ML20154A398

Analysis of Dresden Units 2 & 3 Operation W/One Relief Valve Out-of-Svc

ML20154A398
Person / Time
Site:	Dresden
Issue date:	09/28/1984
From:	Collingham R, Stout R, Swope D SIEMENS POWER CORP. (FORMERLY SIEMENS NUCLEAR POWER
To:
Shared Package
ML17201J425	List: ML20154A398 ML20154A402 ML20154A413
References
XN-NF-84-49, NUDOCS 8809120201
Download: ML20154A398 (24)

Text

I .

o .

L XN NF-84-49

!ssue Date: g/28/84 ANALYSIS OF DRESDEN UNITS 2 AND 3 OPERATION WITH ONE RELIEF VALVE OUT OF SERVICE Prepared by: - - 3 8A D. R. Swope BhR Safety Analysis -

Reviewed by: -[!II MaM#rt-R. E. Collinghara, Manager BWR Safety lpflalysis ,

81 Approved by: .

bi R. B. Stout, Manager Licensing & Safety Engineering Approved by: v '- . s ..f G. A. Sofer, Manager <

Fuel Engineering & Technical Services naa EDj(ON NUCLEAR COMPANY,Inc.

8809320201 000825 7 l

DR ADOCK 050 ,

I NUCLEAm REQULATORY CoheMitSION OtSCLAIMER HitPOaTANT isot!CE ninAnotwe CowTENTE AND USE OP TWI DOCUMENT.

PLE ASE READ CAAIPULLY_

This mePvineel resort om noched thsough reeerei and devWooment proggne sponsored by lamen NwWeer Company, Inc. It is beseg st>

mitted by Emmen Eclear to the USNAC es part of a ten cel contrb Dwdon M futilites esfoty enelysse by lieeneses of the USNRC whch utillas Eason Nueoenfeer6ceted resoed fuel er etter techncel semicoe .

provided by fanon Nuclear for lieht water power rom: tors and it is true I and correct to te best of laam Ntcleer's knowloege. Inkrmation, and belief. The enformenon consined heren may be used by the USNRC in is review of thee 'soort, and by lieansees er espicentt bebre the USNRC which are twommers of lanon Acieer in treir comoretreton of comolierce wie me USN RC's regwtonorsk WItfewt demgeory from tre foregoing reither Ennen Acteer nor any person adrg en is beheif:

A. Menos any werremry, empree or impl% wie respect to the oce#ecy, temph or Wwir.es of IPe iri%r-meten geritained in tNg gecumet, or Pet the bee Of any informaten, operatus, method, er proces diodoned in tNo docurrent wia met intnnge prrvetely omrted riens; or B. AenJmes eiy I.etrilitaes with respect to the wee ef. er be defeget renJitvg from the we et, any 6e%rmeten, em persedL methDd, er proces declosed in the docwment, X N PuI= F @, 764 l

1 i

- - , , . -.-.n- , - , -- - - - - , - -- -----,-,----,,e .~ ~

i l i XN-NF-84 49 TABLE OF CONTENTS Section Page

1.0 INTRODUCTION

............................................... 1 2.0

SUMMARY

.................................................... 3 2.1 LOS S OF COOL ANT ACC I DE NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.2 TRA NS I E NT A NAL Y S I S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 -

3.0 LOS S OF C OOL A NT A C C 10E NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3.1 ANALYTICAL APPR0ACH................................... 5 3.2 RESULTS............................................... 6 3.3 MAPLHGR MULTIPLIER.................................... 8 4.0 T RAN S I E NT A NAL Y S I S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 4.1 ANALYTICAL APPR0ACH.................................. 13 4.2 RESULTS.............................................. 15 4.3 CONCLUS10NS.......................................... 16

5.0 REFERENCES

................................................ 19 9

e .

11 XN-NF-84-49 LIST OF TA8LES Table Pg l 3.1 MAPLHGR for E.NC Fuel with Relief Valve Oct-of-Service.............................................. 9 -

3.2 MAPLHGR for GE Fuel with Relief Valve Ou t - o f - Se r v i c e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 t

I i

LIST OF FIGURES Figure P,aage, j 3.1 GE P8x8R Hot Assembly Heatup, 0.05 ft2 Break l With One Relief Valve Out-of-Service....................... 11 3.2 ENC Fuel Hot Assembly Heatup, 0.05 ft2 Break ,

With One Relief Valve Out-of-Service....................... .

12 .

4.1 MCPR Versus Time Load Rejection Without Bypass with One Relief Val ve Out-of-Service. . . . . . . . . . . . . . . . . . . . . . . 17 '

4.2 Steam Line Pressure at Relief Valves vs. Time During the LRWB Transient with all Relief  !

V a l v e s 0pe r a t i n g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 i

i 9

I b

r I

f

\ .

XN-NF-84-49

1.0 INTRODUCTION

in the analysis reported herein, the operation of Dresden Units 2 and 3 with one relief valve (RV) out-of-service is considered. The impact of such operation on the staximum average planar linear heat generation rate (MAPLHGR) and the minimum critical power ratio (MCPR) limits is detemined for each of the fuel types currently present in the reactors.

Each of the plants has a pilot actuated combination safety / relief valve (S/RV), four solenoid actuated Rys and safety valves. The purpose of the relief valves and the safety valves is to prevent overpressurizing of the reactor vessel. The relief valves are also designed to depressurize the reactor vessel if certain abnormal conditions occur so that core spray and LPCI systems can operate. These conditions might occur during plant transients or postulated accidents.

Only the relief valves and the relief function of the combination l safety / relief valve are considered to fail in this analysis. The potential effect of one RV out-of-service is to change the pressure response of the System during such a transient or accident. This may, in turn, impact the MAPLHGR or MCPR limits. The limiting ASME overpressurization transient analysis for these plants is the closure of the MS!V which did not take credit for any relief valve operation, only safety valve operation; thus, reanalysis of the overpressurizatici transient is not required to support a relief valve out-of-service.

Presented here then is the evaluation of the impact of operation with one RV out-of-service on the MAPLHGR and MCPR limits. The limiting postulated

. p F

.t t

. r 2 XN.NF-84 49 t

I small break accident is analyzed to evaluate the MAPLMGR limit since RVs do  ;

not actuate in large breaks. The limiting load rejection transient is ,

! analyzed for the MCPR limit evaluation.

1 .

r

.i

!- E i

I

.]

i

, r k

I l

I I

I Il 11 l

r

)

i r

) i 1

i 1  !

I i l i r I t i

I i,

, l' e .,

i

[

. t 3 XN-NF-84 49

(

. 2.0

SUMMARY

f 2.1 LOSS OF COOLANT ACCIDENT [

Small break LOCA analyses were separately performed for ENC 8x8 fuel and GE P8x8R fuel in the Dresden Units 2 and 3 to determine MAPLHGR limits during operation with one relief valve out of-service. A MAPLHGR multiplier of 0.891 was calculated for ENC fuel, which, if applied to the MAPLHGR limits [

on ENC fuel for normal operation whenever the plant is operated with one -

relief valve out-of-service, will assure that 10 CFR 50.46 criteria are met in -

J the event of a LOCA. Table 3.1 presents the resultirg MAPLHGR limits for ENC

fuel. L For GE P8x8R fuel, a MAPLHGR multiplier of 0.96, when applied to the lower MAPLHGR limits on GE fuel, was confirmed to provide assurance of

compliance with the 10 CFR 50.46 criteria. Table 3.2 presents the resulting -

I l MAPLHGR limits for GE fuel.

l 1

. A comparison of Tables 3.1 and 3.2 indicate that when the MAPLhGR , j multiplier for ENC fuel with one relief valve out-of-service is applied to the

• higher ENC MAPLHGR limits, it results in a higher allowed MAPLHGR than for the l u

! GE fuel under the same conditions. Note that limits on GE fuel are expressed  !

t a

as a function of average planar exposure whereas limits on EhC fuel are [

t expressed as a function of bundle average exposure.  !

I i

2.2 TRANS!ENT ANALYSIS t

The load rejection transient event, which yields the most limiting j i

thermal margin with all RV's in service, was analyzed to determine the impact of operation of Dresden Units 2 or 3 with one RV out-of service. There was no ,

i 4

impact on thermal margin (MCPR) limits because relief valve pressure settings i

]

j l I I 1

t 4 XN.W .84 49 were not attained until after the time of minimum MCPR. A minimal ispect on peak pressure was found and no pressure limits were exceeded. Thus, no technical specification changes are required to protect thermal margin criteria during such operation. .

l 4

o 5 XN NF-84 49 3.0 LOSS OF COOLANT ACCID (NT 3.1 ANALYTICAL APPROACH A potential for increase in the calculated peak cladding tem-perature (PCT) for a LOCA while operating with one RV out of service exists only if the RV is actuated to depressurire the coolant system. A large break LOCA will not be affected because the break itself rapidly reduces the sys' tem pressure and the Automatic Depressurization System (ADS), of which the RVs are a part, is not required to operate. During a small break of less than approximately 0.2 ft ,2 the ADS may be required to reduce system pressure to the point where the low pressure ECCS systems can operate. If the worst case single f ailure is assumed, in this case, of the High Pressure Coolant Injection system (HPCI), the transient is dominated by the time required to depressurize the system. With an RV out-of-service, this time will increase, resulting in a higher PCT than if all RVs were functioning, i

l A previous analysis, prepared by the General Electric Company (GE) for the Quad Cities Units 1 and 2(I), indicated that the most limiting small break with one RV out-of-service is a 0.05 ft2 recirculation line break with a f ailure of the HPCI. The GC calculations showed that MAPLNGR reductions are needed to assure compitance with the 22000F PCT limit.

The Quad Cities plants and the Dresden Units 2 and 3 are all BWR/3's with similar performance characteristics. The reactor vessel water level,

. system pressure and heat transfer coefficient (HTC) reported in the GE analysis for Quad Cities were judged to be applicable to the Dresden Units as boundary conditions for the small break LOCA calculation with cne RV out of-service. The NRC approved ENC EXEM/5WR Evaluation Model was applied for the e

6 XN-NF-84-49 fuel heatup calculation, using the system boundary conditions from the Quad Cities analysis. The system pressure was used directly in the heatup calculation, and to calculate the fluid saturation temperature for the heatup calculation. The water level was used to specify the quality at the plsne of interest for the heatup calculation, and the haat transfer coefficient was used directly.

The first task in this analysis was a heatup calculation of the GE fuel as a comparison of the ENC model with the GE model. This calculation was identical to the GE Quad Cities heatup calculation except the ENC EXEM/BWR heatup model, HUXY(2), was used. The R00EX2(3) fuel properties code was used to determine the exposed fuel rod properties at the start of the transient. An exposure of 15,000 FWO/MTM was used since this is the most limiting exposure for GE P8x8R fuel. The fuel rod properties thus obtained were input along with the boundary conditions from the GE Quad Cities report, to the HUXY code which performs the actual heatup calculation. The local power peaking distribution as predicted by XFYRE(4) for GE fuel at this exposure was used.

The second task in this analysis was to perform a similar heatup calculation of ENC fuel. This was accomplished in a manner identical to the above procedure, obtaining fuel properties from RODEX2 and the local pe=er distribution from an XFYRE calculation at an exposure of 15,000 KdD/MTM, and system boundary conditions from the GE Quad Cities riport.

3.2 RESULTS The system conditions of the limiting small break (l) are as follows. After break initiation (at zero time) and scram on high drywell pressure, the water level drops below the top of the active fuel at l

l

f i

7 XN.NF-84 49 approximately 260 s. The core level which would experience the highest PCT l uncovers at about 313 s. LPCI flow begins at 540 s., and rowetting of the plane of interest occurs at about 590 s. These event times determine the heat  ;

transfer coefficient (HTC) to be applied in the heatup analysis and correspond  ;

to the times when the HTC changed as reported by GE in Figure 2 of Reference 1: an HTC of 10,000 8tu/hr-f tIF is used until uncovery at 313 s., a HTC of 4 0.0 between 313 s. and 589 s., and an HTC of 25 Btu /hr-f2t .F af ter reflood at .

f 589 s. [

J Figure 3.1 shows the ENC calculation of PCT for GE fuel. Points t I

j from the GE calculation (Figure 2 of Reference 1) are plotted on Figure 3.1 for comparison. The GE and ENC calculations give essentially identical results f

between 0 and 313 s. when the heatup begins, and very good agreement through the heatup period and beyond the time of PCT. The PCT calculated by GE is .

1 approximately 22000F while that calculated by ENC is 21950F. Tne ENC l

i l calculation used the same MAPLHGR as the GE calculation (11.58 kw/ft). j Figure 3.2 shows the heatup calculation for ENC fuel at the same l 1

i MAPLHGR of 11.58 kW/ft. The PCT is 21730F, 220F below the PCT predicted by >

HUXY for GE fuel. The limiting ENC rod 22, is a lower powered rod than the j limiting GE rod 27 and has a lower initial stored energy than does the GE rod j l (1000Flowerfuelaveragetemperature). This difference in stored energy is  !

1 7

I only about 200F by the time of uncovery and then, due to higher power in the GE rod, increases again during the heatup to about 300F at the time of PCT. i The clad tem eratures are identical until the time of uncovery and tend to j

! follow the fuel average temperature during tH heatup. All ENC rods with f l Power similar to the limiting GE ro6 are acerer the canister wall (the  !

h t  ;

i I l I

r t

8 XN.NF 84-49 l

limiting GE rod being as far away as possible) and realize better radiative heat transfer during the latter part of the heatup. This can be seen in the plot of the clad temperature of the highest powered ENC rod 13, in Figure 3.2.

3.3 MAPLHGR MULTIPLIER A MAPLHGR multiplier for ENC fuel is calculated from the normal MAPLHGR(5) in the same manner as was done for GE fuel in Reference 1:

(Maxi LHGR) 3 For the 9x9 LTAs in Dresden Unit 2. MAPLHGR limits for normal operation were determined by inverse proportion to the number of fueled rods as comparsd with ENC 8x8 fuel at the same planar power. Thus, the MAPLHGR multiplier calculated sbove for ENC 8x8 fuel will also be applied to the 9x9 LTAs.

Appl3 ing this multiplier over the full range of exposure yields the results presented in Table 3.1 for ENC fuel.

Since the ENC heatup calculation for GE PSx8R fuel was virtually identical to that reported by GE in Reference 1, the hAPLHGR multipliers reported in Reference 1 are shown to also be valid for GE fuel in Dresden Units 2 and 3. These sultipliers are:

for GE 8x8 fuel. 0.99; for GE 8x8R fuel. 0.97; and for GE P8xBR fuel. 0.96.

Applying these multipliers over the range of exposures yields the resultsin

)

Table 3.2 for GE fuel.

9 XN NF-84-49 e

Table 3.1 MAPLHGR for ENC Fuel with Relief Valve Out-of-Serv,1ce 8x8 9x9 Sundle Average Normal Reduced Normal Reduced Exposure MAPLHGR MAPLHGR MAPLHGR MAPLHGR -

MWO/MTM kW/ft kW/ft kw/ft kw/ft 0 13.0 11.58 10.24 9.12 10000 13.0 11. 58 10.24 9.12 15000 13.0 11.58 10.24 9.12 ,

16000 12.85 11.45 10.12 9.01 20000 12.6 11.22 9.92 8.84 25000 11.95 10.64 9.41 8.38 30000 11.2 9.98 8.82 7.86 35000 10.45 9.31 8.23 7.33 k

10 XN-NF-84-4 9 Table 3.2 KAPLHGR for GE Fuel with Relief Valve Out-of-Service Average Planar Reduced 'HAf : NGR Ex o u t' Ei5 8x5R P8x5R g

200 10.99 11.29 11.00 1000 11.19 11.29 11.10 I i

5000 11.78 11.48 11.38 10000 11.98 11.58 11.58 15000 12.08 11.58 11.58 20000 11.88 11.38 11.38 25000 11.38 10.99 10.81 30000 10.49 10.41 10.24

• Note: An average planar exposure of 30000 MWD /MTM corresponds approximately to a bundle average exposure of 25000 MWD /MTM.

l ORESDEN 3 + HUXY + GE FUEL aE2 4 ssy a E 9 2 u 3Z D 14 35 1 W N 3F M RS M 11 4 1117 21ZB 24 25 as i 5 &E M E3 27 N 29 M s 13 12 14 as 31 W 3s Points from Calculatic, X 7 14 28 25 29 R 34 M e 25 tits M 13 M w

[NC Calculation 5 PIMz7

PIN 18 2195

{ 2000 -

=

I a

g 1500 -

t i

sa 9-E 1000 -

I d i

i i see)k...u -

- n- r.

n a _,

-eu -

-v y

a l T l

2

. . . . . . . . L l 75 15e 225 3e0 375 ese 525 see '"

TIME fSEC1 i

figure 3.1 E P8m8R Ibt Assembly Heatup 0.05 f12 Break i

With One Relief Valve Out of Service

l DRESDEN 3 + HUXY + ENC FUEL tt34 ssF e e s as u nt as as as '

s as as 17 as s u n ,

4 u sr u zs te n s 5 me ns ts p as as m s as as te :s ta st 2s l 7 14 as ts as m M 35 I

e as ti as te ss u m PIMar.

PIN 13

'173 2000 -

=

a )

g 1500 -

t w

e-

@ 1000 -

($ -

e _ . - - -

500 -

- ' ' ' - -  : : _ g e  ;

. . . . . . . . L l

75 150 225 300 375 450 525 See TIME tSEC1 ,

Figure 3.2 [NC ruel iht Assembly Heatup. 0.05 f t? Break With (he Relief Valve Out of Service .

l'

13 XN-NF 84-49 4.0 TRANS!ENT' ANALYSIS 4.1 ANALYT! CAL APPROACH Operation of Dresden Units 2 or 3 with one RV out-of service could affect the maximum change in the critical power ratio (ACPR) in the event of an abnormal operating transient. Previous ENC analyses (6) for the Dresden reactors found that the transient which gave the most limiting ACPR was the load rejection without bypass transient (LRWB). If an RV is out-of service there is the potential for a larger ACPR because of higher pressure and associated reactivity during the LRWB event. l The COTRANSA BWR plant transient analysis code was previously applied for an extensive study (6) of the LRWB event, including sensitivity studies relating the calculated ACPR to important input paramters. The COTRANSA input data was modified to analyze the LRWB event assuming the plant '

was being operated with an RV out-of-service. It was then possible to determine if the prediction of CPR was affected by the assumption of one RV out -o f-service .

A total of four COTRANSA calculations were made during this study.

The first calculation was made assuming that the $/RV was out-of-service with l respect to its relief mode. This valve has the highest capacity of all the RVs. It is set to open in its relief mode at 1149.7 psia, but if the relief r function is out-of service it will open in its safety mode at 1161.2 psia.

This calculation was made assuming the nominal values for the input paraNters .

describing the initial conditions prior to the transient and other boundary conditions important to the analysis of an LRWB transient.

[

l

14 xN.NF.84 49 In the second analysis, one of the two valves with the lowest opening pressure was assumed to f ail. The opening setting for this valve is 1129.7 psia. Nominal conditions were also used in this calculation.

A separate calculation was not made for a failure of either of the final two RVs. These valves open at the same pressure as the S/RV but have a lower capacity and will therefore have less impact on the transient than will the S/RY. They have the same espacity as the two valves with lowest opening pressure, but will open later and again have less impact.

The third and fourth calculations were identical to the first and second calculations respectively, except three parameters were varied as specified in Reference 7 to create a "worst case" situation in terms of the ACPR calculation. These three parateters are the rate of travel of the control rods, the delay time between the scram signal and the beginning of the movement of the control rods, and weighting of the relative reactivity feedback functions of the moderator density and the control rods. The minimum specified control red velocity 100 cm/s (nominal 140 cm/s), and the maximum specified scram delay, 293 msec (nominal 223 esec), were nodeled consistent with Reference 7. Also, the weighting of the relative reactivity feedback of the control rods was reduced 20% while the moderator density feedback function was increased 101. These changes would tend to increase the SCPR and cause the time at which the lowest CPR occurs to be later in the transient. Therefore, the effect of reduced RV capacity on the a CPR calculation was bounded.

15 XN NF 84 49 4.2 AESULTS In the previous studies of plant transients for Dresden Units 2 and 3, the time of lowest CPR for the LRW8 event was always around 1.0 s., and in no case was it later than 1.2 s. By contrast, the RVs started to open after 1.8 s. The four COTRANSA calculations made for this study, therefore, are identical to the corresponding calculations of the previous studies up until the relief valves begin to open (Figures 4.1 and 4.2). Since this time is well beyond the time of lowest CPR calculated for ENC 8x8 and 9x9 fuel and GE fuel, the previous HUXY-XCOBRA calculations of maxine PR apply also to cases with one RV out of service. The analyses herein . carried past the time of relief valve openings to confirm that the time of lowest CPR did not occur later in the transient.

1he limiting overpressurization transient for Dresden 2 and 3(6) is the M51V closure which did not take credit for the relief valve operation.

Thus, it does not need to be reeun. The p(ak pressures during the LRWB are presented here only for informational purposes. A peak pressure of 1270 psia at 3.75 s. occurred in the analysis with a11 the RVs operating normally. The peak pressure was 1271 plia at 3.87 s. for the case with the $/RV out of-service in its relief function and 1275 psia at 4.0 s. for the case with the low opening pressure RV out of service. Thus, no significant differences in the peak pressure were noted for the three analyses.

In the two worst case calculations, the peak pressures were sceenhat higher. A peak pressure of 1301 psia was predicted for the case with the $/RV out of service and a peak pressure of 1313 psia mas calculated in the analyses with the RV out of service, both at 4.0 s.

16 Art.NF.M.49 4

4.3 CMCLV5tMS These results indicate that with one RV out.cf. service there is no effect on &CPR calculated for a LRWS transient for Dresden Units 2 and 33 and therefore no tapact on the MCPR operating limit considerin.; all Nel types currently installed in these plants.

h i

~

2.6 Note: Curve for LRWB with all relief valves operating is identical to this curve.

2. 4 -

2. 2 -

2- '

tr i

j E

1. 8 -

q j 1. 6 -

l 1

1.1 -

1 i

1.2 d 1

5 i . . . . . i i i E"

O 0.5 1 1.5 2 2.5 3 3.5 4 1.5 5 i m I

TIME (SEC1  ?

l Figure 4.1 MCPR vs. Time, Load Rejection Without Bypass with i

One Relief Valve Out-of-Service j .

4 PRESSURE (PSIR)

E I. E. a. s W.

W.

M m i 1

I l$ $

n

.o - E- ,l!!

i'

, m E Er i" "

GI 6 e i '

3 setpoint reached I a

a N l a i i m setpoint reached

> u.

UY h/t j

=0 5

I

g. .. /

- i

s. I m

, m i

(

6#-98-dN-NX 8I 9

a

19 XN-NF-84-49

5.0 REFERENCES

1. "Analysis for Operation with One Relief Valve Out of Service for Quad Cities Units 1 and 2," NEDO-30032, General Electric Company, February, 1983.

2. "HUXY: A Generalized Multirod Heatup Code with 10 CFR 50 Appendix K Heatup Option User's Manual," XN-CC-33, Revision 1, Exxon Nuclear Company, November, 1975.

t

3. "RODEX2 Fuel Rod Thermal-Mechanical Response Evaluation Model," XN-NF-81-58, Rev. 2 Exxon Nuclear Company, January 1983. -

4. "XFYRE: A Multi-Group Two Dimensional Diffusion Theory Code for the Microscopic Depletion of Boiling Water Reactor Assemblies," XN-CC-37, Revision 1, Exxon Nuclear Company, April, 1980.

5. "Dresden Unit 3 Revised MAPLHGR Analysis Using the ENC EXEM Evaluation Model," XN-NF-81-75, Supplement 1, Exxon Nuclear Company, July 1983.

6. "Plant Transient Analysis for Dresden Unit 2 Cycle 9," XN-NF-82-84, Revision 1, Exxon Nuclear Company, November 1982.

7. Letter, L.C. O'Malley (ENC) to L.J. Bridges (CECO), LCO:224:84, "Para- -

meters for Analysis of Dresden Units 2 and 3 Operation with One Safety Relief Valve Out-of-Service," May 9, 1984.

O

7 s,....

XN-NF-84-49 Issue Date: 9/28/84 l ANALYSIS OF DRESDEN UNITS 2 AND 3 OPERATION WITH ONE RELIEF VALVE OUT-OF-SERVICE .

DISTRIBUTION JC Chandler RE Collingham SE Jensen JL Maryott JN Morga

. 1 GA Sofer R8 Stout DR Swope LC O'Malley/ Ceco (6)

Documnt Control e

• 1