered to be the electrical insulation of ethylene propylene rubber which covers the copper conductor. Figures 9 and 10 show the cable construction and the associated thermal model.

l l-2

The asstmptions used in this study are similar to those described in Reference 2 and are sumarized as follows:

1. One-dimensional (radial) modeling has been applied for the power cable. Two-dimensional modeling is assumed for the other equipnent.

2. Where unit orientation with respect to vertical has not been verified, it is assumed that the unit is so oriented as to maximize convection to the most critical cmponent.

3. Units being analyzed are assumed to be surrounded on exposed sides by hot vapor to a sufficient distance (of at least 10 ft) so as to maximize emissivity of the radiant cloud.

4. Convection of heat to the units is modeled by asstuning forced convection at a velocity of 12 ft per sec.

5. anissivity and absorptivity of the equignent cmponent surfaces (internal and external) are set equal to conservatively high values, so as to maximize heat transfer to the equignent surface and within air spaces located inside the equignent auter surface.

6. Natural convection within free air spaces inside the equipnent surface envelope is modeled by using enhanced heat conduction (Reference 3) for the Reliance motor and Target Rock solenoid. A natural convection model is used for the Rosemount level transmitter. These two methods yield cmparable results for similar boundary conditions. However, the natural convection model gives a variable convection coefficient as a function of the surface-to-surface tmperature difference.

7. Critical unit nornnetallic subemponents are determined by review of vendor data and equipnent qualification reports. When two ,

subecmponents have similar projected tmperature sensitivity, alternate heat transfer models are developed to maximize heat flow to each potentially critical cmponent. Each nodel is then subjected to the thermal forcing function for the unit being analyzed. This procedure avoids spurious qualification on the basis of the most thermally sensitive material occurring only in a well-protected location (well insulated, attached to most massive cmponent heat sink, etc.), while anothe/ material, with less inherent sensitivity to high temperature, may occur in a more exposed environment and thus be heated rapidly above its critical tenperature.

All units are assumed to be maximally exposed to the elevated thermal envirorsnent as indicated on the figures supplied. Exposed surfaces are allowed to radiate to the heat sinks which have surface tmperatures ' calculated by CLASIX-3. No credit is taken for any unit being in a convective " dead zone" and thus shielded frcxn the assumed 12-ft-per-sec gas stream. Unit internal air spaces (frcxn 1-3

the shell or case to the heat-sensitive cmponent which is shielded by the' shell or case) are assumed to transmit heat by natural convection,' by conduction, and by radiation.

3.0 RESULTS/ CONCLUSIONS

}

The results of this study are sunnarized in Table 1 which shows the nmber of wetwell deflagration tm perature spikes that each piece of

} equipnent is capable of sustaining, without exceeding the sensitive

4 cmponent taperature limit. %e Ros mount transmitter is predicted to survive six burns on the basis that the amplifier circuit board tenperature exceeds the manufacturer's qualification tmperature.

f The balance of the equipnent analyzed should survive 12 or more of the f 39 wetwell burn spikes, which represents 31 percent of the extended series of wetwell burns. Although none of the equipnent is deonstrated e ~to survive the entire series of wetwell deflagrations, it should be le recognized that the wetwell burn environment is characterized by more numerous and frequent burns due to the small wetwell volume than would be predicted by CLASIX-3 for the larger intermediate volume. Also, since the wetwell thermal profile has been applied directly to the intermediate volme equipnent, no credit for cooling by the contaiment unit coolers is taken.-

%e results of this study reflect other conservatisms worthy of note.

'No credit has been taken in this study for reduced, radiant or convective heat transfer due to local obstructions such as walls, floors, or other equipnent. Heat transfer is maximized to the sensitive

' internal cmponents by assming perfect contact between layers of materials (e.g., cable insulation and outer jacket) . EPR, used as the insulation on the Okonite power cablg, is rated for continuous use in seal applications for 2.0 hr at 440~F. In this analysis, the cable was assumed to fail the first time that the EPR temperature reached 440 F, which occurred after the seventeenth burn. The cumulative exposure of

., the cable to high tmperature is noch less than 2 hrs, and it is

• expected that cables would survive more than 17 burns.

Since the essent;ial equipnent is capable of surviving a mininum of six wetwell type burns, this equignent should be capable of surviving the small number of burns which might occur in the intermediate volume. For the SORV base case, all hydrogen enters the wetwell volume through the suppression pool. Since there is always sufficient oxygen to support embustion in the wetwell, even at the end of the transient, any accmulation of hydrogen in the wetwell will burn when it reaches the ignition criteria. W erefore, virtually all of the burning will occur in the wetwell, volume. Buildup of the hydrogen concentration in the intermediate volume'would be so slow that deflagration burns in the int - v11 ate. volume, if any, would be widely spaced. We long relaxation ~_ time between burns would also allow cooling by the contaiment unit coolers which would further reduce the effect of any burns in the intermediate volume.

1-4

REFEREN3lS

1. GSU Ietter No. RBG-21,218 dated June 7,1985, fmn J. E. Booker to H. R. Denton, Contairrnent Pressure and Tenperature Response to Hyw, .@i Canbustion

2. GSU Ietter No. RBG-21,423 dated July 1,1985', fran J. E. Booker to H. R. Denton, Preliminary Equipnent Survivability Report

3. J. P. Holman, Heat Transfer, New York (McGraw-Hill,1976) , pp 225 to 259 1-5

m -.

TABLE 1 SLMEW OF RESULTS Number of Burns Sensitive Temperature Prior to Exceeding Equignent Cwww.nt(s) - Limit Ternperature Limit Reliance motor Stator coil 340 F 12 (0.13 hp) insulation Target Rock Rectifier 385 F 17 solenoid Rosenount Anplifier/ 303 F 6 transmitter electronics Silicone oil 400 F 29 Okonite power EPR insulation 440 F 17 cable NOTES:

1 Qualification tenperature 2 Maximum continuous service temperature 3 Maxi == continuous service tarperature for exposure time of 2.0 hr 1-6

h FIGURE 1 2200 - l 2100 .

2000 .

1000 .

1000 .

1700 -

1000 -

l 1500 -

@ 1400 -

W 1300 -

$ 1200 -

1100 -

w 1000 -

g 900 -

. .00 -

1. 700 -

400 -

500 -

400 .

300 l l b (13 \.\M

( ( ((,(( ((( A (1 ( ( ( ( ll(( k A l k

200 -

15

( '

100 0 12 14 16 18 20 2 4 6 8 10 0

TIME (THOUSAND SECONDS)

WETWELL TEMPERATURES SORV CASE - RELEASE B RIVER BEND STATION

PIGURE 2

~

MOTOR COIL

'. o

,- s y

g r .

x . ,

_ -. 3

/- , I

'l

- J-Q

( N,; . ,

$h ' '

d

% )

dhh 'r ., ,

d q I~

\

LIMITOROUE VALVE OPERATOR MOTOR RIVER BEND STATION

FIGURE 3 LIMITORQUE VALVE OPERATOR 0.13-HP RELIANCE MOTOR TEMPERATURE FUNCTION FROM CLASIX.3 (RADIATION CoNVECTioNI ll l / / / ,"'/ / / / / / / / / / / / / / / / / / / / /Y INSULATED f '

\ \.

\ '\ \ ' ,

sOUNDARY e' AIR {,  ; ,

\' '

/, x g' N3,gs

/*

/

m g .g

(*cs,d! !! x

\

,N INSULA ,N s

N s N\ \. \ 'N \

s. xxN N

/

f..

~ n n ; TION)

e [ s. N x s ss c,

x ,\ 's sN, lb 1l(R ADI ATION. lb .

/ l>

' if AIR

'/ < ' CONDUCTION) ffI

,\ N O N 'N N.N s,\ \ ' , x soVND R

/ , iron . 'N's , 'N['

y _

7 ///,sT' EEL (([

/ ,

\ \-\\g \ g \ \\\ s s'.

q, Ng\ {wg 'g u

The above shows a two-dimensional model (cylindrical coordinates) of the stator coil (mixture of copper and insulation), which is taken as the critical component. An insulated boundary is used to separate the model from other elements assumed to follow the same temperature transient (typical).

FIGURE 4

~~

RECTIFIER h

50*

REF- 37 NUMBERS CORRESPOND TO THE MANUFACTURERS PARTS BREAKDOWN

\ .

s 35

'. 34 52 33

'a .

g .

39 41 51 50 40 42

• 54 43 32 l

55 56 57 , 31 50 44

\

N ~ .,.[

\

\ '

6.

' h*: Y' <N ,y w4

• • $+ 5 .

\  !

J

.40 APPROX CG+

• M

4. =

,} f

\T N

/

TARGET ROCK SOLENOID VALVE RIVER BEND STATION

FIGURE 5 TARGET ROCK SOLEN 0ID VALVE TEMPERATURE FUNCTION FROM CLASIX-3 (RADIATION, CONVECTION)

// /

R ADIATION, CONDUCTION INSULATED AIR '

BOUNDARY x/ _

W RADIATION, W SILICON CONDUCTION The above shows a two-dimensional model (cylindrical coordinates) of tlie rectifier block, which is considered to be the critical component. The body of the valve is not considered as a heat sink as it is also sub-jected to the thermal transient.

1 FIGURE 6 i

l l

\

e MODEL-ELECTRONIC N HOUSING IT 7

'4 . .

o (( . . -

~~

.,N ,

8

(

7 ay '% , ,

~, . .e

. g  %

g 4

=

• s. t ,

o ' c N y Q \. 7 l

.. s maa MODEL-t

'@ 11 cHuna SENSING MODULE d

• e e PRESSURE / LEVEL TRANSMITTER

• FIGURE 7 6

7 s. c.1 )

\ \,

i B. C. 2 l

ks. c.1  ;

Y '

m g

(* c-l l

\l

\c  ? , .

~ lNSUL TEDDYP)

MATERIAL 50UNDARY CONDITIONS 1

1. b LOW COPPER ALUMINUM 1. FORCED CONVECTION

& RADI ATION

2. b NICKEL /lRON ALLOY

2. FREE CONVECTION l

! 3. E SOLDERING MATERI AL & RADI ATION

4. E ETHYLENE PROPYLENE RUSSER (EPR)

5. 0 EP0XY GLASS LAMINATE l

l THERMAL MODEL OF PRESSURE / LEVEL TRANSMITTER (ELECTRONIC HOUSING) (ES-002-0) l l

- - _ ---..-r- - - - - - _ , , - , - . - - .

FIGURE 8 1

Z

^

INSULATED (TYP)

,7 9

m R

MATERIALS BOUNDARY CONDITIONS

1. 316 STAINLESS STEEL 1. FORCED CONVECTION

& RADI ATION

2. U SPRING STEEL

~

3. SILICONE DIL THERMAL MODEL OF PRESSURE / LEVEL TRANSMITTER (SENSING MODULE) (ES-002-0)

. . . . . . - . . - - . - _ __- .,...,,,,,_,,,.-,..-,%, _ . . , . _ - - - , , - - - , _ . , - , . , , _ _ , , , _ . , , , , , . . . ._,,._.4, , - , , -

. FIGURE 9 i

C W '#

gotttt JACKEI gtopt JACKET p ATION UY) i C A8LE (TYP)

CABLE TRAY CONSTRUCTION OF POWER CABLE _

t l

l l

FIGURE 10 o

= l B. C.1 R

WSULATED WITIAL AR G AP f

MATERIAL BOUNDARY CONDITIONS

1. COPPER

1. FORCED CONVECTION

2. M CHLOR 05ULFONATED & RADI ATION POLVEHYLENE (TRADE NAME HYPALON)

3. 1 ETHYLENE PROPYLENE RUSSER (EPR) l .

• l l

I I I

POWER CABLE THERMAL MODEL (ES,-0.01 -0) l 4

Attachment II SEQUENCE OF EVENTS FOR SORV AND DWB BASE CASES

SEQUENCE OF EVENTS FOR SORV BASE CASE Time (sec) Event 0.0- Relief valve stuck open coincident with loss of all EOCS 240 ADS manually open at top of active fuel 409 Core uncovery

-2000 Core 3/4 uncovered 3400 Reflood initiated (5000 gpn) 3645 Reflood cmplete 19832 Hydrogen production ceases k

2-1

Sequence of Events For DWB Base Case Time (sec) Event 0.0 Drywell break (break area equal to 1 SRV) coincident with loss of all ECCS 1077 Auto ADS open 1283 Core uncovered 2300 Reflood initiated (50'00 gpn)

Suppression pool drawdown (for revised DWB) 3645- Reflood cmplete 21569 Hydrogen production ceases 2- 2 i

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

INrRODUCTION 4

a 4

In the previously subnitted report on equipnent pressure survivability,

• Hysurn Deflagration' Pressure Effects on Equipnent" dated July 5, 1985, the essential equipnent was shown to survive the pressures prrvW by hydrogen burns during hydrogen release. During a meeting with the NRC staff held on July 24, 1985, Gulf States Utilities was p requested to evaluate the effect of the burn which was forced to occur

[ after the h A upn release was terminated.

. RESULT t

%e pressure peak resulting frm the. forced, sintitaneous wetwell, intermediate and contaiment burn was 35 psig. All essential equipnent

-has been evaluated to establish its capability to vithstand a 35 psig static pressure, which may result fr m a hy& wesi deflagration burn.

Pressures for which the equipnent has been qualified are given in Table

1. As shown on this table, all equipnent listed, with the exception of the containment unit coolers and the hWur.n -mixing systs fans, is qualified..for pressures greater than or equal to 35 psig. We unit coolers and hylur.n mixing system fans, although not qualified for .35

.psig, have no ' um p nents that are susceptible to high pressures. h unit coolers are of open construction, and all ccuponents such as valves

' and motors are of sturdy construction and are capable of withstanding 35 psig. h unit cooler piping and coils are constructed to ASME Section III, Class 3 and will withstand 150 psig. h hyi % ess mixing system fans are also of sturdy constructicn and are not susceptible to high pressures. In addition, the hy& u pri mixing syst s fans will not be operating during the period when hydrogen deflagrations are occurring.

1 4

e

. e*

3-1 t

i .

y m.. ,__....,,i.m,',,,,,v. ..m..-,,., , ~ . . , , y....,, _ , , , , , , . . _ . . . _ _ _ _ _ _ _ . , _ . _

Table 1 Equipnent Description Make/ Manufacturer Qualified Function Vendor Model/ Catalog No. Pressure (psig)

DRYWELL Autmatic Depressurization System (ADS)

Main steam safety / relief Crosby 8 x R x 10, Style HB-65-DF 35 valves (ADS)

Hydrogen igniter Power Systems Model 6043 70 ignite hydrogen / air-cmbustible mixture during degraded core event Resistance thennal Pyco, Inc. 65 detectors CONTAINMD7f CNS Contalment Atmosphere Monitoring Contaiment and drywell Solenoid valve, Target Rock 66 atmosphere sampling TRCP 77KK-003 CPM Containment Hydrogen Mixing Mixing fan (typ) Fan motor, Buffalo Forge West, 15 TBEC 145T Exhaust valve (typ) Motor-operated valve, Posi-Seal, 105 IMIQ SMB-000-2 Supply valve (typ) Motor-operated valve, Posi-Seal, 105 IMIQ SMB-000-2 E12 Residual Heat Removal LPCI Injection Motor-operated valve, Velan 70 IMIQ SB-2-60 HCS Hydrogen Recmbiner Hydrogen recmbiner Hydrogen recombiner, Westinghouse 70.0 Model 4B Hydrogen Igniter System Hydrogen igniter Power Systes Model 6043 70.0 Ignite hydrogen / air cmbustible mixture during degraded core event I

l' 3-2 ,

L

RIuipnent Description Make/ Manufacturer Qualified Function Vendor Model/ Catalog No. Pressure HVR Ventilation - Reactor Plant

-Unit' cooler Unit cooler motor, Buffalo Forge 15.0 Mitigate tarperature West 445'ICZ increase during event and .

(return temperature to normal following event JRB Superstructure - Reactor Building Cont. personnel airlock Door access, Graver Woolley 118.0 Drywell personnel airlock Door access, Graver Woolley 118.0 Drywell equipnent hatch Door access, Graver Woolley 118.0 2

Cont. 'equipnent hatch Door access, Graver Woolley 118.0 Instrumentation Reactor pressure vessel Rosemount Model 1152 73.0 level transmitter RHR/LPCI permissive irm L.m entation Resistance thermal detector Pyco, Inc. 65.0 Contairunent Electric Penetrations Electrical-Penetration Conay Corp / Unique 70.0 Assembly

1. 72.0 psig based on ultimate capacity analysis

2. 56.0 psig based on ultimate capacity analysis 3-3

t g .-

-o ,

4 Attachnent III a.

HYDROGEN DEETERATION PRESSURE EFFECTS 2

b N

j RESPONSE 'IO FORCED BURN

]

I -

f-i f

s I

I P

i i

l l

i p - , -m,~ n -s n - ,--we>.m .--s--,;-,---