ML20072F955

Forwards NRC Requests for Addl Info Re Sbwr Isolation Condenser Performance

ML20072F955
Person / Time
Site:	05200004
Issue date:	08/02/1994
From:	Marriott P GENERAL ELECTRIC CO.
To:	Borchardt R NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM), Office of Nuclear Reactor Regulation
References
MFN-089-94, MFN-89-94, NUDOCS 9408240141
Download: ML20072F955 (27)

Text

.

GE Nuclear Energy f)t*WfAI 0f]C?l'( rl)Al(OOY

!?$ Curtper hevue,5m.bv. CA 95!25 l

August 2,1994 MFN No. 089-94 Docket No. STN 52-004 l 1

1 Document Control Desk l U. S. Nuclear Regulatory Connnission l Washington, D. C. 20555 Attention: Richard W. llorchardt, Director Standardization Project Directorate i

Subject:

NRC Requests for Additional Information (RAIs) on the 1 Simplified Boiling Water Reactor (SBWR) Design

References:

Transmittal of Requests for Additional Information (RAI)s Regarding the SilWR Design, Letter irom N1. Malloy to P. W. Marriott dated April 8,1994 l

i The Reference letter requested additional information regarding the SilWR Isolation Condenser performance in fulnllment of this request, GE is submitting Attachment I to this letter which transmits the response to RAI 440.6.

Sincerely, i 1 l- 1

& - 5 l<I'H**f 1

1 4

P. W. Sfarriott, Manager i

. Advanced Plant Technologies

. # M/C 781, (408) 925f>948 Attachment 1, " Response to NRC RAl"-

cc: M. Malloy, Project Manager (w/2 copies of Attachment 1)

F. W. llasselberg, Project Manager (w/l copy of Attachment 1) 3 200098 go '

I 9408240141.94o802 (

PDR ADOCK 05200004 ,

n .PDR d

g. -

RAI Number: 440.6 Question:

RAI SRXB.28 requested information concerning IC performance in currently operating BWRs and application of this information to assessment ofIC performance Ihr the SBWR. The staff requests that GE provide further infbrmation in this regard, to allow the staff to continue its review. Specifically, the staff requests that GE provide:

a. Operational data for ICs in current-generation plants, demonstrating the heat removal capability of these systems for extended periods of time. This includes heat transfer performance and accumulation (and possible venting) of non-condensible gases.

b. Assessment of the operational data demonstrating the applicability of the data to the IC design proposed for the SIlWR, accounting for the effects of any design difTerences between the currently operating ICs and the SilWR system. This includes a quantitative comparison, using scaling methodology if appropriate (i.e., dimensionless parameters), showing that the range of thermal-hydraulic conditions experienced by currently operating ICs is similar to that predicted by GE for the SBWR. The range of conditions includes (but is not limited to) accidents during which the IC is expected to operate, regardless of whether specific credit for IC operation is taken in GE's standard safety analysis report analysis of accident.

c. Demonstration of the capability of the TRACG computer code to model the performance of the ICs in current plants, including the effects of non-condensible gases that can accumulate in the IC.

d. A detailed summary of the data from the CON 1 PASS database from which GE drew reliability information presented in its revised response to SRXB.28

( AIFN No.103-93, June 30,1993).

GE Response:

Item RAI 440.6 a.

There is a long and well-documented history of successful Isolation Condenser (IC) performance at operating plants. The design rules and practices for ICs are the same as fbr other heat exchangers used in licensed nuclear reactors such as the Reactor IIcat Removal (RilR), Service Water, Feedwater Heaters, etc. Plant startup tests have confirmed that IC performance has consistently exceeded the conscivative predictions of standard sizing methods as documented in this respone,e. Operating plant experience has demonstrated that the ICs are a

I simple and reliable way to remove decay heat following isolation. The j methods used to size the SilWR IC heat exchangers are consistent with proven l conservative methodology used to size operating plant ICs.

Table b.I of this response contains a listing of plants with isolation condensers.

The longest recorded use of an IC was at Dodewaard. Figure a.1 shows that after i eleven (11) hours the heat removal curve appeared to follow the decay heat generation rate. There was no discernible reduction in performance due to buildup of non-condensibles. Dodewaard has eight hours of coolant inventory in the IC heat exchanger, the greatest of any operating plant. Therefore, it is expected that this is the longest continuous use of an IC in service.

No record ofloss ofIC heat exchanger efficiency due to buildup of  !

non-condensibles has been fi>und. The Oyster Creek FSAR 6.3 refers to a GPUN Topical Report No. 056 titled," Evaluation ofIsolation Condenser Performance with Non-condensible Gases in Steam" The FSAR states that: .

"The report concludes that closure of the vent line isolation valves or blockage of the vent line will not preclude proper operation of the system." Operational performance data for operating plants primarily consists of startup test results verifying the perfi>rmance of the isolation condensers.

The ibilowing are excerpts from the startup tests found in GENE records. The general conclusion from the tests is that IC performance was up to 200% greater than predicted using standard heat exchanger sizing rules. This may be due to-overly conservative fouling factors that are not appropriate for nuclear water quality systems. The excess capability of the ICs resulted in excessive water carryover at some plants requir ng modification of the system to reduce flow.

Water carryover has been addressed in the SilWR by the addition of a moisture separator at the discharge to the IC/PCCS pools.

.L i 2

1

I i

i 10 ,

I 9 -

[8 -

3 z o 6 9 W

$7 -E $

c< 5 n.

2 4

~

T E w

$ $ 7 EMERGENCY a 2 5 z CONDENSER g <

o o CAPACITY c)

Z Z w 84

-8> 8 H

O O w 3 z >

c3 -c 0 m

W a

W g

2 2 w 2 $

3 1

0 102 2 3 4 5 6789103 2 3 4 5 6 7 89104 2 3 4 5 6789105 TIME IN SECONDS Figure A.I. Dodewaard IC Capacity 3

hulaljon Condenser Startun Tests Prior to plant operation each system must be tested to verify its performance conforms with design requirements. The following are extracts of the IC tests from the Dresden, Millstone, Nine Mile Point 1, Tarapur, Fukushima 2, Oyster Creek and Tsuruga startup test reports. Although each test was conducted using slightly different methods, all tests confirmed that IC performance exceeded design requirements by a significant margin.

A. DRESDEN IIISOlATION CONDENSER (Ref. NEDC-10430A)

Purpose To determine the heat removal rate of the isolation condenser (IC) and the time for the system to come into operation. To demonstrate primary to secondary system leak tightness.

Criteria The heat removal rate shall be no less than 250 x 106 Iltu/h, and the system ,

shall indicate primary to secondary leak tightness.

Results The first Blu/hr heat removal capacity test was attempted at 10% reactor power with all steam bypassed initially to the main condenser. The capacity of the isolation condenser, and the decrease in reactor subcooling due to the isolation condenser return water, were suflicient to cause closure of the bypass valves and termination of the test.

The test was next attempted at the 25% power level but, because ofinadequate indication of isolation condenser water level and excess carryover, the capacity was evaluated from a reactor heat balance instead of the shell-side heat balance as originally intended. Bypass valve movement was already calibrated as a ihnction of feedwater flow thus enabling the following heat balance to be made. q l

As the isolation condenser was brought into senice, the APRMs showed a _ ,

decrease of 3% reactor power or 258.0 x 10 6 Btu /hr. This reduction in reactor l

power was due to a decrease in subcooling resulting from the IC return water. l One bypass valve plus 35% of another bypass valve closed as the system went into service. The change in feedwater flow was determined to be 0.6 x 106 lb/hr. By assuming that the change in feedwater flow was equal to the change in steam flow to the main condenser, a heat balance was performed.

AQcore = AWFW (Abcore) - AQIC AQlC =-258 x 106 Btu /hr + (.6 x 106 lb/hr) (1191 - 106 Btu /lb)

= 392 x 106 Btu /hr or 157% Design.

Automatic initiation of the isolation condenser was not employed for the test mentioned above because for operational reasons it was desirable to prevent the initial cold water slug associated with automatic starting of this system.

However, the test was repeated later with automatic initiation, and a capacity 4

check was again uade by the same method as for the earlier test. In this case the net bypass valve movement associated with establishment of the isolation ,

condenser was 1.58 valves (SIC). The corresponding APRM decrease was l again 3%, and substitution of these numbers in the heat balance equations gave a value of 480 x 10 6 litu/h or 192% fbr the isolation condenser capacity. The reactor was initially operating at approximately 20% power with all steam being bypassed to the main condenser.

The isolation condenser was leak tight and the time to come into operation was measured to be 18 seconds. This time was defined as the time from the initial valve movement to the time that the bypass valves stopped moving. The readings of the gamma monitors on vents A and 15 increased from 0.1 mrem /hr to 1.5 mrem /hr when the isolation condenser came into service.

This low value indicates that the system has no primary to secondary leakage.

During this test the steam and condensate return flow indicating switches were observed. In the case of the condensate return flow a maximum of 23" of water was recorded during operation of the condenser. The present setting of the condensate return flow trip is 32" of water.

Subsequently, on February 24,1971, a design change was made wherein the  :

opening of the condensate return valve was limited such that the efTective '

capacity of the system was 120% of design. This modification was intended to limit the secondary carryovt. and minimize the risk of system isolation from high steam flow on initiation.

With the reactor operating at 2330 MWt and 717 MWe with 1.65 bypass valves open and 96 percent core flow, the isolation condenser was brought into operation using condensate return valve 1301-3 to control flow (a special circuit was installed to interrupt the normal " scal-in" feature with a switch on the control panel byjogging the gate valve open or closed as desired). After two attempts, which resulted in calculated capacities of 132% and then 110%, the -

drive motor limit switch was placed at a position estimated to give about 120%

capacity. Subsequent initiation of the isolation condenser resulted in a calculated capacity of 120% of design.

During the test of February 24,1971 data was obtained to enable a calculation of carryover to be made. The calculation was based on a comparison of the actual isolation condenser water level drop rate, with the level drop rate equivalent to the condenser capacity with no carryover. The level drop rate based on 120%

design capacity is calculated as follows:

QlC (litu/hr) =

hfg (lltu/lb) x p (Ib/ft3) x A (ft2) x R (ft/hr) 120% capacity = 300 x 10 6 litu/hr 300 x 106  :

970 x 543/0.01672 x R (ft/hr)

R = 9.55 ft/hr or 0.16 ft/ min (no carryover). ,

t 1

5

Using the measured water level drop rates, estimates were made of the carryover at different water levels for the shell-side of the isolation condenser.

At a water level of about 6 feet, the carryover was calculated to be 8% by weight for water in the discharged mixture.

llecause of general concern over the degree of carryover it was decided to modify the isolation condenser by fitting baffles to reduce the water loss.

Following these modifications further carryover tests were made on May 27, 1971. The table below shows the change of shell-side water level with time.

ISOIATION CONDENSER CARRYOVER Indicated Level Drop Less Expected Carryover Shell-Side Level Rate Rate Ib Water /lb Mixture (fl) (ft/ min) (ft/ min) 6.3 0.64 0.48 75 5.2 0.39 0.23 59 4.4 0.27 0.11 41 4.0 0.22 0.06 27 3.6 0.22 0.06 27 3.1 0.22 0.06 27 2.5 0.18 0.02 11 2.2 ~0.16 4).00 ~0 The test of May 27,1971 also indicated an efTective capacity of 120 percent of design, thus again confirming the throttled setting of the condensate return valve; the calculation of capacity being made as for the previous tests. The isolation condenser system was again demonstrated to be leaktight, as evidenced by the relatively small increase in the "A" and "II" vent radiation monitors from 0.2 and 0.4 mrem /hr before the test, to 3.1 and 3.0 mrem /hr, respectively, during the test. l Discussion The capacity of the isolation condenser was shown to be more than adequate.

Primary to secondary leak tightness was demonstrated and a satisfactory automatic initiation time recorded.

The purpose of the isolation condenser may be expressed as follows:

To provide 16.3 minutes of operation at 310 MBtu/hr without adding makeup to the shell-side.

For the following reasons the May 27,1971 test was considered to be successful-on this basis:

a. The test duration was 16.5 minutes.

b. The shell-side water temperature was initially 212 F rather than the 1 normal 90-100 F. 1 i

6 I J

i l

c. The pre-warming of the shell-side water resulted in the loss of about l OA-0.5 feet, so that the initial shell-side inventory was less than normal.  :

d. The test could have been extended for some time longer. There was no particular reason for stopping it at 16.5 minutes.

Carryover, as measured during the May 27,1971 test, was shown to be strongly dependent upon the shell-side water level. However, the test demonstrated that an adequate quantity of water existed without makeup and that carryover was in fact almost zero below a shell-side water level of 2.5 feet.

II. MILLSTONE 1 ISOLATION CONDENSER TEST (STP 13) ,

(Ref. NEDE-13201 A)

Puroose Determine the heat removal rate of the system. Determine the time fbr the ,

system to come into operation. Determine that the system is leak tight.

Critena Determination of the heat removal rate of the isolation condenser, determination of the time for the system to come into operation, and .

verificati_on of system 'cak tightness will constitute satisfactory completion of the test.

The expected heat removal of the isolation condenser is about 206 million litu/hr.

The time Ihr the system to come into operation is expected to be on the order of one to three minutes.

The isolation condenser is expected to be leak tight under all conditions. j Results The valve on the steam line is normally open so that the isolation condenser was pm into service by controlled opening of the condensate return valve. The automatic shcIl side fill system was inhibited by closure of a manual valve and ,

the stationing of an operator at that location to put the automatic fill system in  ;

service if required. i 1

1 The time to put the system in senice was based on analysis of the transient '

recorder trace from first movement of the condensate return valve to the time bypass valve cam position reached steady conditions.

The capacity of the system was determined from steam flow change through bypass valves and a condensate return flow at 370 F. Leakage of reactor water through the heat exchanger tubes could be detected by the radiation monitor on the vent to atmosphere. Test result is shown in Table C.8-1.

The isolation condenser was put into senice in 26 seconds and has a capacity of 2.5 times rated.

7 )

- .- = , .

The radiation level in the condenser increased 0.1 mr/hr, indicating satisfactory leak tightness.

On fast activation, there was a high flow problem in the differential pressure ,

detectors in the line leaving the reactor vessel and in the return line.

Therefbre, the isolation condenser return line valve was set to open to prevent high flow and subsequent isolation. After the valve was adjusted, the isolation condenser was tested again and its capacity determined by a tube side heat balance. This test was performed by fast actuation of the return valve and the system came into service in 3 to 5 seconds.

The capacity was determined to be 373 MBtu/hr or 5.4% of rated power, This is a f actor of 1.8 times rated capacity for the system.

Tabic C.8-1 ISOLATION CONDENSER TEST RFSULTS Test Conditions Reactor Power 25 %

Core Flow ~35%

Bypass Cam Position 25 %

Bypass Valve No. I full open No. 2 93 %

Turbine Generator Off Line Before Test IC in Operation End oftest Bypass Valve No. 2 % Open 93 20 _

IC Vent Radiation Monitor 1 0.2 0.3 0.8 Level (mr/hr) Monitor 2 0.1 0.2 0.6 Q1C

= (Bypass Steam Flow Change) x (hg - hreturn)

= 0.73 x 833,000 lb/hr x (l 199.3 - 344.1) Btu /lb

520 MBTu/hr,7.6% of rated power i Where, QIC = heat removal rate ofisolation condenser hg = main steam enthalpy (reactor pressure = 979 psig) breturn = condensate return line enthalpy (return line temperature

370 F).

C. NINE MILE PT.1 ISOLATION CONDENSER TEST (Ref. NEDE-10278)

Purpose To determine the heat removal rate of the system, the time for the system to come into operation, and demonstrate the leaktightness of the system.

8 i

A Criteria Determination of the heat removal rate of each set of condensers, response time of the system, and verification of system leaktightness constitute satisfactory completion of the test.

Results The isolation condenser system is comprised of four individual condensers.

The shell side of each condenser is water-filled and is vented to the atmosphere.

The tube side (the flow path through which vessel steam flows) is condensed and then returned to the recirculation loop through a condensate return valve.

The condensers are paired such that each of two condenser sets shares common steam supply and condensate return piping.

Each isolation condenser set was tested individually. The reactor core thermal power before initiating condenser operation in each instance wa.s approximately 200 MWt. Data were recorded both from process instrumentation on the control room panels and on the transient recorder.

Condenser capacity was calculated from the control room instrumentation data using both a shell-side and a tube-side heat balance. Capacity was also calculated using a tube-side heat balance based on the transient recorder data.

The calculated capacities are given in C-3 below, " ISOLATION CONDENSER Calculated Capacity" Due to meter fluctuations, the transient recorder produced the most precise data. Table C-3 also presents the capacities calculated using these different data sources. The most drastic of the fluctuations mentioned above was noted on the shell-side water level instrumentations. The fluctuations ranged almost over the entire scale of the instrument. The source of these fluctuations was obviously the varying shell-side water density as a result of the boiling. The subjectivity used in reading the shell-side water level is reflected in the capacity calculated with these data. Likewise, the lesser magnitude fluctuations in the process instrumentation introduced subjective crror in the tube-side heat balance. The permanent record obtained from the transient signal recorder made possible a more precise evaluation of the plant parameters.

Table C-3 ISOLATION CONDENSER Calculated Capacity Calculated Capacity Condenser Set (106 litu/hr) Design Number A 13 C Minimum 111/110 996 520A7 405fA 190 121/122 453 223.78 370.86 190 A = Shell side heat balance from process instrumentation 11 = Tube-side heat balance from process instrumentation C = Tube-side heat balance from transient recorder data 9

i O

Assuming an absolute error in the transient recorder data of +/-2.5% (based on bypass valve position oscillation), the minimum capacity of the condensers would be: (a) 222.14 (106 ) litu/hr for condenser set 111/112 and (b) 187.36 (106 )

litu/hr for condenser set 121/122.

The clapsed time required for the system to come into operation is defined by the greater of: (a) the condensate return valve opening time, or (b)the time required for the main steam bypass valves to reach their new position. The time required for the condenser to come into operation was: (a) 3 minutes for condenser set 111/112 and (b) 1.5 minutes for condenser set 121/122.

The leaktightness of the system was measured by the change in radiation level at the isolation condenser shell-side atmospheric vent. . The measured change )

j is shown in Table C-4. The system was considered leaktight since the change j in magnitude was extremely small. This observed change in radiation level was attributed to the increase in steam flow through the condenser near the 1 radiation monitors.

l Table C-4 ISOIATION CONDENSER VENT RADIATION DATA l

Radiation Level (mR/hr)

Condenser Set liefore Operation During Operation 111/112 0.340 0.385 I 121/322 0.234 0.310 J D. TARAPUR EMERGENCY CONDENSER (Ref. NEDE-13111)

I Emergency condenser capacity can be measured by two methods: a shell side

, heat balance or a tube side (reactor) heat balance. The shell side heat balance is l'- made by measuring the evaporation rate of the condenser cooling water. The l i

rate at which the condenser coolant evaporates is measured by the i rate-of-change of water level in the condenser tank. l liy shell-side heat balance, the heat removed by the emergency condenser is:

Qec = W x (hs - hr) where hs = enthalpy of saturated steam at atmospheric pressure hr = enthalpy of saturated water at atmospheric pressure .

W = rate of evaporation = [AL x A)/t] x p where AL = change in condenser water level in time 't' A = condenser area of cross section ]

and p = density of water near 100 C  !

10 ,

(

l

__ - - - _ - - - - - - - - - - - - - - - - - - - - _ - - - - - - - - - - - - - - - --------J

i The tube side heat balance is given by the equation:

Qec = (Wp s1 - Wp s2)hs - (Qc1 - Qc2) - (Wfwl - Wfw2) h tw where subscript I refers to measurements madejust before the emergency condenser operation and subscript 2 refers to measurements made during the emergency condenser operation.

Wsp and Wrw = primary steam and feedwater flow readings i Qc = core thermal power, given by the average of the six Power Range Monitor (PRM) readings hs and hrw = enthalpics of saturated steam and water at reactor pressure.  !

i Recirculation pump power addition, heat lost by cleanup, and rod drive coohng ,

systems are assumed constant under both conditions. Negligible secondary steam was being used for gland seals and air ejectors during this test.

Unit I was the first emergency condenser to be tested. During the initial power operation of Unit 1 (Phase IV) the emergency condenser was put into operation.

This test operation was with one tube-bundle at a time (of the two in the unit) in order to avoid reactor depressurization, which could be possible if the heat rejection through the emergency condenser was greater than the reactor power. Once the single tube-bundle capacity was established, the condenser was again tested with both tube bundles in service, and the reactor at a power level high enough to avoid depressurization. With the experience of the Unit 1 testing, Unit 2 was tested only once, with both tube bundles in service. The results obtained from the testing are given in Table 14A-1.

Table 14 A-1 EMERGENCY CONDENSER CAPACITY  !

Tube Tube Sum of Tube 150th Unit 2 Design Value Bundle lluadle  !!undles llundles 150th llundles lloth Ilundles 1 Only 2 Only 1&2 in Senice In Senice In Service .

From Tube Side l Ileat llalance 19.9 MW 21.2 MW 41.1 MW 67.6 MW 46.5 MW j From Shell Side lleat lialance 31.7 MW 31.2 MW 62.9 MW 69.4 MW 59.2 MW 39.6 MW l Core l'ow(r i Before Testing 50 MW 302 MW ' 132 MW Results from the shell side heat balance calculations show reasonable agreement between the two units. The tube side heat balances performed at low ,

power levels are grossly inaccurate mainly due to the error in the l measurement in primary feedwater flow at low Dows. The accuracy of the shell side heat balance calculation is independent of the core power level, and is reasonable at all power levels. Good agreement was found among all l I

calculations made by this method.

11 l

During all testing, the emergency condenser's vent side gamma activity was monitored and found to be less than 0.1 millirem / hour, a negligible quantity.

A visual inspection of the condensers during operation indicated no serious interference between structures due to thermal conditions, and no vibration problems.

E. FUKUSillh1A 2 ISOI ATION CONDENSER (Ref. NEDE-10426)

Purpose To determine the heat removal rate of the isolation condensers and the time' for the system to come into operation, and to demonstrate that the system is leak tight.

Criteria Level 2 The expected heat removal rate of each isolation condenser is about 36.2x106 Kcal/hr.

The time for the system to come into operation is expected to be on the order of one to three minutes.

The isolation condensers are expected to be leak tight under all conditions.

Results The isolation condenser is comprised of two individual condensers. The shell-side of each condenser is water-filled and is vented to the atmosphere.

The tube-side (the flow path through which vessel steam flows) is condensed and then returned to the recirculation loop through a condensate return valve.

The condensers are paired such that each of the condensers has separate steam supply and condensate return piping.

Each isolation condenser set was tested individually. Data were recorded both on process instrumentation in the control room and on the transient recorder.

Condenser capacity was calculated from the control room instrumentation data using both a shell-side and a tube-side heat balance. Capacity was also calculated using a tube-side heat balance based on the transient recorder data.

The reactor was operating at 234 N1Wt (17% rated) with all the steam bypassed to the main condenser. The feedwater system was in manual mode, and the

$1PR (Alechanical Pressure Regulator) was in senice. Alakeup water to the isolation condensers was valved out of service.

Table C.6.1 summarizes the results of the measurements and shows that the measured capacity was much greater than designed. The disparity between shcIl and tube calculations would indicate a large amount ofliquid carryover from the condenser shell. This, in fact, was observed at the condenser vent during the test. The carryover was large enough that it was decided to reduce the steam flow to the condenser by limiting the opening of the valve in the return line.

12

a The isolation condenser was tested again at 50% power to find the return line valve opening that best matched the measured and design heat removal capacity of the isolation condenser. The results are presented in Figure C.6.1. It was decided to limit valve V-1301 to about 15% opening. However, this was never implemented because of the complexity cf the valve's positioning mechanisms.

Additionally, radiation levels at the shell-side vent locations were monitored during the tests. Only negligible increases in the radiation levels were observed. Likewise steamline pressure differences were monitored. These are shown in Table C.6.2.

1

)

I i

4 13

- - . _ =.. . . .. = _ -- . _ -

Table C.G.1 ISOLATION CONDENSER CAPACFIY Capacity (10 6 Kcal/hr) Time to isolation Shell-Side Tube-Side Design Operation Condenser Nicasured Alcasured (min)

A 68 48 36 1.8 11 123 54 36 1.8 Table C.G.2 SHELL-SIDE VENT RADIATION LEVELS Radiation Sensor Reading (mr/hr)

Isolation Condenser In Service liefore/During A(a) 13(a) C(a) D(a)

A liefore 0.2 0.3 0.2 0.2 During 0.4 0.6 0.5 0.5 11 liefore 0.2 0.3 0.21 0.2

.During 0.23 0.41 0.31 0.31 (a) Radiation Sensors A/B Monitor Vent A; Sensors C/D Monitor Vent B Shell and Tube-Side Temperatures and Differential Pressures Temperature ( C) Differential Pressures (%)(b)

Tube Steam Water Isolation Condenser liefore/During Shell A/[3(c) C/D(C) A/ll(C) C/D(C) A/ll(C) C/I)(C)

In Service A Ilefore 46 54 55 0 0 0 0 During 88 tra 165 38 40 17 16 11 llefore 43 42 41 0 0 0 0 During 88 106 175 40 40.5 18 18

-i l

(b) Differential pressures are monitored in tube-side piping for pipe breaks.

Steam - 635 cm H2O = 100%

Water - 503 cm H2O - 100%  ;

(c) Sensors A/C monitor condenser A; Sensors B/D monitor Condenser B. j 14 l

l

70 50% POWER /100% FLOW 60 -

50 -

DEstGN VALVE 42.l uwt g ,o _

c f

6 5

5 30 -

e 8

O CONDEN$ER A 20 O CONDENSER B 10 o

A 1 I v

5 10 15 RETURN VALVE OPENING (%)

(v-tm-s) i l

Figure C.6.1 Isolation Condenser Capacity Versus Valve Opening i i

l 15 l

l

F. OYSTER CREEK ISOl ATION CONDENSER TEST (Ref, NEDE-13109)

Puroose To iletermine the heat removal rate of the isolation condensers and the response time of the system.

Results The test was performed at the following reactor conditions Reactor Power 200 MWt Reactor Pressure 990 psi r Core Recirculation Flow 58 x 10 )lb/hr Vessel Steam Flow 470,000 lb/hr The condensate return line valve was tripped to initiate the transient.

Condenser water level was recorded using plant instrumentation and the transient was recorded on the Sanborn recorder. IIcat removal rates were calculated for both condensers using shell-side and tube-side heat balances.

Tube-side heat balance Q = (.Ws1 - Wsg) (hcg - hC r) where:

Wsl,Wsg = Turbine steam flow before and after condenser initiation mg2 = main steam flow after isolation condenser initiation hc g = enthalpy of steam entering the condenser her = enthalpy of fluid leaving the condenser Shell-side heat balance Q= p AV/AT hf g p = density of fluid in condenser AV/AT = Change in volume of water with time hfg =hg - hr Evaporation enthalpy at shell temperature The measured heat removal rates for each condenser are given in Table D.3.1.

Iloth condensers have a measured capacity of at least 70 MWt, compared to a design capacity of 56 MWt. The response time of the system, defined as the l time for the bypass valves to reach steady state,is about 10 seconds. Figure D.3.1 shows the transient traces obtained on the Sanborn recorder for the test of 1 isolation Condenser A. Power (neutron flux) rose about 20% above its initial l value and settled out slightly below the initial condition. There were two effects 1 which caused the power swing. First was the leg of cold water entering the  ;

reactor when the condenser return line was opened, and second was the )

effective increase in feed-flow from the condenser since the feedwater system was on manual control and, therefore, did not respond t > level and steam flow l

1 16 l

l I

changes except by operator action. Pressure initially dipped 10 psi and returned l to its original value within 15 seconds. Water level rose 7 inches and slowly returned to its initial value. Radiation levels measured at the condenser were not excessive, indicating satisfactory leak tightness. The increases in radiation level that were seen near the condensers are attributed to the flow of reactor steam through the system.

Each condenser blew significant amounts of water out of the vents during the first 5 or 6 minutes of shell-side boiling, therefbre, a heat balance based on change in shell-side water level used the rate oflevel change after this period of excessive carryover settled out.

Table D.3.1 '

ISOI.ATION CONDENSER HEAT RENIOVAL Condenser A Condenser 13 (N1Wl) (51Wt)

Shell-side heat halance 131 86.5 Tube-side heat balance 70 82.5 G. TSURUGA ISOIATION CONDENSER (Ref. NEDE-10224)

November 23,1969 to November 23,1969 l'urpose To determine the heat removal rate of each condenser and the response time for the system, and to demonstrate the leak tightness of the system.

Criteria Level 1 Determination of the heat removal rate of each isolation condenser, determination of the time for the system to come into operation, and verification of system leak tightness constitute satisfactory completion of this test.

Reactor Conditions Reactor thermal power 168 AfW, reactor pressure 67.8 kg/cm 2 bypass valve 70% open, isolation condenser temperature 98 C.

Results The response time and heat removal rates calculated from the tests of the isolation condensers are shown below.

Condenser A Condenser il Response time (sec) .. . . .. . . . . . . . . . . 10 14 Shell side heat balance (51Wt) ...... .... 65 46.5 Tube side heat balance (N!Wt) ... . ... . . 62.2 60.8 17

4 t

/0/ U y ,. ,

-fH M. _ _ , , , . ,y+4, s i 41 -

e i Ve s3 u

. 4-+-.M 'l  ; .- 1 -'i.. '

Fees w e-Q _'4!Dh

.. . + + - [ b.

_dm

.[

'l . N.

'] i ' ' i 7'-T ; #

'-  : M ,*4 -= ~ ..I i a ] l I i e v

~r+-1 +---- - , d 4d. l-+ + 1-4 +w ' ; l H 4 44 . ;,

Iu'r e . ' .

(( ,J g g t

i .

' s' . '

Inz rum scve <

.5 r e 4., +

-t h, t- .% ~d.r !(t

.- 1 I--, -a Q_ p.4 , .

i. p ' , , i t .. l EM M dI~'. Ei

• hi-r .hh._h

.i+J y .., ,_ . . .

r i 441 r

+f- -+,---t + + -*--t 7 4, -4~4-4 .

, l- I'-

l --+ -.

r q yp 4 . e

. .,4, q .,,

- t,., -

o p . 1*+l+i-+y . t -e

.-+--

+3 9. JA .. L..s

%. 9.. + ,A 9,,,4,

> ii

'

• I l

l ) l.

I Io / 00 - . , _ ,,,,

i 4

~y- '5 -- a~"

I Ls.a yn .- + ...s , .! It g,,,,,

m g,,,4a_,.

i

- , . ... _.1 ,

i r a ,

r-.a , ., ; ... .4 ..

Go.g l-4 -

7 4.- , ..g 4 I f wp.. w p'M-.L  ! .

(g, y' ,,g ) - - o.".

i I- -t

.bi - ,

%q . ,. g *-H

.,. + I; i.!

M-o .c.-4.-

+ I . i .

O imme-.dd M . I y L1 '~L .

1 ;I f, ll ll'lI  ;, e [ ,,, 8 .

l i , t, ,; 't

=

. i,,

A'4 reg  ! 1.. t_. h [' _

i Lf ,,h))]

we. mr , , .

pq 'Tr?- ,

7-

- ...-.bm l +-,

~

.-.a ,. .n

[

v ,-+ 9- --A--. l. ,

e I,

o '_ . i i i t-+-H,.*T*tm.

/24 _

-- .4+, -

f I

y Lj Arem o _ _n_.x

$%d Mffnc s9*g Q g, ( */,) TT' 'r , S l Mroru.ar~ w mA . .. .  ; v

- N .,l '

t< s<, r u a ., masi y 4[

u T.

i J

F. , '$  ! .1.

2.0 R- g ,' l' I.

I i jiI I 2.

O r /3 l ggg., i l l

' - a

/s.

m ,

---+r -, 3 , ,

L . I

! l' p' . =. v *;w4.,,

.*./ee W.

I gg f .; ;

t.7--. g ,

3 g 4ths*%de +4 4..

a l

bbr (4., -

~4-fpf-t-t -

._r- b.. .f. A

~ #

O 1,F W . _.p.,i T-g- q.i_4

4 ._. .

gp) ,,--q. ._

-..14 , p. p . .+.+..

....,_,.,t

. u D M I I 1 I I a i d;L . , .

g i i 'illl

.:- 1 ll Il "i

n m

-_ rJ.-m+

1 . ,

12.a r, .. .

~

c -2<-s r4 A 2e vu,,

Ut! 6 4 -..i;,. E '

,m y,

o > 1  ; m . .

fC9 s t .

_.. 4

.t.

.a. .,

v., .

%.g

_ .. a ,.

n . ,, r ,

m m v ,, u -

l =.

.L .nsei6 1

T-rd'..< l .., _-

i  !  ! ..;

o .

.m a m . w.. ~ -

rs. w ,.. ,

c- r., , ,

m w '

,J^^'

+,.

~n. ,..

?9

,-_ i:'

F n -r

>.*m~3(

[M,4/ W yff" T ;,

4,. ,.

. .. '_r.. , - a e..s,. +

,~ +

ne g.a m, u

.r

...s 14 i z E It . i .i.ii.i u. a.Q J_4-m M

( L II J ' .1 -I _4_ _ l _ .1. L J .1 :,14 al C --u- ----

2 Figure D.3.1 Isolation Condenser Test, Run 6 l

1 18

1.

Discussion The tests of the isolation condensers were performed with the bypass valve J initially open about 70% and with the condensers isolated from the water storage tanks. The bypass valve position was recorded during the test and, after the condenser return valves were closed, the volume of water drained from the storage tanks to fill the condenser shell back to its original level was determined. The heat removal rate calculated from the change in bypass valve position is considered to be a more reliable measure of steady-state condenser capacity than the shell side heat balance because there are fewer variables involved in the calculation and the determination of changes in the parameters ofinterest are not as prone to the influence of outside variables such as carryover. i l' sing the design bypass valve capacity (16.7% of rated power of 968.4 N1Wt) the capacity of each of the isolation condensers is at least 60 N1Wt or about 200% of - ;

the design capacity of 31.9 hlWt per condenser. It was visually verified during l operation that the carryover at the shell side vents was insignificant. l When the condensers were put in operation the neutron flux went through a one-cycle oscillation of about 10% and then settled out at essentially the initial value. The bypass valve closed to an equilibrium value of about 30%. The feedwater system was being operated manually during the test to maintain reactor water level, and neutron flux, and bypass valve opening increased in phase with the feedwater injection, However, these traces consistently returned to the same deflection between feedwater transients. The reactor pressure dropped very slightly (~ 0.125 kg/cm2) while the bypass valve was closing during the initial transient but quickly returned to and maintained the original pressure during the transient. The radiation level in the condensers increased less than 0.1 mr/hr, indicating satisfactory leak tightness.

t 19

Item RAI 440.Gb.

Table b.1 is a tabulation of operating plant IC parameters compared to the SilWR.

Most operating plants and the SilWR ICs have been sized to handle approximately 3% reactor thermal rating. The major difference between most operating plants and the SilWR is that the SilWR has three (3) ICs of 1-1/2%

capacity where most operating plants have one or two condensers of 3%

capacity. Also, the SilWR has an IC coolant inventory sized for 72 hrs compared to between eight hours and one-half hour for the operating plants.

Operating plant IC heat exchanger thermodynamic parameters used to calculate heat removal capability are based on tube side saturated steam pressure of approximately 1,000 psia and shell side temperatures of approximately 228 F.

The SilWR IC condenser capacity is based on 1,050 psia saturated steam on the tube side and shell side temperature of 212 F. The SilWR IC is in an open pool, rather than a slightly pressurized vessel as in operating plants, which accounts fbr the lower shell side temperature. The SilWR IC has vertical tubes, whereas the operating plant units have horizontal tubes. The range of operating plant IC thermal-hydraulic conditions is similar to that expected to be experienced by the SilWR. Therefore, operating plant IC experience is clearly applicable to the SilWR.

Item RAI 440.6 c.

Capability of the TRACG computer code to model the performance of the ICs, including the effects of non-condensible gases that can accumulate, will be evaluated and documented following completion of the SilWR Testing and Analysis Program (TAP).

1 l

20 l

1

.~ . .. - .

f 4

l l

Table b.1

)

OPERATING PLANT VS SilWR 1 ISOLATION CONDENSER PARAMETERS

-l Minimum IG Reactor Number ilundl :s Capacity hr. Capacity per IC Thermal of per and  % Thermal .

Megawatt Condensers Shell (ib. water) Megawan j j

Dresden I 700 1 2 Note 4 Note 4 Humbolt llay 165 1 Note 4 Note 4 Note 4 Garigliano 508 1 2 Note 4 24 KRIl-A 800 1 2 4 6 l

Ilig Rock Point 157 1 2 Note 4 Tarapur 1 & 2 1,322 1 2 43 3 (l83,000)I Dodcwaard 165 1 2 83 6 Nine Mile Pt. I 1,538 2 sets of 2 1 1.5(2) 3.6 per set  !

(l70,000)I i Oyster Creek 1 1,860 2 2 1.5(3) 3 (170,000)1 Dresden 11, III 2,527 1 2 Note 4 3 Millstone 1 2,011 1 2 0.5(3) 3 (129,000)

Tsuruga 1,061 2 2 Note 4 3 (117,200)I Nucienor 1,381 1 2 1(3) 3 (105,000)1 Fukushima 1 1,380 2 2 Note 4 3 OKG1 1,500 1 2 8 6 I (500,000) l S11W R 2,000 3 2 72 1.5 (3,083,700)  :

Notes.

- 1. The amount of water above the IC condenser tubes is the minimum specified in the IC condenser purchase specification and may be significantly greater in the as-built heat exchanger.  ;

I

2. An additioral eight hours of cooling water is available from makeup tanks. I

3. The time of available cooling capability is based on decay heat rate curves and the amount of cooling water available above the IC tube bundles. Times will vary depending on decay heat assumptions and as-hulk tank capacity. '

4. Values not found in IC purchase specifications or specification not found in GENE records, i

l l

i 21

e d hem RAI 440.6 d.

Provided below are summary reports from COMPASS and the Nuclear Plant Reliability Data System (NPRDS) on the reliability of ICs in service.

Summary Report on NPRDS Data on IC Failures in Operating BWRs NPRDS (Nuclear Plant Reliability Data System) is operated by the Institute for Nuclear Power Operations (INPO) to collect component failure rate data from operating nuclear plants. On September 26,1990, NPRDS was queried for data on all failures in isolation condenser systems at U.S. BWRs. Failures were reported at the following operating plants:

Dresden 11 Dresden III Oyster Creek Nine Mile Point 1 Millstone NPRDS responded with data dating back to 1974 (the approximate time that plants started reporting data to NPRDS). Using a factor of 70% average plant availability, an estimate of the total plant operating time for this span of data is 450,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> over 50 reactor years for the above plants, which represent only about half of the IC equipped BWR ficet.

The components included in the NPRDS definition of the IC system, and the number of failures reported for each component are as follows:

Accumulator 0 Annunciator -0 Circuit Breaker 0 IIcat Exchanger 2 Bistable Switch 49 1.evel Controller 2 Indicator / Recorder 5 Integrator 6 )

Power Supply 0 i Transmitter 15 i Pipe 29 i Relay 14 Sup[x rt 33 Valve 31 Valve Operator 36 l

l 22

i l

I Summary discussions of the reported failures for each of the components follow:

i IIcat Exchanger: Only one of the two reported heat exchanger failures  !

contained narrative description, and that narrative was unclear except to i indicate that the heat exchanger was declared inoperative due to "line break l sensing" (Note: The heat exchanger failures are believed to be for Millstone 1 due to a salt water intrusion into the reactor.) l Bistable Switch: Of the 49 reported fitilures,43 were reported to be due to instrument setpoint drifting out of technical specification limits in the high-flow delta-pressure sensor. This recurring problem occurred at all plants except Nine Mile Point and continues to the present time at most plan ts. The remaining 6 failures were random.

Transmitter: Of the 15 reported fitilures,9 were due to the same setpoint drift problem as above, except that at Dresden 111 they were classified as transmitter failures. The remaining 6 were due to miscellaneous random problems.

1.evel Controller: The two level controller failures were random and unrelated.

Indicator /Itecorder: The five reported fitilures were random and mostly due to wear and aging.

Integrator: The six reported fitilures were random and largely due to aging.

Pipe: Of the 29 reported pipe failures, only 15 contained descriptive narratives. These 15 failures were all due to cracks in lines external to the IC, mostly outside of the drywell.

llelay: The 14 reported relay failures were due to various causes and apparently did not afTect plant operation.

Support: All but one of the failures were due to pipe snubber problems.

Valve / Valve Opciator: Although the data classifies 31 litilures as valve litilures and 36 as valve operator failures, the classifications are not ahvays distinct. Of the 67 total reported failures,13 were failures of valves to open and 12 were failures to clow. In addition to the 12 failures to close, there were an additional 14 reported cases of excessive leak-through. It is sometimes diflicult to distinguish between a leak-through and a fitilure to completely close. Additionally, there were 7 reported fitilures of excessive out4cakage. The remaining 21 failures were miscellaneous problems invohing valve packing, torque switches, limit switches, and other items.

23

Of the total 67 valves,32 were either unidentified or the identity of the part number was not known. Of those that were identified,15 were in the condensate return line. An unknown number of these were isolation valves. There were 1I reported failures in steam inlet isolation valves,7 failures in vent valves, and 2 failures in level control valves.

From the infbrmation included in the report it was not possible to determine the effect of individual failures on plant operation. However, the availability of IC appears to be high compared to other more complex systems. The vast majority of problems have been with electrical components. The isolation condenser itself is highly reliable and would be operable even in the case ofleaking tubes.

Summary Report on COMPASS Data on IC Failures in Operating IlWRs COMPASS (Comprehensive Performance Analysis and Statistics System) is the LFNE reliability data system for operating plants. This system receives daily reports from most domestic operating plants. Data on equipment performance is mostly in relation to failures that result in plant outages or power reductions.

Also included are critical path and other important activities that are conducted during plant outages. Not included are routine maintenance activities conducted while the plant is operating. Also not included are descriptions or logging of test activities or failures found during test (unless the failure resulted in plant shutdown).

On October 2,1990, COMPASS was queried for data on all failures in isolation condenser systems at the following operating plants:

Dresden II Dresden III Oyster Creek Nine Mile Point 1 Millstone COMPASS responded with data dating back to 1970 and into 1989. The total calendar time for these 5 plants is 874,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />. The total operating time fbr these plants is 607,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> or over 70 reactor years of operation. This represents about half of the IC equipped llWR fleet.

The IC components that had reported failures and the number of failures for each component are as fbliows:

Cable 3 IIcat Exchanger 5 Instrumentation 9 Pipe /Nonle/ Safe End 19 Support 5 System 6 Valve 57 24

6 Sununary discussions of the reported failures for each of the components follow:

Cable: The three cable entries were to replace cables with environmentally qualified cables at NRC request.

• Ileat Exchanger: One of the heat exchanger failures was for a major modification; one was for sand-blasting and painting; and one was for repair of a crack in a diaphragm weld. The other two fitilures were tubing leaks, in one case requiring a complete retubing.(Note: llelieved to be Millstone 1 after a salt water intrusion.)

• Instrumentation: Two of the instrumentation failures involved a timer, three were due to setpoint drift, and four were miscellaneous or unspecified failures.

• Pipe /Nonle/ Safe End: Sixteen of the piping failures were due to cracking, primarily due to IGSCC. Two of the failures were due to water hammer problems, and one failure was to replace a head gasket on a steamtrap.(Note:

Two water hammer incidents occurred due to condensate pockets in the steam lines caused by poor drainage. Addition of drain lines eliminated the pockets and water hammers have not reoccurred.)

Support: All of the support entries were for problems with snubbers.

System: Three of the system entries were for conducting heat capacity tests, and two entries were for miscellaneous problems. Two of the entries involved contaminated spills outside of the plant.

Valve: The distribution of the valve failures accordiog to type of valve is as follows:

check valves 3 vent valves 7 isolation valves 22 solenoid valves 1 condensate return valve 3 general (unidentified) 21 There is no indication in the data as to how many of the unidentified valve failures might be failures in the condensate return valves.

The distribution of the valve failurcs according to type of failure is as follows:

packing leaks 13 limit or torque switches 7 broken / bent stems 2 misc./un.specified 35 25

Of the three identified condensate return valve failures, the nature of two of the failures was unspecified. The other was a failure of the valve to open during a pressure transient resulting in operation of a safety relief valve. This failure to open was due to either a faulty contact or a torque switch being out of adjustm ent.

Of the 104 total events recorded,15 caused the reactor to be shut down including one automatic and one manual scram. In 37 of the events, IC work was on critical path contributing approximately 4216 hours0.0488 days <br />1.171 hours <br />0.00697 weeks <br />0.0016 months <br /> to plant outage time, approximately 0.5% unavailability.

The apparent largest contributor to the reported event frequency was the valve, causing 57 of the 104 entries. Valves also caused 11 of the 15 reported IC causes ,

of plant shutdown, including the single automatic scram. In contrast, valves '

contributed only about one-fifth (881 hours0.0102 days <br />0.245 hours <br />0.00146 weeks <br />3.352205e-4 months <br />) of the total critical path unavailability, a proportionately small amount. From the reported data, it is not possible to determine how many of the valve failures would have prevented proper system operation on demand, but it would appear to be only a small fraction of the entries.

Another item ofinterest is the heat exchanger. There were only 5 heat exchanger events reported, and none of them was a failure that would have prevented the IC from responding to a demand. In addition, the heat exchanger events contributed only 43 hours4.976852e-4 days <br />0.0119 hours <br />7.109788e-5 weeks <br />1.63615e-5 months <br /> of critical path unavailability. However, the total component (and system) outage time for maintenance was 1565 hours0.0181 days <br />0.435 hours <br />0.00259 weeks <br />5.954825e-4 months <br />, and one of the events caused the plant to be operated at a reduced power level of 25-40% for about 5 weeks while tube repairs were being made. Most heat exchanger repairs were made off of the critical path during outages for other work.

The largest contributor to critical path plant unavailability was piping (3125 of the total IC contribution of 4216 hours0.0488 days <br />1.171 hours <br />0.00697 weeks <br />0.0016 months <br />). Piping also required 11,220 hours0.00255 days <br />0.0611 hours <br />3.637566e-4 weeks <br />8.371e-5 months <br /> of component and system unavailability due to maintenance, repair, or replacement. Almost all of this unavailability was due to cracking, probably mostly due to IGSCC.

Setpoint drift in the IC differential pressure instrumentation, which has been a chronic problem, did not appear as a major contributor. This is because most setpoint drift problems appear in test and do not require plant outage time.

Hence they did not appear in the rOMPASS data.

When comparing the IC COMPASS data to other systems it is apparent that the reliability of the IC benefits from simplicity.

26

,