ML20085E850
| ML20085E850 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 05/17/1976 |
| From: | Heather Jones, Michelson C, Tyler T TENNESSEE VALLEY AUTHORITY |
| To: | |
| Shared Package | |
| ML20085E842 | List: |
| References | |
| NUDOCS 8308160607 | |
| Download: ML20085E850 (28) | |
Text
. .
~
o 9
ENCLOSURE 2 TENNESSEE VALLEY AUTIIORITY BROWNS FERRY UUCLEAR PLANT UNITS 1-3 ,
RilR PUMP PROTECTION AGAINST OPERATION IN EXCESS OF DESIGN RUNOUT
'}
C. Michelson H. L. Jones T. G. Tyler MAY 17, 1976 O p
$$k ho G ,
I
. . O O~
TABLE OF CONTENTS g Panc No.
1.0 INTRODUCTION
. . . . . . . . . . . . . . . . . . ...... 1 2.0 BACKCROUND . . ....................... 1 2.1 Emergency Core Cooling .. .... . .. ....... 2 2.2 Containment Cooling . . . . . . . . . . . . . ..... 3 3.0 RllR PUMP PROTECTION .................... 4 3.1 Punp and System Head and NPSI! Curves . .. ...... 5 3.2 Units 1 and 2 Pump Protection . . . . . . . . . . . . . 7 3.3 Unit 3 Pump Protection . .... . ... ....... 8 3.4 Verification of Emergency Core Cooling Flows ..... 8 4.0 RHR PUMP TEST RETURN LINE FLOW AND FLOW INDUCED VIERATION . 9 4.1 Flow and Vibration Tests . .. .. . .. ....... 10 4.2 Units 1 and 2 Flow and Vibration Verification . . . . . 11 ,,
4.3 Unit 3 Flow and Vibration Verification . ....... 11 4.4 Proposed Corrective Measures ...... ....... 11 5.0
SUMMARY
AND CONCLUSIONS ........... ....... 12
6.0 REFERENCES
. . ....................... 13 TABLES ............................. 14 FIGURES . . . . . . ....................... 18
?
h s
e
o
. . O LIST OF TABLES Tabic % Title Page No.
1 Minimum Total RilR Pump Flow 14 Requirements for Adequate ECCS Response 2 ERR Pucp Protection and Test 14 Return Line Square-Edge Orifice Plate llole Sizes 3 RHR Pump NPSH Required .
15 4 Total RllR Pump Flow Test Results 15 in LPCI Mode ,
5 MIR Pump NPSH Available Test Results 16 in LPCI Mode 6 Periodic RHR Pump Surveillance 16 Flow and Discharge Pressure Acceptance Criteria ,
7 Typical RHR Test Return Line 17 Y Flow E
I l
6
~
. . O O LIST OF PICURES Figure Titic g Page No, 1 Modified LPCI System--One Pump 18 Running Without Pump Orifice 2 Modified LPCI System--Two Pumps 19 Running Each Without Pump Orifice
- 3 Modified LPCI System--One Pump 20 Running With Pump Orifice 4 Modified LPCI System--Two Pumps 21 Running Each With P, ump Orifice 5 Original LPCI System--Pour Pumps
- 22 Running Each Without Pump Orifice 6 Original LPCI System--Pour Pumps 23 Running Each With Pump Orifice 7 Original LPCI System--Three Pucps 24 Running Each With Purp Orifice 4*
4*
\
I e
0 t
. . O O 1.0 INTRODt?CTION The Browns Ferry Nuclear Plant hccrgency Core Cooling System (ECCS) design and performance for units 1 and 2 have been the subject of a recent review.
This review has led to a proposed change in the system, as reported in reference 1, which provides a significant reduction in the peak cladding temperature for the 7X7 fuel following a postulated recirculation line break. This reduction in peak cladding tcrperature has been accorplished b'y climination of the Low Pressure Coolant Injection (LPCI) System rceirculation loop selection and keeping the residual heat removal (RHR) crosstic valve closed.
The operating modes of the LPCI pumps are changed by the LPCI modification such that two purps discharge to cach injection header thereby changing the discharge flow characteristics from that previously established.
Additional flow resistance must be added on the discharge side of each pucp to ensure satisfaction of not positive suction head (NPSil) require-ments at the design runout of each pump that is delivering flow to the y ,
postulated recirculation line break. The original LPCI design was evaluated on the same basis and found in need of additional flow resis tance.
Major areas of discussion in this report include the RilR pump and system performance with and without the added flow resistance for the three Browns Ferry units. Also considered is the effect of the added flow resistance on the RllR purp test return line flow and flow induced vibration.
8
2.0 BACKGROUND
During a single f ailure analysis of proposed modifications to the LPCI mode of RIIR System operation for units 1 and 2 at Browns Ferry, a new potential failure was identified which was common to both 'the original design and the proposed nodification. The identified failure occurring af ter a loss-of-coolant accident (LOCA) could result in short-term Ri!R pump operation in excess of design runout. This was considered an unacceptable challenge to pump availability since two Ri!R punps are required for long-tcrn
containment cooling. His potential deficiency was reported to the Nuclear Regulatory Commission (reference 2). The Cencral Electric Company also informed its customers of the problem (reference 3). The details of the failure and the potential effects on emergency core cooling and contain-ment cooling are discussed below.
2.1 Emergency Core Cooling The original LPCI mode of the RHR system at Browns Ferry was the standard BWR/4 configuration, using four pumps and loop-selection logic. For this d: sign the LPCI injection valves are clos,cd in normal operation end the RHR cross-tie valve is open. On receipt of an accident initiation signal following a rceirculation line break in one recirculation loop, the loop cslection logic uses a network of pressure transducers to determine which
)
rceirculation loop is unbroken. The LPCI injection valve in that loop is l signaled to open, the recirculation pump discharge valve in that loop is cignaled to close, and LPCI flow from all four pumps is directed to the 4*
[]
unbroken loop. Failure of the injection valve to cpen would disable LPCI #
l j
entirely. A single failure such as incorrect loop selection would result in flow from all four RHR put:ps being lost to the broken loop. !
ne final ECCS acceptance criteria adopted by the AEC reduced operating flexibility and introduced possible power level restrictions for the I standard BWR/4 design with 7X7 fuel. To offset the affect of the new -
criteria, a modification to the LPCI mode of RHR operation was developed which takes advantage of credit given for the flooding effect achi2ved through the availability of at least some of the RHR pumps under certain oingle-failure conditions. For the modified design, the LPCI injection valves and the RHR cross-tic valve are closed in normal operation. On j rsceipt of an accident initiation signal following a recirculation line
(
t,rcok in one recirculation loop, both LPCI injection valves are signaled I to open, both recirculation pump discharge valves are signaled to close, and the LPCI flow from two RHR pumps is directed to the unbrok'en loop. The LPCI . flow from the other two pumps is lost to the broken loop. A single failure such as a spurious accident signal from the opposite unit would result in LPCI flow from one RHR pump directed to the unbroken loop and one i
. , O O' 3 RilR pump flow lost to the broken loop. Acceptable emergency core cooling under various sing 1c-failure conditions is assured by swapping Rl!R pump <
motor power supplies and makinggccrtain other electrical and mechanical hardware and wiring changes as outlined in reference 1.
The limiting singic failure for the modified design is that failure which results in the longest reflood time and consequently the highest peak cladding temperature (PCT). Sensitivity studies have been performed, and reported in -
reference 1, which demonstrate that a typical liciting failure for a recirculation pump suction line break in the modified system is the failure of the LPCl injection valve in the unbroken loop to open.
For a recirculation pump discharge line break, the compicment of ECCS equipment availabic af ter single failure is identical to that of the unmodified system following an injection valve failure. The blowdown transient, however, is much less severe for a break in this location because the f rictional losses in the suction piping, the cyc of the pump, and the pump casing, combine to reduce the ef fective size of the break. The blowdown is so.much less sovere that in fact the calculated peak eindding temperature 'g is lower than for the suction side break. The suction line break remains 1
l the design basis accident for the modified system, but with a lower calculated peak cladding temperature than for the original LPCI design.
2.2 containment Coolinr.
The RllR system operating in the LPCI mode, as discussed in section 2.1, performs a short-term, post-LOCA emergency core cooling function. The RilR system provides a long-term containment cooling function when operating in the containment spray or suppression pool cooling modes. For adcquate containment cooling a minimum of two RllR pumps and associated heat exchangers, and two EllR Service Water pumps and associated valving on service water headers which supply the heat exchangers must remain availabic in the long term.
The availability of an adequate ntunber of RilR purcps for long-term contain-ment cooling was investicated and reported in reference 1. In the LPCI node
\
1
. . O O 4 of the RilR system in a standard BWR/4 configuration, a single failure could result in all four RllR pumps discharging to a broken recirculation loop.
For the modified design, two Rily pumps are assured of discharging to a broken loop without a single failure, and one RilR pump discharging to the break is possible with a singic failure. For such cases, it ic necessary to assure that the pumps required for long-term contain= cut cooling remain avr.llable af ter operating for a short time at high ficw rates with limited not positive suction head.
The flow characteristics of the RHR pumps discharging into a broken
~
recirculation loop will be different from that established for an unbroken loop. 1hc difference results from the loss of jet pump flow resistance l
(over 50 percent of the total LPCI system flow resistance) when the recirculation loop breaks. Additional flow resistance is required in the LPCI system to limit RilR pump flow to an acceptable value. The effect of any added flow resistance on long-term containment cooling capability must be considered.
l Singic failures which might influence long-term containment cooling were i' ,
considered in reference 1. An adequate nua.ber of RHR pumps, heat exchangers, and service water pumps were assured for all postulated single failures provided the RHR pumps are protected against excess flow effects.
3.0 RHR PUMP PROTECTION The protection of RllR pumps against excessive flow experienced when delivering to a broken recirculation loop requires consideration of the minimum flow requirements for adequate ECCS response and containment cooling, and the mximum flows which are commensurate with the estimated availabic net positive suction head. Since the LPCI modification has been added to Browns Ferry units 1 and 2 but not unit 3, the required protection was evaluated on a unit basis.
The minimum flow requirements for adequate ECCS response were established by Cencral Electric and are given in Table 1. The maximum allowable RilR puq ficvs were established by TVA on the basis of system flow resistance studies and the pump operating characteristics as determined from the I
. . O O '
manufacturer's certified pump performance data and verified by in-plant tests. The maximum flou limits were assured by the insertion of a flow liraiting orifice plate downstream of each RilR pump as indicated in Table 2. The results are outlined below.
3.1 Pump and System Ilcad and NPSit curves Figures 1 through 7 arc included to show system head loss and available net positive suction head (NPSil) curves for one, two, three, and four pump opera-tion, with and without a pump protecting orifice. Also shown is typical EllR pump total dynamic head (TDil) and required NPSit curves derived from certified performance curves supplied by the pump manufacturer. -
System head loss curves are included for 20 psid and 0 psid (reactor to drywell differential pressure) to show the effect of small variations in pressure drop through the postulated break. EllR pump protection is based on the 0 paid curve. Also shown is the system head loss which will be experienced at 0 psid if the jet pump resistance is removed from the flow .
path. This will be the system head loss curve for a pump delivering flew to a broken recirculation loop.
Calculated system available NPSit curves are included for 130*F and 170*F suppression pool water temperatures at 0 psig suppression chamber (torus) pressure to show the effect of increasing water temperature during isolation or post-accident heatup. For all cases it is assumed that no credit for containment pressure above atmospheric can be taken when calculating the system available NPSil. Also shown is the benefit to be derived if credit could be taken for containment overpressure (5 poig shown) during short-term heatup.
Pump performance data is shown over the full range certified by the manu-facturer. Data extrapolations are indicated to facilitate prediction of performance; however, sustained pump operation at flow rates exceeding the certified data range would have to bc verified by additional testing.
In cach case, the predicted system flow will be that shown where the pump TDil curve crosses the system head loss curve. The pump NPSil required will be that corresponding to the predicted system flow. Stable pump operation
__u_ _ _ _ _ _ _ _ _ _ . . _ - _ -
. . o O 6 should be assumed if the pump NPSit required does not exceed the syntom NPSil available for the predicted system flow and torus temperature and pressure conditions.
\
In v'iew of the importance of assuring ad;quate NPSil at design runout, tests were performed on an instclied PJIR pump to d 2termine if additicnol margin was available in the certified NPSit curvas a;pplied by the pu:7 manufacturcr. The Hydraulic Institute procedures for determination of limiting suction requirements preacribe a 3 percent reduction in head at a given flow as a usually accepted criteria for prescribing NPSH.
Since pump protection and not loss of head is the principal concern during pump runout, tests were performed to determine the NPSH at which the onset of unacceptable pump vibration and audible cavitation could be detected.
These were considered more important evidences of acceptable short-tern operating limits.
The NPSil tests were performed at Browns Ferry on the 3A RllR pun:p while operating in the suppression pool cooling mode. Reduced suction preccures were achieved by throttling a gate valve in the pump suction line. The .
pump suction pressure was monitored by a Bourdon-type pressure transducer and the installed suction pressure gauge. Pump motor vibration was monitored by two accelerometers at the top of the motor (one in-line nnd one at right angics to the flow). The pump discharge pressure was monitored at the installed discharge pressure gauge at the local panel. The pressure transducer and the purp motor accelcrometer outputs were recorded on high speed strip charts.
The pump NPSil tests were performed at 8,000 and 10,000 gpm. The pump suction throttling was terminated in both cases before the " breakout point" (sudden and severc loss of discharge head) of the pump was reached. The maximum throttling point was selected on the basis of severe audible cavitation but acceptabic motor vibration for short-term pump operation. The results of these tests indicate that the certified pump NPSH required curves supplied by the pump manuf acturer may be reduced by an additional nine feet of head without adverse effects on short-term pump operation. A reduction of pump discharge head of 10 to 12 percent at a given flew rate was observed but this, does not
~
. . O O .
concern the purp protection prob 3cm. The results of the tests are given in Tabic 3.
, \
i 3.2 Units 1 and 2 Purp Protection The modified LPCI system head loss and availabic NPSH curves are shown for one and two pump operation without RHR pump orifice protection in Figures 1 and 2. These curves indicate that when one RHR pump is dalivering flow to the break as a design bases (af ter a single failure),
the pump flow rate will try to exceed 15,000 gpm and require , pump NPSH which is simply not availabic in the system. Continued availability of the pump is unlikely and the pressure boundary components and piping umy be jeopardized. The same situation applies for two pump operation except that it does not require a single failure for the pump flow to try to exceed 15,000 gpm.
Since it is a design basis to have one or two pump flow to the postulated racirculation line break, it is necessary to add RHR pump protection in j U
the form of a flow restricting orifice at the discharge of each RHR /
pump. Thc modified LPCI system head loss and available NPSH curves are shown for one and two pump operation with pump orifice protection in Figures 3 and 4 With pump orifice protection, one RHR pump will not exceed 12,000 gpm
- when delivering flow to the break. The calculated system NPSH 4 available exceeds the pump NPSH required until the torus water temperature reaches 152'F (about 100 seconds af ter the postulated break per reference 4). Above this temperaturci credit must be taken ,
for 5 portion of the proven margin on the pump NPSH required curve discussed in section 3.1. If operator action to align the pumps for c:ntainment cooling is assumed in 10 minutes, the torus temperature will ,
b2 about 160*F and the NPSH deficiency will only be approximately 2 feet.
This'is well within the 9 foot margin determined by tests to be available.
, When two RHR pumps are flowing to the break, each pump flow will net sxceed 11,500 gpm. The calculated system NPSH available will exceed 1
the pump NPSH required until the torus water temperature reaches nearly ;
i t
~ '
. . O O 170*F (over one hour af ter the postulated break per reference 4).
Operator action to trip the pumps is anticipated earlier in the heatup.
\
3.3 Unit 3 Pump Protection The system head loss and availabic NPSH curves for the original LPCI mode of RHR system operation are shown for four pump operation without RHR pump orifice protection in Figure 5. These curves indicate that when four RHR pumps are delivering flow to a recirculation line break (due to a sinel e failure), each pump flow rate will try to exceed 14,000 gpm and require pump NPSH which is significantly greater than that available in the system. Continued availability of the' pumps for long-term use is questionable and pressure boundacy leakage may develop.
Since it is a design basis to accept a loop selection logic failure, it is necessary to add RHR pump protection in the form of a flow restricting crifice at the discharge of each RHR pump. Even if the loop selection .-
icgic crror were excluded as a single failure by appropriate design, it would be necessary to consider the failure to close of the recirculation line discharge valve in the unbroken loop. This failure to close would provide a low resistance flow path to the broken loop about equivalent to a direct flow path to the break. The original LPCI h:ad loss and availabic NPSH curves for four pump operation with pump crifice protection are shown in Figure 6.
With pump orifice protection, the flow from cach of four RRR pueps will not exceed 12,300 gpm when delivering flow to the break. The czlculated system NPSH availabic exceed; the pump NPSH required until the torus . water temperature reaches 152*F (about 100 seconds af ter the postulated break per reference 4). Above this temperature, credit cust be taken for a portion of the proven margin.in the pump NPSH required curve as discussed in section 3.2. .
3.4 Verification of Emergency Core Cooling Flows The RHR pump protection orifice will restrict the ability of each RHR puup to deliver flow to the unbroken recirculation loop. The minimum total RHR
O O pump flow requirerents for adequate CCCS response are those given in Tabic 1. An inspection of Figures 3 and 4 indicate that Table 1 minimum flow objectives are exceeded on'a calculational basis for units 1 and 2.
A similar inspection of Figurc 7 indicates that the objectives are also exceeded on a calculational basis for unit 3.
To back up the calculations, tests were performed by TVA to verify the RilR system capability to meet the minimum pump flow requirements indicated in Table 1. For those tests, the MIR system was aligned in the LPCI mode with the RilR pumps taking suction from the suppression chamber and discharg-ing through their respective pump protection orifice plates directly to the reactor vessel. The tests were performed.on unit 3 but with the , unit 1 and 2 punp protection orifice plates installed where applicable. The results of these tests are given in Table 4 All flow requirements stated in Tabic 1 were exceeded with adequate margin. The corresponding small increase in NPS11 required to accommodate a larger than predicted flow to a broken loop was well within the margin in NPSH denonstrated for the pumps in section 3.1.
The test results compared favorably with the calculations. y The RilR pump NPSil available was also measured during the test to verify the NPSil calculational methods used as a basis for the system MPSil availabic curves shown in Figures 1 through 7. The results are given in Table 5.
The suppression pool tenverature during testing was 90*F. The test results verify that the calculational methods yield conservative results.
Periodic surveillance tests are proposed to reverify pump performance. This will be achieved by establishing equivalent pump flows and discharge pressures in the RilR test return line. Proposed RilR pump flow and discharge pressure acceptance criteria are given in Tabic 6.
4.0 PJIR PUMP TEST RETURN LINE FLOW AND FLOW INDl'CED VIBRATION The long-term containment cooling function. of the. R11R system, as discusacd in section 2.2, requires that the flow of two RilR pumps pass through their associated EllR heat exchangers and through a single RHR test return line to the suppression pool to provide the required cooling.
Both test return lines are not availabic for flow return because of certain postulated singic failurcs.
, , C) O "
The large flow in one 18-inch test return line has introduced troublenoine l
f1w-induced vibrations downstream of the 12-inch globe valve used to \
i throttic the flow. The problem gas been reported to the Nuc1 car Reguintory '
Commission (reference 5). The proposed solution has been to add a ficw restricting orifice downstream of the 12-inch globe valve and thereby provide significant back pressure on the valve. Tests have verified'that the back pressure effectively reduces downstream cavitation and its associated induced vibrations.
The addition of the RllR pump protection orifice has greatly reduced the allowabic pressure drop for the cavitation suppressing orifice.
As a result, this method of limiting flow induced vibration is no longer viabic in all cases as discussed below. In addition, the EllR punp protection orifice is limiting the availabic flow in the test r2 turn line.
4.1 Flow and Vibration Tests I
s.
Af ter installation of the RilR purp protection orifice and the RllR test
- j return line orifice plates identified in Table 2, several tests were l
parformed to determine their effect on test return line flow and flow induced vibration downstream of the 12-inch globe valve. Typical results of the flow tests are included in Tabic 7. Cencral Electric established &
c minimum flow requirement of 8,000 gpm for cach of two RilR pumps and associnced heat exchangers for long-term containment cooling on cach Broun3 Terry unit.
The readings given in Tabic 7 were taken from a flow meter in the test raturn line. On all units, the test return line flow meter consistently ruds somewhat lower than the flow meter in the corbined pump discharge lin2. The lower reading is thought to be due to the close proximity of ;
the flow clement to the upstream branch connection to a Tee.
S:;uthwest Research Institute (SwRI), under contract to TVA, investigated thn flow induced vibration effects. SwRI has not coupleted its report on tha tests, but preliminary results and recomendations are included below.
[
. . o O 11 4.2 Units 1 and 2 Plow and Vibration Verification The renults included in Table 7g for units 1 and 2 are typical of those observed during testing. They meet the minitaum flow requiremento established by Cencral Electric for two 131R putap operation but with littic or no margin.
Preliminary results on the vibration measurements perforced by SwRI indicate that cavitation and vibration 1cvels have been reduced very littic from those observed earlier without the pump or test return line orifice plates (original design) . The very large hole required in the test'rcturn line orifice plate because of the addition of the lutR pump protection orifice has rendered it relatively ineffective in reducing cavitation. The SwRI recommendation in to replace the 12-inch globe valve with a drag-type 1
valve (or equivalent) of appropriate size.
4.3 Unit 3 Flow and Vibration Verification b
l The results indicated in Table 7 for unit 3 are typical of those observed #
durinh testing. They exceed the minimum flow requirements established by Cencral Electric for two RHR pump operation but with limited margin.
Preliminary results on the vibration measurements performed by SwRI indicate that cavitation and vibration icvels have been reduced significantly from those observed earlier without the pump or test return line crifice plates (original design) . Ilowever, the relatively large hole required in the test return line orifice plate, because of the addition of the RHR pump protection orifice, has rendered it somewhat ineffective in reducing cavitation to very low levels. SwRI is also recommending the replacement of the 12-inch globe valve with a drag-type valve (or equivalent) of appropriate size on unit 3.
4.4 Proposed corrective Measures TVA is taking steps to replace the 12-inch globe valve in the test return line with a larger glebe valve and orifice plate combination or with a drag-type valve (or equivalent) on all Browns Ferry units. This cormitment
O O is intended to provide increased margf na in the test return line flow and reduce the flow induced vibrations downstream of the valve to low values.
Af ter im".a11ation of the new valves, flow and vibration measurer.cnts will be taken to verify the final con, figuration.
Prier to replacement of the valves, the test return lines on units 1 and 2 will not be operated at high flows (?_12,000 gpm) unlet.s required for special testing or cmcrgency operation. This restriction has been recommended by swr 1 and r.hould minimize flo4-induced vibration and associated fatigue.
In addition, the test return line orifice plates will be recoved from units 1 and 2 because of their ineffectiveness. A sparger will be installed on each test return line discharge to disperse the flow. Whenever feasible, both test return lines will be used to avoid prolonged high ficw operation.
The test return lines on unit 3 will be equipped with orifice plates and spargers. The ef fect of spargers on test return line flow and vibration was cound during testing to be inconsequential. They are being added to minimize suppression chamber flow induced vibration. There will be no operatin;;
restrictions on unit 3.
M.
5.0
SUMMARY
A'!D CONCLUSIONS The application of the I.PCI system modification to Browns Ferry units 1 and 2 with 7X7 fuel adds to the overall capability of the plant to continue opera-tion in a manner that ensures the health and safety of the public while pro-viding benefit in the productior, of electrical energy.
Adeqtate PJ1R pump protection against operation in excess of design runout has been provided and demonstrated for the modified units 1 and 2. PJiR test return line flow is within NSSS vendor requirements but flow induced vibration still exceeds long-tcrn objectives. Replacement valving will be provided which will add substantial margin to the flow and reduce the vibration to lower levels.
EllR pur.p protection has been added to Browns Ferry unit 3 to assure protection against R11R pump operation in excess of design runout. RilR test return line flow on unit 3 has been verified to be within NSSS vendor requirements and fleu induced vib ration is ~relatively low. Ilow eve r, replacement valving will also be provided on unit 3 to enhance the flow margin and reduce vibration f urther.
. . . O O 6.0 REFERE:CES 1 1. " Browns Ferry Nucicar Plant bnits 1 and 2 Emergency Core Cooling ,
Systems Low Pressure Coolant Injection Modifications for Perfornance l Improvement," Revision 1, February 1976 (Transmitted by letter from J. E. Cilleland, TVA, to B. C. Rusche, NRC, dated February 12, 1976).
2 " Browns Ferry Nuclear Plant Unit 3 Potential for RllR Purp Operaticn in Excess of Design Runout," DDR 224 Interim Report (Enclosure to j letter from J. E. Gilleland, TVA, to D. F. Rnuth, NRC, da'ted thrch 4,1976) . *
- 3. "Long-Term Containment Cooling Requirements for BWR/3 and BWR/4 ,
Plants," General Electric Plant Services Information Letter SlL ,
No.151, dated February 18, 1976.
- 4. Figure 14.6-12, " Torus Water Temperature Af ter LOCA," Browns ,
Ferry Nuclear Plant Final Safety Analysis Report.
- 5. " Browns Ferry Nuclear Plant Units 2 and 3 Potential for Failure of the Wcld Between the Yoke and Motor Mounting Plate for Flow Control Valves (FCV's) 74-58 and 74-72," DDN 191 Eighth Interim .
I Report (Enclosure to letter f rom J. E. Gilleland, TVA, to J. G. [
Davis, NRC, dated thrch 30, 1976).
P 4
e
~
14 TABLE 1 MINIMUM TOTAL PJIR PUf!P FLOW REQUIRE!!ENTS TGR ADEQUATI: LCCS RESPONSE (At 0 paid Reactor to Drywell)
BFNP BFNP Units 1 and 2 Unit 3 (Epm) _
(gpm)
One EllR Pump )
One PJIR Loop _) 10,800 One LPCI Path ) -
- Two PJiR Pumps ) *
- One RllR Loop ) 20,000 .
+
One LPCI Path )
Three RIIR Pumps) i ho RilR Loops ) 28,200 One LPCI Path )
TABLE 2 4
RilR PUMP PROTECTION AND TEST RETURN LINE SQUARE-EDGE ORIFICE PLATE IlOLE SIZES
, r BFNP BFNP Units 1 and 2 Unit 3
- (in.) (in.)
RilR Pump Protection Orifice 6.8 7.9 "
RHR Test Return Line Orifice 11.3 9.6 i 9
4 k
1 l
1 ..
t L
I t
i -
~
. . . O O
TABLE 3 MIR PUIG llPSil REQUIRED Flow Rate Certified Measured (imm) by Mfr. (f t) by TVA (ft) 8,000 25.0* 14.2 10,000 25.5 16.4 12,000 34.5 25.0*
- lly extrapolation.
TABLE 4 TOTAL RilR PUlIP FLOW i
TEST RESULTS Ill LPCI MODE (With Pump Protection Orifice Plates Installed)
BFNP BFNP Units 1 and 2* Unit 3 (gpm) (gpm)
One RilR Pump )
One lulR Loop ) 11,100 One LPCI Path )
Wo MIR Pumps )
One RilR Loop ) 20,900 One LPCI Path )
Three El!R Pumps)
Two RIIR Loops ) 30,500 One LPCI Path ) *
- Test performed on unit 3 using units 1 and 2 orifice plates.
16 TABLE 5 IGIR PUMP MP0ll AVAILABLE TEST RESULTS III LPCI MODE l
6At 90*F) i BIMP BlVP Units 1 6 2* Unit 3 Calc. (ft.)
Meas. (ft.) Calc. (ft.) Meas. (ft.)
One 131R Pump )
One PJIR Loop ) 42.0 45.2 One LPCI Path )
Two RilR Pump ) -
One RilR Loop )' 41.5 44.2 One LPCI Path ) +
'lbree RilR Pumps)
Two RilR Loops ) 41.8 44.1 One LPCI Path )
- Test performed on unit 3 using units 1 and 2 orifice plates, f
ii TABLE 6 'l PERIODIC Ri!R PIR!P SURVEILLANCE FLOW AND DISCllARGE PRESSURE ACCEPTANCE CRITERIA BIMP BFNP Units 1 and 2 Unit 3
( gpm) * (ps ig) * * (gpm)* (psir)**
One R'IR Pump )
One EllR Loop ) 19,000 h125 One Test Return Line )
ho EllR Pumps )
One RilR Loop ) 115,000 1200 f One Test Return Line )
Three EllR Pumps )
I Two RilR Loops ) {,
128,200 1160 .
ho Test Return Lines) I (equal flow split) !
. t
- Combined pump flow. .
ccPump dischargu pressure.
.- l l
. . . O O~ 17 TAllLE 7
'lYPICAL IUIR 1EST RETUI'Ji L1!!E FLOW (With Pump Protection and Test Return 1,ine Orifice Platen Installed)
DFllP BFliP U.its 1 and 2 Unit 3 (rpm) (ppm)
One Ri!R Pump )
One RilR Loop ) 10,500 11,500 One Test Return Line) -
Two RllR Pumps )
- One RilR Loop ) 16,000 16,5,00 One Test Return Line) 1 i.
.. . O O 45 y5
'~ i p
'm . . , . _ ' W_ 170o7 Ib P3,,IG)*
~~ N y ,
I W
~
t- r I30"F N(0 ps G)~
o 'J Yi 39 Q
35 d f}Y m
--~W 2 N m eo 170*F (o p3jg);
i i 25 25 5 l E' G 0
1000 1000
~
o 800 ,
800
=
_ SYSTEM WITH JET PUMPS E
- 600 - x
- 5. ---
N 600 -
SYSTEM WITHOUT JET PUMPS
- r:
e o PUMP TEST DATA RANGE g-
[3 g ..
DATA EXTRAPOLATION 400 - : TORUS -
yg ,
- REACTCR TO DRYWELL (20 PSID)i '-
200 / /
' .' 0 0 y /
~
.L
)i ]
0 2,500 5,00C 7,500 10,000 12,500 15,000 m
FLOW (GPM)
FIGURE I MODIFIED LPCI SYSTEM--ONE PUMP .
___ RUNNING _WITHOUT PUMP ORIFICE
-~
19 95 O -
O
,i ,
45 L' ~
](OPSIG) '
\
s 35 -
n -
35 (PSIG E
$ % [
i
=>
25 ' '
25 1
9 1000 IC00 800 %
N EC0 c
m.
600 -
SYSTEM WITil JET PUMPS 600 g ---
SYSTEM WITHOUT JET PUMPS g e o PUMP TEST DATA RANGE E -
DATA EXTRAPOLATION .
400 ~
- TORUS 40 '
(20 PSID)' " ,
REACTOR TO DRYWELL
- O PSID)" ,
y y 'y -
//
0 5,000 10,000 15,000 20,000 25,000 30,000 FLOW (GPM)
FIGURE 2 MODIFIED LPCI SYSTEM--TWO PUMPS RUNNING EACH WITHOUT PUMP ORIFICE -
o 45 9 n, 45 I
b {Q
^
- -- f '
g- ,
h g
a 35
'% s .
35 NI e
se d
a h
PSIG ::
F, N
52 25 (' 25 3 A O i
! 0
)
- 1000 1000 l
l 800 -
800
- I
^
600 -
SYSTEl4 WITH JET PUMPS 600 bl .
3 .q j t
SYSTEM WITHOUT JET PUMPS ,.
a e o PUMP TEST DATA RANGE o
' DATA EXTRAPOLATION (0PSID)4 / p k 400 - 4.- TORUS ,A *
, 400
#G (0 PSID)
,I' '
REACTOR TO DRYWELL
/
/< ,
/ . .
; /
200 [/ ' 200
- j -
0 s - 2,500 5,000 7,500 10,000 12,500 15,000 FLOW (GPM) FIGURE 3 MODIFIED LPCI SYSTEM--ONE PUMP
~~
RUNNING __WITH PUMP ORIFICE i
O O li5 X ._ W i 45 . t --- - . g70eF (5 pSIG " ' o I o ' U \
-kO fSIG)-
S
' u E; r~
c>. 35 * - U -
~
35 : M I/0op h
- a. 9 PSIG I
- E 25 ,
25 : 5 , u 0 1000 ' 1000 800
\ gc,,
(20PSID)o'- SYSTEM WITH JET PUMP 5 w. p --- SYSTEM WITHOUT JET PUMPS - (0 PSID)' ic 600 -
/ e c o w C00 g
{a PUMP TEST DATA RANGE {5 % DATA EXTRAPOLATION TORUS / c~
=
- OC
/ <
r 400 REACTOR TO DRYWELL - / .
'400 ?
/ '. C.
/
/
/ .
200 / s 200 p 0 x 5,000 10,000 15,000 20,000 25,000 30,000 FLOW (GPM) FIGURE.tl MODIFIED LPCI SYSTEM--TWO PUMPS RUNNING EACH W1TH PIIMP ORTFTPfp
' ~
(] 22 45 - 1, , 45 w w aso~. GZ' t
~ s a
. 35 -- ~
=
- 35 YO&~PSIG Jc- ~ ?
N E-25 ' - t 25 5
$ +
- IC00 1000
' e 800 '-
(20 PSID)O ~ 09 ! SYSTEM WiTH JET PUliPS } ? (o psig) r/ :
--- SYSTEM WITHOUT JET PUMPS 600 -c E'
3 PUMP TEST DATA RANGE '
- s 600 -
I DATA EXTRAPOLATION t n TORUS 400 A TOR TO DR M ' m 7 - i
/ '.
^
/ . \
/
(O PSID)4
- 200 '
,s' ~~ 'l
~#
0 ~ 10,000 20,000 30,000 40,000 50,000 60,000 FLOW (GPM) 8' 9 FIGURE 5 ORIGINAL'LPCI SYSTEM-- FOUR PUMPS i RUNNING EACH WITHOUT PUMP ORIFICE 2 k
45
. O O 45 170op' .
7 )- [ S (O PSIG N a E-l 35
\
%) *%
T g - 35 r cv -= 0 PSIG)v m % z
% N 5 h 25 (
25 E R e t = 1000 1000 (20PSID): 800 ~ ' / 800
-SYSTEi.1WITHJETPb l
--- SYSTEM WITHOUT JET PUMPS
{I 600 UMP TEST DATA RANGE m 600 [ x 7 ,,
.... DATA EXTRAPOLATION Z
/ j
[ x TORUS f' E REACTOR TO DRYWELL / . 400 -
- 40:,
./ '
, / (0 PSID)' ' * . -
.J.
/ .
/ '
/ . .;
/
200 y / 200 i 0 10,000 20,000 30,000 { 40,000 50,000 60,000 ' FLOW (GPM) i FIGURE G ORIGINAL LPCI SYSTEM--FOUR PUMPS RUNiiING EACl1 WITH PUMP ORIFICE .
.. _ - _ - _ - l
. l im. y l
. . (v~ v -l
~ '
% I7fo *
~ '
n ~W - ( (O PSIG). : P s
- d. l O
- 2 g 35 .
35 4 8 ~W G = 170o7 O Q: *f _ R x it F 25 L 25 El 0 i 1000 1000
' (20 PSID) 800 N ' i 800 i N '
(0 PSID) ; SYSTEM WITil JET PUMPS N 7 , - - - SYSTEM WITHOUT JET PUMPS E 600 - _/ gg3 e
- i. c a PUMP TEST DATA RANGE $
e -- S DATA EXTRAPOLATION / l s s TORUS j (0 PSID)'
~
400 -. -- #
- REACTOR TO DRYWELL t;GO
,7 ,
/ .
/
200 / r-, s '-
/_-
/
l 0< l 7,500 15,000 22,500 30,000 37,500 45,000 FLOW (GPM) ,
. FIGURE 7 ORIGINAL LPCI SYSTEM--THREE PUMPS RUNNING EACH WITH PUMP ORIFICE .
1 r J n 76 At DATE OF DOCUMINT DATE RECitVED NO-Tamacsrs VAL 13Y MFh0RITY 3/4/76 3/9/76 60202 \ L " MEMU REPOHT OTHtH 10 UNiG CC OTHEM X 4 IE ( ROY ) ACTH)N NECESSARY CONCuHRENCE ] DATE ANSWEN&D CLA$brF POSI UFFICE NO ACTION NECEssANY b COMMENT ] gy f ILE CODE - REG NO l * * *
. DisCHtPTION 4 Muse De Urirlana f=>i: REFENHED TO pfE RECf fvt0 SV DATE EBOWNS FEBRY NUCLEAR P1JJrf UNIT-3 REPORTABLE MFICIENCY POTENTIAL FOR RH1 P W OPEltATEGN SETouD RUNOUT CONDITION.
GOWER {/ ', R.A. MARTFIELD ENCLO&uME5 V COPIES FOR P91t,14 CAL PBR, NSIC, AIS MIE Start TO REC 1011AL C00RD111ATOIL Fog DISTRIBUTION. 1 i H t M A4.E E ( - PY SINT TO RECION II L: f I U S NUCLE AR NEGULATORY COMMISSION FOHsA N 3?65 1 i { l i t
- ,}}