ML19210C779

Forwards Evaluation of OPS Nov 1979 Response to ACRS Re Floating Nuclear Plant Core Ladle Design

ML19210C779
Person / Time
Site:	Atlantic Nuclear Power Plant
Issue date:	11/02/1979
From:	Vassallo D Office of Nuclear Reactor Regulation
To:	Fraley R Advisory Committee on Reactor Safeguards
References
NUDOCS 7911200095
Download: ML19210C779 (66)

Text

{{#Wiki_filter:TCEA -sa ate + o UNITED sTATFs 8'i ,'i NUCLEAR REGULATORY CMMISSION C WASmNG TON. D. C. 20555 ,,. 1 ~h ' /

• ?*

\\,,,',, November 2, 1979 Docket No. STN 50-437 MEMORANDUM FOR: Raymond F. Fraley, Executive Director, Advisory Cormiittee On Reactor Safeguards FROM: D. B. Vassallo, Acting Director, Division of Project Management, NRR

SUBJECT:

ACRS LETTER OF JULY 25, 1979 - FNP CORE LADLE DESIGN This is in response to your July 25, 1979 memorandun to me indicating that the Comittee desired additional infomation concerning the Floating Nuclear Plant core ladle design and related matters. Of fshore Pover Systems has addressed the contents of your July 25, 1979 memorandum in their September 14, 1979 letter to us. We have reviewed this information and the Enclosure to this letter provides the Comittee with our comments. I understand that an ACRS Subcommittee meeting on the FNP core ladle design has been scheduled for November 17, 1979. The NRC staf f wishes to proceed to conclude our review of this matter as promptly as possible. Therefore we would like to have the concept and preliminary design of the core ladle reviewed by the full Comittee at the December 1979 meeting, so that we could consider - " Comittee input and ccments as part of our review effort. D. B. Vassallo, Acting Director Division of Project Management Office of Nuclear Reactor Regulation

Enclosure:

Staff Review and Evaluation Of Offshore Power Systems Response To ACRS Letter Of July 25,1979 1366 056 096 7911200

STAFF REVIEW AND EVALUATION OF 0FFSHORE POWER SYSTEMS RESPONSE TO ACRS LETTER OF JULY 25, 1979 NOVEMBER 1979 1366 057

TABLE OF CONTENTS Page No. Introduction y Responses A. Items Related to the Imoact That the Core Ladle I Will Have on Other Containment Structure 1. Calculate the fraction of decay heat radiated from 1 the cool for the proposed design 2. Calculate the effects of heat radiation in Item 1 11 on the rate of (a) disintegration and collapse of exposed concrete, (b) disintegration and collapse or melting of concrete behind the six-inch ~ magnesite brick wall and (c) collapse of steel from the reactor cavity 3. Discuss the consequences of Item 2 with respect 12 to (a) loss of integrity of superstructures, (b) loss of hearth capacity, (c) impact resistance of the hearth and its supports, and (d) integrity of structural steel members 4. Discuss the stability of the six-inch magnesite 15 brick wall above the hearth level with respect to (a) loss of brick by spalling, (b) differential motion with respect to the hearth concrete walls and anchors, (c) loss of concrete behind the wall by spalling, disintegration and melting at cal-culated temperatures, or at temperatures indicated in Figure IV-6 of OPS Topical Report No. 36A59, and (d) slagging reaction between the brick walls and melted concrete 5. Discuss the fluxing of magnesite brick by silicious 17 material falling into the hearth 6. Discuss the properties and merits of basalt as a 17 concrete aggm gate 18 7. Discuss the possibility of the heat flux being higher on the sides of the molten mass than on the bottcm (FRG conclusion for concrete melt) with melting going horizontally faster than vertically i 1366 058

page No. B. Items Related To Three Mile Island Accident 1. Discuss the possibility of the Upper Head Injection (UHI) 20 System releasing nitrogen into the primary system and impeding the ability to establish or maintain natural circulation 2. Discuss the acceptability of the single failure 22 criterion 3. Discuss the timed sequences of events upon the loss 23 of all AC power before core damage will result 4. Discuss the reliability of the auxiliary feedwater 25 system 5. Discuss how H buildup in the ice condenser con-26 tainmentisd3aitwithfoilowingaraIevent and following a core melt 6. Discuss how the FNP compensates for the diffi-34 culty, due to the remote location and the lack of space available in improvising new systems and techniques in case of an accident 7. Discuss how one faces lack of flexibility for 35 design change due to the compactness and lack of available space on the FNP C. Items Concerning the Effects of Chancino Base Mat 36 Mater 1a1 1. Discuss the effects of changing the bast mat from 36 concrete to magnesium on the probability of a major air release during a core melt accident. C'scuss the comparisons of probabilities and dose levels for air releases associated with concrete and magnesium oxide during a core melt accident 2. Discuss the consideration given to the use of a 40 vented containment. Discuss the consideration given to the use of sea water for venting and/or cooling a molten core 3. Discuss the change in position for allowing the 42 FNP to be placed on riverine and estuarine sites. Has the proposed installation of the core ladle changed the NRC staff's position on this matter, if so, why? What actions and in wnat time period, are censidered practical to isolate the core for a riverine or estuarine site? 1366 059 11 i

i Pace No. 4 Discuss the NRC Staff's position that the FNP Core 46 Ladle is considered an environmental issue and not a safety issue D. Additional Information Reouested from the NRC Staff 51 1. Provide available information on the Sandia 100 51 plant liquid pathway study 2. Provide available information on the WASH-1400 51 type study of the ice condenser type plant, along with a comparison for non-ice condenser type plants iii 1366 0'0

TABLES Page No. 1. Pool Surface Temperature Histories 6 2. Scoping Study using MELSAC 7 FIGURES l. Pool Temperature Histories Predicted by MELSAC 8 2. Fraction of Integrated Decay Heat Absorbed by 9 Walls and Steel Above Mg0 Ladle as a Function of Time After Start of Mg0-Melt Interaction 3. Erosion Depth as a Function of Time After Start 10 of Mg0-Melt Interaction For Vertical and Horizontal Erosion APPENDICES A. ACRS July 25, 1979 Letter 52 8. 0PS September 14,1979 (without attachment) 55 C. Near Term Requests For Improving Emergency Preparedness 57 D. References 59 iv 1366 041

s INTRODUCTION ~ On June 27, 1979 the ACRS Subcommittee responsible for the review of the Offshore Power Systems application to manufacture eight floating nuclear plants, met with the applicant and the staff to review the FNP core ladle design as submitted by the applicant in Topical Report No. 36A59, "FNP Core Ladle Design and Safety Evaluation." This conceptual design feature was proposed by the applicant in response to the staff's environmental assessment (FES, Part III) and the LPGS Report (NUREG-0440). Subsequent to the ACRS Subcommittee meeting, the ACRS at its July 1979 meeting issued a letter (see Appendix A) requesting additional information from the applicant and an evaluation of the response by the staff. By letter dated September 14, 1979 (see Appendix B) the applicant provided additional information which the staff has reviewed and evaluated. The staff response utilizes the format of the ACRS July 1979 letter. e V 1366 062

1 A. Items Related To The Impact That The Core Ladle Will Have On Other Containment Structures 1. Calculate the fraction of decay heat radiated from the pool for the proposed design. STAFF RESPONSE The OPS response is based on calculations which assune the radiation losses from the pool surface to be decouoled from the pool heat transfer processes. It is recognized by OPS that the problem is highly coupled and a computer program is being developed at CPS to solve the couoled problem. The MELSAC Code (I) ~ which is being developed by NRC staff consultants at the Brookhaven National Laboratory (BNL), solves the coupled problem and the results obtained from MELSAC differ from the calculations and assumotions made at OPS. A major difference between the OPS and the staff results relates to the pool surface temperature histories. OPS assumes the two surface temperature histories shown in Table 1. These temperature histories differ appreciably from the predictions of the MELSAC code. Typical results from MELSAC are shown in Figure 1 for two different pool heat transfer correlations. Case 10* uses the downward heat transfer coefficient given by the Kulacki, Goldstein(3) correlation for an internal heated molten pool. Case 6* models the dentity driven heat transfer coefficient (4), which arises because the sacrificial material being melted is less dense than the pool material, and a buoyancy-driven motion is induced. ' Case numoers refer to cases reported in Reference 2. Cases A, 8 and C most closely represent the ladie and cavity configuration described in Reference 5. The assumptions used in Cases A, B and C are included in Table 2. 1366 063

2 The temperatures shown in Figure 1 are the bulk pool temperature; the uoper pool surface temperature is typically 80 to 100 K below this value. The effect of the higher surface temoerature history credicted by MELSAC is to transfer more heat to the upper structures. A typical run (Case B) using MELSAC is compared with the results provided by OPS in Figure 2. After one day the fraction of heat stored in the walls and vessel ranges fran 0.24 to 0.46 using OPS assumptions, whereas MELSAC predicts the heat stored to be 0.76. Simi'arly, after 5.79 days (melt-through for Case 8) MELSAC predicts a fraction of 0.5 in the walls and vessel compared with 0.11 and 0.17 at 6 days using OPS assumptions. The result of this additional heat transfer to the walls is that MELSAC predicts (Case A) that the concrete in the cavity wall (24" Mg0, 3" gap, 21" concrete) will reach its decomposition temperature (1473 K) after only 1 day. In order to determine a wall configuration that protects the concrete (<1473 K) and steel bulkheads (<810 K) beyond the two-day period, a scoping study was carried out and the results are prasented in Table 2. MELSAC, predicts (Case B) that a Mc0 -li of 36" thick would be required to protect the concrete and steel for 2 days, given the input assumptions listed in Table 2. The times to melt the whole of the reactor vessel by thermal radiation from the molten pool surface varied between 0.5 to 4 days under the assumptions made by OPS. In Table 2, we have included the MELSAC estimate of reactor vessel melting, which is about 1.8 days. The model used in MELSAC is different from that assumed by OPS and is also exposed to the higher pool surface temoerature history. In MELSAC the reactor vessel is modeled as a series of connected masses. Heat transfer between each of i366 36A

I f 3 the vessel masses.s by conduction. However, the MELSAC prediction is not inconsistent with the range of reactor vessel melting suggested by OPS. ihe erosion rates presented by OPS are decoupled from the upward heat transfer and were simply obtained by using a constant fraction of the volumetric heat capacity. The erosion rate predicted by MELSAC (for the wall configuration which protects the concrete and steel, namely, Case B) is comoared with the OPS erosion rates in Figure 3. The MELSAC erosion rate clearly shows the coupled effect. At early times the heat transfer to the upoer structures is high (because of the large temoerature ' differences) allowing only a small fraction (F) of the decay heat to be directed into the Mg0. At later times, the upp'er structures are at higher temperatures and the upward heat transfer is reduced allowing half (F = 0.5) of the decay heat to be directed into the Mg0 at the point of melt-through (5.79 days). This is consistent with Figure 2 in which the fraction of decay heat stored in the walls is predicted to be 0.5 at 5.79 days. It should be noted frem Table 2 that MELSAC Case 3 predicts that the 25" of Mg0 will protect the concrete and steel for 2 days, whereas the core ladie will hold-up the molten pool for 5.79 days. Under the assumptions of the MELSAC code, and using the above wall configuration, considerable damage would be expected to the upper structures in the reactor cavity after the 2 day period and before the molten pool is released frcm the ladie at 5.79 days. Finally, our connents must ce qualified as they are based on a first version of the MELSAC code. At oresent MELSAC asstsnes the molten pool to be initially pure U0. As the meltfront moves into the sacrificial bed, the code comoutes 2 1366 065

4 the dilution of the UO2 with molten Mg0. However, the dilution of the pool by steel and zircalloy cladding (both of which will certainly be present) is not presently modeled. The omission of molten steel addition to the pool from the reactor vessel is significant as the mass of steel in the reactor vessel is 1.5 x 106 lb compared with 0.22 x'106 lb of UO - 2 It is not clear at the present time what the total effect of introducing such a large quantity of molten steel will have on the pool conditions. One ~ effect may be to decrease the pool temperature. Whether the resulting mixture will remain molten and in what configuration (layered or mixed) is not currently known. A simple calculation, which brings the pool into thermal equilibrium with the molten steel just at the time vessel melting is complete, 0 yields an equilibrium temperature of about 2000 K. This is considerably below the melting temaerature of the UO -Mg0 binary system, but both far above the 2 steel melting temperature and far below the steel boiling temoerature. Intro-ducing molten steel into the pool at the rate it is being melted may have the effect of lowering the pool temperature, thus decreasing the radiative heat transfer to the vessel and its resulting melting rate. On the other hand, quenching the pool with molten steel may drop the pool temcerature below the Mg0 melting tempe' ature. If penetration of the ladle stcos, then dilution of the pool with molten Mg0 would also stop. The question then arises as to whether or not the quenching effect of the vessel steel would ccrnpensate for the loss of the dilution previously provided by the molten Mg0. The 2000 K equilibrium temperature is clearly a lower limit. The pool temoerature history, allowing for the effect of steel dilution, cannot yet be predicted but would be between the present MELSAC predictions and 2000 k. 1366 066

5 Future work on MELSAC will be Prected to addressing the effect of dilution the molten pool with steel and zircalloy. We do, however, consider that the current predictions by MELSAC represent early times with respect to ladle penetration, heating of the structures in the cavity and melting the reactor vessel. The effect of diluting the pool with steel and zircalloy would tend to increase the time scale of the above events. However, cuantifying the changes in time scale at this stage is extremely difficult. The ladie penetration times predicted by MELSAC are also dependent on the assumed equal lateral and downward heat transfer correlations (see coments on Question a.7) If evidence becomes available to suggest that penetration should be faster in either direction, then the MELSAC code could easily be modified. The conclusions of the staff, regardir.J this que: tion and the remainder of the questions in item a of the Connittco's July 20th !cttor, is that the ladle concept is feasible and can be engineered to provide retention of a molten core for a period of time in the range of two days to one week. As noted above, the applicant is in the process of developing a coupled calculation model. Any significant differences between that model and the staff's model can be resolved during the early phases of the final design. Once a calculational model is agreed upon, the ladle configuratior can be optimized for the available space to provide the largest possible core retention time considering &ll the factors raised in items a.2 through a.7 of the Comittee's letter of July 25, 1979. .i L366 067 g g

• ,w

6 TABIZ 1 PCCL SC*6 ACE TEGERA~tBE HISICRIES TIME (DAYS) TEMPEPATEZ I. Sardia Estimate 0 0 4712 F (2600 Q l 3632 F (2000 C) 2 3524 F (1940 C) 4 3308 F (1820*C) 6 3092 F (1700 Q TIME (CAYS) TEMPE?An 2E II. CPS (Black Sody) Estimate 0 3641 F (2005 Q 0 1 2394 F (1312 C) '~l - 2 2232 ? (1222 Q 4 2092 F (1144 C) U 6 1952 F (1067 Q + 4cPacnvece Facm Asstaswcr @ i366. 635

7 %BLE 2. SCOPIN C S=ruDY DSING MELggt "IT4 cWuut Twicio*EST vmt ra " Tim e To " Timer ro ' Time 70 $ TEEL @ g-CF m3C OF @CII PIN ettA M STRAT miEL t-mrtT rgmP W eN wp L tN WAL( CORS M M N' VE$5EL CCN cagTE STa4T OF LAsi.E VEISEb CcNC.mtu (D mVS) (HILS) (W) A o Go's 0 'c% ozs 2 44 1 7S lo+ 3'o B 0 914+ 0304g y ya) 2 53 1-23 2.*01 %"70 C l 0468 o t s14-6 co 2 r3 ln 24s t 146 t-AssumeTicM S :- m Ass op vos tot on WS mass op ve5EL G 7c 32.'s Kg L. ant.c TwicwNETs 16m lhtLTiaJ4 rEme fn C 3 0 T3 '< 3 mELTe%(, rEmp cometfTE 14T3 K FDot H/ r anted ON DENstTV DCivCW ce s.a. c L a n c N Pool Sud# ace.- 3Te.vv. nn.E1 EFFEC rsvE m issav m C S Mk'MT T<1ANSTER BE%ND vc5tQ 3 WA LL $ LADLE SY Nm"TVAA L ccNva.new. 1366d69

~ m 4 ) 6 E C e S A A S (C )0 L E n 8 g R r s Y E E I l F S S B S N C D A ( E R T T M

c T

E u I n W T E E s W H f o. l i N o E E t c v n )s o 4 4 r D -s u. l n Y u a I a I T c a. m e t s ( t N t s E u e n D K m n a u e r m e T t oo P 1 l I Ea u c / i F N ~ [ \\'y. / o 0 0 0 0 0 0 0 0 0 0 4 2 O 8 6 3 3 2 2 ,,YG g 6 s- ,oO 3 N sr r_ D -.OC . C i

4 ^* + r. = D**D *]D'3'l 9 o Ju 1..\\ .3 oe Fi avde.1 ,,,4rner,en a e summa,w n nrav Mn-wr sa nonen w uwnes cvmrt Asa n,ss- . r.- - A.3 M MU AJc,.T aans Co-r //www Arr a M a dmz Ga= nuwa-m /hz, xMr ION . \\ Mg G I,,neG4.2- .. _ ' ~ ~_._1._ BNi _ccur_lLResults l I I i !I i .u : : i O MEb5Nb ~i.;W ALL. I i m~ Q MEl.Snc ! 'S7~EE L ' i ' ~ ^ ~ 1 a A16LSRC - s War.L + S7"EGL. l l r;, ! ~ ,i i i l l' OPS CURVE I OPS CURVE II ~ i tw -/ +-n w,rre Jr" Y

: : c ' _," ( 8 3

)_ / / \\ ,,a. f / _.g ~~~ n .7 ~ - g ~ '"'M "{' )&

• ^

l6-V/ E$\\- 382 5_._...__._ W\\ .A

/

i .c, e N. M \\.... = .2 o h Temo e r w e m 'W OPS du e s-I i _._. 'b.. n, w. cir.*Ath w'r.) E D A,Y. 363%Y(3,000%) ! l I 4" s c49.ssaycnoew hm! i!., t r = m a 3 r w m m._co m g.9 .W. tum R ~ ni<t=m er-r=o {+' at_{ } <d f ~' '~ ~ 'Q(O s _ _ _s e. 2 Aoi').i E' p qg -%f ,g b

• 2 O$d k

b f fjf g 4 ,6-~ ~ ! i i f. ;~'-.- Aj e;,-ti.J mfir1D4cr7eas 217'4475 . ; et 'l $5 ' I ! i g-1 3 =*. ,i;: i

• j i

l ~ ' I. .I i I i , _.s. ~ TMf A VJf ff.OT* A M*. ., QRg}lCf**)? tBF*dn $) -1 t! - ) ~ ~* ! - l l! l l ,2) AYS -P ! /i.M. 7,; i ,r I 4 4 i. i;, a, i - c i-i 1 .fo ' fo= 2 . _7C /0 s s 4 CPC opuc.211%) Mm d@ E24 CE. h) }66 Ob ee

10 54 vtt. Y ' son _s >w,,rm a A suacnau or, w r arran -r or su.a. un,- JM AL'*9*1oAl C o n VB*'D'P*1/ A/ AdB A/_A_99 7,nA/""f*AL /=*CM FY f s g I f./my-#

• i t

8 l l f NA dB' MAP A WAA e JIT M S A'7" 4'T*ADT* Af fMWWEA,L*9 A Af 4 ,L.. . z .. w ::" :== x:= ::::.,:::*:.- :.=,; i i f i .T,, Vol.uMdTMIC J/KA7* CAPAC/TY 05 A*fe C l's A J*U^/47*104 of _&4. i Coed 90dITJ ed.' / M/77At. VAA. add' e 3. 9 t rQ S 537'll/eo 2 A DM12nd ad

a. _n_A4* ML u v rM M, A W

/ / ~' t o/ 8 , /. / o.8/ M / FY s / / S / / / / /

o. V 9?

/ / FY / / / / / / / / / / // ~ / / /

s. F

~ ' n / / /. FY/ / / / / /

• ~~ ~

/ / / / N-(_ mt.uc case s ,7,,,.-. ll / / / ., // / / __/ ~~ ~ // / / / ~ "~ / ~ -- /// / ,/ r Afrza ram es s<.a-uc .w x2Ac. art cay.O 4 ? -- z. a +.. -. p --. s -..7 p) i 566 372

• - car a..~.a. mm area. men

?@@ ID

11 2. Calculate the effects of heat radiation in Item 1 on the rate of: (a) disintegration and collapse of exposed concrete STAFF RESPONSE We agree with the response but with qualifications discussed by us in Question a.1, namely, that we do not accept the pool surface histories and recommended a wall configuration with at least 36" of Mg0 protection. (b) disintergration and collapse or melting of concrete behind the six-inch magnesite brick wall STAFF RESPONSE ~We agree with the response regarding the fact that the walls can be protected for two days using suitable high temperature insulating brick. The melting point of basalt aggregate is exactly the value we have used in MELSAC. We do not, however, believe that the config tration of 24" Mg0 and 21" Basaltic concrete will be sufficient to protect the concrete (refer to Question a.1). We recomend at least 26" of Mg0. (c) collapse of steel from the reactor cavity We agree with the response and note that the MELSAC prediction for vessel melting of 1.8 days is within the range of 0.5 to 4 days suggested by OPS. The addition of this steel to the pool cannot be modeled in the present version of MELSAC and the effect of dilution of the molten pool by this amount of molten steel was discussed as part of the coments to Question a.1. We again emphasize that the current predictions of MELSAC, therefore, represent early times with respect to penetration, heating up of the structures in the cavity, and melting of the reactor vessel. 1366 173

12 3. Discuss the consequences of Item 2 with respect to: (a) loss of integrity of superstructures STAFF RESPONSE The applicant has stated that proper protective barriers can be utilized to protect structures above the reactor vessel which may be affected by upheating following the vessel melt-out. The protective barriers include high temperature insulating brick (Mg0, ZO r A10, or possibly ceramic fibres (Fibrefax). These barriers 2 23 will prevent the concrete from disintegrating, collapsing or melting ~~ when properly desigr:ed. The applicant proposes a shield that would limit the maximum temperature of the concrete to less than 2200*F, which is celow the melting temperature of basaltic aggregate. However, the staff requires that the applicant prove that the melting temperature of basaltic aggregate is the controlling factor and not those of other components of the in-place concrete. With regard to the steel struc-tures, the applicant plans to limit the surface temperature of the primary steel components to 1000*F. Also, for steel components subject to high thermally induced stresses the temperature limit will be reduced. This position meets the requirements of the American Institute of Steel Construction. However, the applicant has not identified any specific structure that will require the above specified. protection. The app 11-cant plans to identify the specific protection that will assure the integrity of the superstructures from radiation heat during the final design phase of the FNP. 1366 174'

13 Based on the evaluation of the above material, we conclude that the applicant has providedan adequate preliminary design. This design provides enough information to give reasonable assurance that the final design will satisfy all of our requirements. However, our final approval is subject to our review of the final design. '(b) loss of hearth capccity STAFF RESPONSE The applicant has increased the ladle capacity to over four times the previous capacity and has identified additional space where a properly designed supplementary ladle can be placed to take care of any overflow from the basic core ladle. These actions should dismiss any concerns on the adequacy of the ladle capacity. We consider the core ladle support structures to be acceptable since these structures have been redesigned to comply with the structural acceptance criteria as out-lined in the Standard Review Plan. However, the applicant should demonstrate in the final design, the adequacy of the structural systems supporting the ladle in order to determine the time dependent structural capacity to support the actual ladle configuration. Basically, the applicant should demonstrate that the structural members supporting the core ladle will not fail prior to any failure of the core ladle by melt-through of its contents. 1366 i)75

14 (c)' impact resistance of the hearth and its supports STAFF RESPONSE The applicant has performed analyses with the following postulated new loading conditions: (1) Reactor Vessel 30ttom Head Impact, and (2) Upper Reactor Vessel Impact, in order to consider the appropriate impactive loads in the evaluation of the core ladle. The analyses indicated that case (1) noted.bove, controls the design. The analyses for this case showed that the ladle can resist the impact load without failure. We agree with the applicant's conclusion. However, the applicant should document in the final design the adequacy of the struc-tures supporting the ladle for their capacity to resist these impactive loads. We find the approach used by the applicant acceptable, subject to our review of the final design details. (d) integrity of structural steel members STAFF RESPONSE The applicant plans to shield all of the primary structural steel mem-bers within the reactor cavity (bulkheads, floor and deck) frem energy radiated from the core ladle. The protective barriers are designed to reduce the temperature to a level that will avoid rapid deterioration of the steel members or excessive thermally induced stresses. The applicant plans to follow the require-ments of the American Institute of Steel Construction. This design code considers steel components fire resistive if the average tempera-ture of the steel members does not exceed 1000*F. The applicant plans to limit this temperature to the surface temperature of the steel mem-bers and plans to lowe this limit for areas of possible high thermally induced stresses. 1366 076

15 4. Discuss the stability of the six-inch magnesite brick wall above the hearth level with respect to: (a) loss of brick by spalling STAFF RESPONSE The response to the question compares magnesia and zirconia brick and can be construed as implying that Zr02 is more resistant to spallation than Mg0. Although it is true that magnesia brick does have a relatively high rate of expansion among refractories, it should be noted that zircenium dioxide undergces a phase change from monoclinic to tetragonal at apprcximately 1150 C. A polycrystaline samole of Zr02 brick expands approximately 0.8% from room ~0 temperature to 1150 C. In going through the transition the brick shrinks 0.9% in the 1000C interval above 11500C so that the volume at 1250 C is 0.1% less than its volume at room temperature. This phase change has caused-mechanical problems in zirconia bricks that have been used above 1250 C. Consequently, 0 although magnesia has a greater, but montonically varying expansion, structures constructed from magnesia brick should suffer less mechanical damage than structures constructed from Zr0g. (b) differential rr.ation with respect to the hearth,. concrete walls, and anchors STAFF RE.CPONSE There is considerable information available from past experience with furnance design for steel-making operations that should enable the applicant to design a satisfactory structure from the standpoint of differential motion. As part of the final design effort of the FNP core ladie, we will request design details and drawings of the croposed auchoring system. We find this criteria acceptable, since their refemnce is a standard acceotable 1366'077 to the staff.

16 (c) loss of. concrete behind the wall by spalling, disintegration and melting at calculated temperatures, or at temperatures indicated in' Figure IV-6 of OPS Topical Report No. 36A59 STAFF RESPONSE We cgree with the OPS resconse subject to the qualifications discussed in cur resconse to Questions a.1, a.2(a) and a.2(b). As long as the brick wall remains substantially intact the concrete will heat uo slowly so that spallation is not expected to be a problem. (d) slagging reaction between the brick walls and melted concrete STAFF RESPONSE We agree with the CPS response subject to the cualifications discussed in our r sponse to Questions a.1, a.2(a) and a.2(b). Slagging reactions shout c not be a problem as long as the wall remains intact and the concrete remains below its melting point. 4

17 5. Discuss the fluxing of magnesite brick by silicious material falling into the hearth STAFF RESPONSE We agree with the OPS response, however, please refer to our response to Questions a.1, a.2(a) and a.2(b) with respect to preventing the concrete side walls from melting prior to two (2) days. We would also like to point out that the response mentions an experiment conducted at Sandia Laboratories in which basalt concrete was melted in an Mg0 crucible. It should be noted that the Mg0 brick in the experiment was also heated along with the concrete to 1400 C and this presumably facilitated the formation of the glass-like matrix observed. In the reactor system under discussion, the Mg0 brick would initially be relatively cold. Therefore, it is questionable whether the experir.ent is prototypic. 6. Discuss the properties and merits of basalt as a concrete aggregate STAFF RESPONSE We agree with the OPS response. Of the relatively common materials that can be used as concrete aggregate, basalt is a good choice. It does not generate gas when it is heated. However, it does contain silica which tends to fann low melting mixtures with other materials. 1366.379

18 7. Discuss the ' possibility of the heat flux being higher on the sides of the molten mass than on the bottom (FRG conclusion for concrete m melting going horizentally faster than vertically STAFF 7ESPONSE We agree with the response and it is also our understanding that the high lateral erosion rates observed during penetration of concrete is due to gas genera tion. The concrete results are obviously not aop11 cable to erosion of Mg0. ~- A number of simulant experiments have been carried out to address the above concern and a number of heat transfer correlations have been proposed. In particular, the experiments carried out under L. Baker at ANL and I. Catton at UCLA have indicated that the correlations used oreviously(3) are inappro-priate to molten core penetrating Mg0 because the pool is less dense than the molten Mg0. The simulant esperiments indicate that there is considerable bouyancy-driven motion under these circumstances. The earlier corre1*ationsI3) developed for cool heat transfer were based on a Rayleigh number fomulation, which uses the difference between the bulk pool temperature and the melting interface temperature as the driving force. The experiments at UCLA and ANL indicate that the Rayleigh number fomulation should be based on the density difference between the pool and the melting material. The density-driven correlations result in much higher heat transfer correlations. 1366 y0

19 Apolying these correlations to prototypie conditions was done through the GROWS code. The original version of the GROWS code used the temperature-driven formulations and the heat transfer correlations tended to favor lat ~ penetration in preference to downward penetration. A later version of GROWS (GROWS-2) was issued at a recent meetingD) held at ANL. GROWS-2 uses the density-driven fonnulations, and while there are large differences in density between the pool and the molten Mg0, the code predicts downward penetration rates much faster than the lateral penetrations rates. However, as the pool is diluted with molten Mg0, the density difference between the pool and the molten Mg0 is obviously reduced and the density-driven heat transfer correlations are reduced to the original temperature-driven fannulations. The net result is that for core penetration of Mg0 over a period of several days, the lateral and downward penetration are approximately the same. For the above reasons we have made the lateral heat transfer coefficient equal to the downward heat transfer coefficient in MELSAC. We only distin-guish between the effects of density-driven heat transfer and the original convective formulations using a temperature difference driven Rayleigh number. There is clearly not any experimental evidence yet available that would suggest changing the heat transfer formulations in MELSAC. If such evidence bec::mes available, it could easily be incorporated into MELSAC. }3bb;

20 B. Items Related To Three Mile Island Accide

• 1.

Discuss the possibility of the Upper Head Injection (VHI) System releasing nitrogen into the primary system and impeding the ability to establish or maintain natural circulation. STAFF RESPONSE Upper head injection systems have been reviewed on a generic basisO), and on specific plants (e.g., Sequoyah). The methods of analysis for UHI systems are generally considered to be conservative modifications of the methods used in analyzing other PWRs, and, subject to the reservations stated in reference 1, are considered acceptable. Questions similar to the ACRS have been raised previously by the staff in particular UHI reviews (e.g., Sequoyah) about the mechanical action of the valve system that shuts off the flow of water from the UHI accuiralator tank after injection. These valves are opened when the primary system is brought up to pressure and refaain open during reactor operation. Their only action is to close after injection to prevent nitrogen from following the injected water ir.to the upper head. Two lines in parallel from the UHI accumulator provide the reliability needed in the injection path, and two valves in series in each of these lines provide the reliability needed in shutting off the water flow. The system has been reviewed in several PWR plants, and it has been determined that, since the valves have separate and independent pcwer suppifes and water-level sensing devices, the valve system meets the single-failure criteria. (1) 5. L. Israel, et al., " Safety Evaluation Report on Westinghouse Electric Cmoany ECCS Evaluation Model for Plants Equipped with Upper Head Infection," NUREG-0297, April,1978. 1366 %2

21 It has further been required that the actual installation in each plant be tested prior to plant operation. The tests are to determine the actual amounts of dissolved plus entrained nitrogen carried over to the upper head with the UHI water by a direct sampling technique. These amounts have generally been of the order of <2% compared to an allowable specification of about 4.38% by volume. The tests are conducted into a system at atmospheric pressure, which gives maximum flow velocities and turbulence, with a maximum possibility for entrainment. The tests are considered conservative in this sense. The tests indicate that the amount of nitrogen injected into the uocer head would be conservatively estimated to be of the order of 20-40 cubic feet at atmospheric pressure. These small amounts would not be significant in the ECCS operation, nor would they interfere with natural convection if that were required. The applicant's response to the ACRS question has developed essentially these same arguments. The staff has concluded, therefore, that because of the unlikelihood of more than a single valve failure, and in view of the small quantities of nitrogen injected in preoperational tests, the problem of accidental nitrogen injection has been satisfactory resolved. The effects of the injection of non-condensable gases are the sub;et of an experimental program that is currently underway. The results of these egoeriments wil.l..be factored into the staff's evaluation as they become available. 1366 183

22 2. Discuss the acceptability of the single failure criterion STAFF RESPONSE The OPS response acceptably describes the single failure criterion as currently applied by the staff. This criterion hcwever, does exclude some passive failures and some operator errors which have been identified by the TMI-2 Lessons Learned Task Force. Recommendations by the Task Force may modify this criterion.

• e w

+b 4 1366 184

23 3. Of scuss the timed sequences of events upon the loss of all AC power before core damage will result STAFF RESPONSE The OPS response to this questics reviews a scenario in which failures are limited to the loss of on-site and off-site AC power sources, without recovery. (DC power supplies were. assumed to be available) The turbine driven auxiliary feedwater pump is assumed to function properly, even though operator action could eventually be required to control this system. In the OPS analysis, the principal loss of primary coolant inventory was through pump seals, estimated at 5 gpm per pump. Under the set of assump-tions used in the OPS discussion, the consequences are probably satisfactorily discussed. OPS has concluded that the core will remain covered for about 17 hours in this sequencc. Although some core boiling will occur, the core will be relatively undamaged during this time. Heat removal through the auxiliary feedwater system is expected to remain iable for about the same length of time (20 hours) based on the use of the turbine driven auxiliary feedwater pump. Instrument systems in the containment building are qualified for a temperature environment that would not be exceeded during this period. ~ The staff notes that its own studies, using more restrictive assumotions, have predicted core damage in a shorter period. For example, with loss of both AC and CC pcwer supplies including loss of the turbine driven feedwater supply, core uncovery would take place in one to three hours. For this scenario, rapid loss of inventory from the primary sy:?e would take olace 1366 0'85

24 through the pressurizer valve, since no heat sink would be a :411able to keep the pressure down. Battelle Columcus has studied accidents :equences for the Sequoyah plant, which has an ice-condenser containment similar to OPS. Battelle has used the MARCH code to estimate independently that the core will become uncovered in three hours. The staff estimates that the assumption of failure of the turbine driven feedwater supply introduces a factor of 10*I to 10~2 into the overall probability of the sequence, depending on the accessibility of the turbine controls, which varies widely from plant-to-plant. We also note that similar sequences have been calculated in WASH-1400 (App. V, page 39) for the Surry Plant. Here failure of the turbine driven auxiliary pump was assumed. In the unlikely event that the sequence was completed, core melt could begin in two to three hours. The staff believes that, considering the less restrictive conditions proposed to the applicant, his estimates are probably consistent with the others. 1366 'J86 4

25 4. Discuss the reliability of the auxiliary feedwater system STAFF RESPONSE The estimates on unreliability by OPS for the three scenarios investigated are believed to be an appropriate characterization for the proposed AFWS design. Relative to those generic reliability perspectives derived recently for AFWS designs in 33 operating pressurized water reactor plants, the proposed FNP-AFWS would be as characterized being of high reliability. With exceptions of the scenario involving total loss of AC (where only the steam turbine driven train would be available for autcmatic actuation), the expected dominant contributions to AFWS unreliability would be undetected human errors (pre-existing) that result in incorrectly positioned manual valves in the system. Such errors could arise from either incorrectly positioned va? ves in the suction portion of the system or from failure to correctly position valves following surveillance testing for pump opera-bility. The staff believes that such human interaction potentials can be minimized through appropriate procedural and administrative controls being put into place prior to FNP operation. To this end, the staff suggests that those generic recommendations derived from the recent 33 plants AFWS evaluations be considered in establishing the appropriate procedural / administrative centrols for operation of the FNP-AFWS. 1366 187 4.

~ 26 5. Discuss how H buildup in the ice condenser containment is dealt with 2 following a TMI event and following a core melt STAFF RESPONSE The accident at Three Mile Island, Unit 2 (TMI-2) resulted in a substantial release of hydrogen gas due to an extensive zirconium-water reaction in the re-actor core. A deflagration inside containment followed which creduced a pressure buildup in the containment on the order of 28 psig. Since the TMI-2 containment structure was built to an internal design pressure of 60 psig, there was no loss of containment integrity. However, the immediate question arose as to the con-sequences of a TMI-2 type of event were it to occur in an ice condenser contain-ment, which has both a smaller volume and a lower internal design pressure. In te r-nal containment design pressures of an ice condenser vary between 12 and 15 p We will first discuss the staff': 'o:itten whi:n 1 clud:: tho direction of our current efforts and the proposed basis for continued operation and licensing of nuclear power plants utilizing ice condenser containments. We will follow this with our critique of the OPS response to the ACRS concern The OPS response, which considered 100% metal-water reaction, did not, however address the consequences of a core melt, It is the OPS view that the prior work on core melt as reported in Topical Report No. 36A59,"FNP Core Ladle Design and Safety Evaluation *. satisfies the present information requirements.

27 The OPS plant has a core ladle which is designed to hold the postulated mol-ten core for two days. The applicant only addressed 100 percent metal water reaction and not a core melt accident since it Was concluded in NUREG-0440, " Liquid Pathway Generic Study," that a core melt would violate con-tainment' integrity. Staff Position The accident at TMI-2 was one that exceeded the design basis for a nuclear power pl an t. The failure of the PORV was accompanied by coerator error (turning off the safety injection pumps), procedural error (misalignment of the auxiliary feedwater control valves), design error (pressurizer water level indicator) and a host of deficiencies in the accident analyses. One of these deficiencies was the assumed five percent metal-water reaction in the reactor core. Nuclear powered plants are not designed to withstand multiple-failure events similar to that which occurred at TMI-2. Uthough the TMI-2 accident exceeds the design basis, the staff has considered the consequences of a TMI-2 event, including 100% metal-water reaction, in an ice condenser containment. We are unable at this time to give a quantitative re-sponse to the question of hydrogen control. There exists a sumar.antial number of uncertainties in such an analysis. Many of these are brought up in critique of the OPS response. Qualitatively speaking, the assumption of 100% metal-water reaction seriously challenges containment integrity under almost any scenario. The introduction of high temperature, non-condensible hydrogen gas resulting from a complete metal-water reaction following a LOCA could possibly overpressurize and cause contain-ment failure. The assumotions of deflagration and/or core melt with the intro-duction of additional non-condensible gases would almost certainly cause containment failure in an ice condenser. 1366 M9

28 In Chapter 3, particularly Section 3.3, of NUREG-0585 "TMI-2 Lessons Learned Task Force Final Report" there is a discussion of the need for and feasibility of hydrogen control features in all LWR's that would go beyond the current design bases specified in the NRC regulations. In recommendation 10 of that report, the Task Force recommended to the Director of Nuclear Reactor Regulation that the Commission give notice of intent to conduct rule making relating to the consideration of design features to mitigate degraded core and core melt accidents; in particular, systems for preventing the uncontrolled combustion of hydrogen that could be produced in such accidents. The Director of NRR has asked the ACRS to review and comment on the recommendations in NUREG-0585, and the Office is currently reviewing the recommendations in context with those of the President's Commission on the accident at Three Mile Island and others. The results of that review will be presented to the Commission for decision. A decision on whether and how to proceed with the proposed rule making is not expected to be made by the Commission for several months. The staff proposes to defer imposition of requirements on hydrogen control beyond present requirements pending a decision by the Commission on the proposed rule making. Of primary importance will be the new metal-water reaction rate to be assumed. In the course of rule making, we would expect to consider the feasibility of various alternatives, such as inerting, filtered venting, and controlled burning of hydrogen. I366 190

29 We believe deferral of further action at this time relative to hydrogen control is justified, particularly foi a manufacturing license. The design of the FNP will be required to accomodate the accident prevention measures currently being introduced by the Lessons tearned and Bulletins and Orders Task Forces, those recommended by the President's Commission, and others. These measures include changes in safety equipment design, operating training, accident response, and diagnostic instrumentation to reduce the probability of future accidents which, like TMI-2, might exceed the current design basis and produce large amounts of hydrogen. On this basis, we proposed that for an interim period, until rule making can be conducted, the licensing process for FNP can continue to be conducted in cccordance with the current regula-tions and guides for the design and installation of post-accident containment combustible gas control system, i.e., paragraph 50.44 to 10 CFR Part 50 and Regulatory Guide 1.7. During this interim period the staff does not foresee that viable alternatives will be foreclosed. 1 !3bb 10I

30. Criticue of Offshore Power Systems Resocnse Offshore Power Systems has concluded that core damage similar to that at TMI-2 with burning of hydrogen or 100 percent metal-water reaction in the core without burning of hydrogen would result in a containment pressure of 40 psig. They also concluded that the ice condenser containment would remain intact at 40 psig. The pressure resulting from 100 percent metal-water reaction and burning of hydro-gen would cause containment failure. The applicant concluded that the containment would not experience gross failure at 40 psig. However, we believe that leakage around the penetrations may reach unacceptable levels at pressures as low as 22 psig due to failure in a gradual, ductile manner. Moreover, the increased leakage of containment penetrations due to pressurization above the design limits will have to be evaluated in the analysis of hydrogen buildup inside containment. The applicant's conclusion that the containment would remain intact with 100 per-cent metal-water reaction and no burning of hydrogen implies that preventing the burning of hydrogen by inerting the containment would enable the containment to survive the 100% metal-water reaction. This conclusion is based on the applicant's calculation that the contain' ment pressure will reach a maximum of 40 psig. The containment pressure attained for 100 percent cetal-water reaction without hydro-gen burning depends on the following infomation: 1366 092'

31 Amount of water available in the vessel for metal-water reaction and a. steam generation; b. The portion of chemical energy transferred to hydrogen and to steam; The chemical energy release as a function of time; c. d. The amount of chemical energy abso-bed by heat si,nks (vessel, piping, etc.) before it enters the containment; The amount of ice available for pressure suppression; and e. f. The number of trains of the containment spray system available for pressure suppression. In calculating the containment pressure of 40 psig for 100 percent metal-water reaction without hydrogen burning, the applicant made the fo11cwing assumptions in order to account for the above inform 1ation: a. The applicant assumed 1.4 x 107 8tu's was absorbed by the hydrogen. In addition, the applicant assumed that the total chemical reaction energy was absorbed by steam and transported into the containment. (This appears to involve a double accounting for ten percent of the chemical energy.) With this assumption, the applicant accounted for the chemical energy absorbed by the hydrogen and steam and the amount of water available in the vessel; b. The chemical energy was released at a constant rate over a period of one hour; All of the chemical energy was released to the containment; c. d. Approximately twentf-five percent of the ice was available for pressure suppression; e. Three out of four containment snray tr: ins w<!re available for pressure suporession. 1366 093 --e-- mme ---mm-- e e--m,-,-.m e

32 The applicant's assumption on the amount of energy absorbed by the h>Jrogen and the steam was based on the hydrogen leaving the reactor vessel at 1800*F. Hy-7 drogen at 1800*F accounts for 1.4 x 10 Stu's of the chemical energy. This as-sumption Jay be non-conservative for that accident wherein only that amount of water needed for the metal-water reaction is available in the core. If there was not sufficient water available for steam generation in the reactor, the hydrogen produced would absorb a larger portion of the energy produced by the metal-water reaction. The temperature of the hydrogen entering the containment would then be higher than 1800*F which results in a containment pressure higher than 40 psig. The applicant's assumption that the hydrogen is generated over a period of one hour is dependent on the type of accident that causes the core damage. If the hydrogen were generated over a shorter period of time, a containment pressure greater than 40 psig would be expected. The applicant assumed none of the chemical energy was absorbed by heat sinks on the path out of the reactor vessel. This is the most conservative assumption they could make in this area. The applicant's assumption that twenty-five percent of the ice was available for pressure suppression may be non-conservative for some accident scenarios. Less ice may actually be available resulting in higher containment pressures.

33 After a core mel+down, more than one train of the containment spray system may be lost due to debris from the accident being pumped through the system and dam-aging the cumps. If less than three out of four containment spray system trains were available, the resulting containment presstre would be above the predicted value of 40 psig. The impact of the assumptions on the containment pressure depends on the actual accident scenario. Numerous accident scenarios will have to be studied to deter-mine which one results in the maximum containment pressure for the case of 100 percent metal-water reacticn without hydrogen burning. We do not presently have sufficient information to verify the applicant's conclu-sions on the effects of a postulated accident involving 100 percent metal-water reaction without hydrogen burning. Moreover, we do not have sufficient staff re-sources in the short term to analyze the pressurization of ice condenser contain-ments due to 100 percent metal-water reaction without hydrogen burning for various accident scenarios. Based on a cursory review of the applicant's analyses, we conclude that there are too many uncertainties in the applicant's assumptions to place much credence on the associated conclusions. 1366.395

34 6. Discuss how the FNP compensates for the difficulty, due to the remote location and the lack of space available in improvising new sy:;tems and techniques in case of an accident. STAFF RESPONSE We agree with the OPS response. It should be noted, however, that over the past several months following the Three Mile accident, the staff has been conducting an intensive review of the design and operational aspects of power plants and the emergency procedures for coping with potential accidents. The purpose of these efforts was to identify measures that should be taken in the short-tern to reduce the likelihood of such acci-dents and to improve the emergency preparedness in responding to such events. To carry out this review, efforts were established in four areas: (a) licensee emergency preparedness, (b) operator licensing, (c) bulletins a6d orders followup (primarily in the areas of auxiliary feecsater systems reliability; loss of feedwater and small break loss-of-coolant accident analysis; emergency operating guidelines and procedures) and (d) Short-Term Lessons Learned. The results of these efforts are a set of require-ments that the staff has recomended for implementation. The Ccmmission may add to or modify these staff position: after reviewing them. Additional staff requirements may be developed as the Lessons Learned Task Force com-pletes its long-term recomendations. Efforts are underway within the NRC to review all aspects of emergency planning, including the adequacy of present planning and the need for coordin-ation with and participation of other agencies in developing emergency planning. Appendix C outlines the requirements developed to date resulting }30b

35 from the staff's Emergency Preparedness Studies. Further, the Commission has initiated a rule making procedure, now scheduled for completion in January 1980 in the area of Emergency Planning and Preparedness. Additional requirements are to be expected when rule making is completed and some modifications to the emergency preparedness requirements contained in the Appendi' may be necessary. Moreover, an NRC-EPA Task Force i<eport, NUREG-0396 dated December 1978 recommended 10- and 50-mile emergency planning zones and the Commission has endorsed this recommendation. The results of our ongoing studies and rule making hearings will be applied to the FNP design as well as to the utility-owner for site dependent matters. 7. Discuss how one faces lack of flexibility for design changes due to thecompactness and lack of available space on the FNP STAFF RESPONSE We agree witn the OPS response. The cons deration of compactness and lack of space has been raised as early as 1971 during our preapplication review. We have recognized this aspect and have consf 'ered it in our subsecuent review and evaluation and we have not found lack of space to be a design constraint at this time. D s i366 097

36 C. Items Concerning The Effects Of Changing Base Mat Material Discuss the effects of changing the base mat from concrete to magnesium oxide on the probability of a major air release during a core mel accident. Discuss the comparisons of probabilities and dose levels for air releases associated with concrete and magnesium oxide curing a core melt accident. STAFF RESPONSE Large amounts of airborne radia-isotopes can be dispersed outside con-tainment by the release of either gas or liquid to the environment. Gaseous radioisotopes or aerosol particles can be released in a gas ~ onase, or, as at TMI-2, radic-isotopes dissolved in liquid phase can be released to subsequently emit gaseous radio-isotopes. To the extent tha: the presence of the core ladle delays or inhibits one or more steps in such releasts, it may be considered to reduce the probability of release by the affected mechanisms, as noted in the applicant's response. Of greater 'nterest is the possible existence of re! ease mechanisms unique to ne ladle design. The presence of a core ladle alters the possible release mechanisms during a core melt accident by the following means: 1. The radio-isotopic inventory is maintained for a longer time and at a higner temoerature within the reactor compartment. 2. The to:al amount of material melted is increased, and its chemical composition altered. 3. The free volume of the reactor compartment is reduced by the excess volume of the ladle over that of the concrete. To scme ex ent, :nese differences be:,veen core ladle and concrete mat design coun:eract one another. (2 vs 3) The lower free volume means that for a given accident-induced

1. 10w :nrcugn or frem :ne reactor compartment a larger fraction of volatile 1366 19R

.ffO

37 fission products would be swept out of the reactor compartment as a step towards their oossible ultimate release. Counteracting this is the recuced generation of gases due to the altered chemical composition of the melt, sucn that the net effect of the ladle is to reduce mass flow from tne ' reactor compartment. Exceotions to this generalization are those accidents in which water at late times in the accident is introduced into the ccmcartment, and for these cases :he adverse volume effect is less favorable by only a very small fracticn. (1 vs 2) The higner :emperatures of melts in the ladle design increase the molar entropy of fission product vapors in thermal equilibrium with

he melt.

Counteracting this is the increased molar entropy of fission products wi:nin the melt due to dilution by Mg0 and the much larger amount of molten steel in the ladle as opposed to the smaller concrete-basait melt. For those species having pure phase boiling points above about 2500* K the net effect would be expected to be a reduction in equilibrium vapor pressures due to dilution in the ladle, while for the more volatile species, such as cesium, the higher temperatures in the ladie would lead to higher equilibrium vapor pressures. Since no non-condensible gas generation is expected in the ladle design, there is little~ driving force available to remove fission product vapors from the reactor compartment for the formation of aerosols, hence little, if any, adverse affec: upon airborne source terms. The core ladle design allows a much higher energy density to occur, since it results in the storage of later. and internal energies (heat of fussion anc heat capacity integrated over temperature) which, in the i366 09(T 4

38 concrete mat design, are discharge by pyrolysis and volatilization of the concrete and comparatively rapid ultimate discharge to the water beneath the hull should water come into contact with the melt to proeduce a steam explosion, the thermodynamic efficiency (conversion of heat energy to mechanical work) could be higher with the higher energy density in the ladle. The probability of a steam explosion, however as well as the con-version efficiency is a complex function of a large number of functions which were previously documented in NUREG'0440, " Liquid Pathway Generic Study," and also discussed with the ACRS. It is the staff's conclusion that neither the probability nor the themal to mechanical work compression efficiency will be changed in any appreciable way by the presence of the core ladle. Tempera:ures in the reactor compartment when it is largely lined with magnesia would be o# the order of hundreds of degrees hatter than if unlined. Significant vapor pres'sures of silver control rod material and. steel ccmconents could be maintained. The generation and condensation elsewnere of these vapors could ccnstitute a significant heat transport mechanism frcm the melt. There would be larger thermal and concentration gradients within the volumes and openings connecting the reactor com-partment Oc the remainder of the containment. Diffusive and Soret effect transoort in these connecting volumes could lead to aerosol formation by condensation and reaction of metal gases with the containment a =osphere. In addition, the molar volumes of gases within the reactor compar=ent would be several times larger than those in the upper containment, wnile the mean molecular weights would be only two to three times larger, making possicie connective transpcrt. There effects are also present in

ne concrete ma: design, but to a lesser degree.

Eacn of :ne effects outlined above have differences between ladle and

ancrete ma: designs which, to some unknown extent, counteract one anctner in Neir ecles in cetemining the expected air release source

}3()() 100

39 The effects differ in the two designs only by degree, such that term. they offer 90 airborne release mechanism unique to the ladle design. On iciance, the choice between a concrete base mat or an Mg0 core ladle dces no sucstantially affect the airborne release probability or tne inventory of radio-isotcces susceatible to release, except insofar as the ladle design inhibits or delays eventual creach of the F,'iP hull. Since this exceo:f on is significant, the applicant's unquantified assesscent of overall risk reduc: ion aopears warranted. 1366 101 ~ ~ _n-

40 2. Discuss the consideration given to the use of a vented containment. Discuss the consideration given to the use of sea water for venting and/or cooling a molten core. STAFF RESPONSE A vented containment design accepts small, controlled leakage in exchange for reducticn of the likelihcod of massive uncontrolled release from containment failure. Venting in principle can protect containment from failure cue to steam or nydrogen flame burning overpressuri:ation, but not from failure due to detonation. Atonations are, however, inherently less likely than less violant, though rapid, pressurizations, since the fematien of strong : hecks is possible only under restricted circumstances. Kycr:cen gas is much more easily disscciated, either thermally or by ionizing radiction, than oxygen, nitrogen, or water vapor. It is therefore, susceptible

scentaneous ignition when in locally high concentration (diffusion flame : nditens) and corresocndingly less likely to form detonable mixtures t9ecughcut the containment.

Slow hydrogen generation, on the other hand, is ntroilable by the hydrogen recombiners. Steam explosions due to the rapid nixing of watcr and molten material are less likely to contribute containment failure than more slower overpressuri-zation events (See NUREG-0440, " Liquid Pathway Generic Study," pages A-16 to A-24). The accident sequences leading to containment over presturization are, therefore, dominated by comoaratively slow pressure increases, which a n! susceptible to mitigation by containment vent systems. 1 % 6 102 4

41 Seawater is generally s!igntly alkaline (pH8), and thus could serve as a reducing agent for iodine. In addition, it contains a mean concentration of 50 mg per tonne of natural iodide and could function well as an isotopic exchange reservoir for vented radio-fodines. If gas were vented into seawater at depth, i.e., under pressure, the solubility of xenon in seawater would also be significant. Venting to seawater could, therefore, reduce the octantial for airborne release, although at a cost of unnecessary seawater contamination during accidents which did not challenge containment integrity. The enlaride ion concentration of seawater renders it unsuitable for use within contair. ment as a coolant due to corrosive effects on steel and Sufficient feedwater is present on the barge to cool the concrete. molten core if provisions were made for this purpose. The simples: and most direct method of venting to the sea would be a well or standpipe ccmmunicating the contair. ment to the underside of the huil. A skirt or inverted wall surrounding the bottom of the barge could :nen capture vented gases and guarantee that containment pressure woulc c.ever rise above the hydrostatic pressure on the hull bottcm.

42 3. Discuss the change in position for allowing the FNP to be placed on riverine and estuarine sites. Has the proposed installation of the core ladle changed the NRC Staff's position on this matter, if so, why? What actions and in what time period, are considered practical to isolate the core for a riverine or estuarine sice? STAFF POSITION 9 The NRC staff position related to the generali:ed siting of RIPS in estuarine and riverine areas has remained unchanged througnout the course of :he s:aff's environmental review.1,2,3 This position is that "... finding acceptacle FNP sites in estuaries, rivers, or near barrier islands, will : cst likely be extremely difficult, but [the staff] cannot conclude that :nere are no accept-able estuarine, riverine or barrier island locations for RIP emolacement *., hen appropriate mitigative actions are taken." Both the staff and,the U.S. EPA ~ concluded that siting FNPs in such areas could produce a significant potential for adverse environmental impact, particularly with actions associa:ed with construction and maintenance dredging. Furthermore, in its assessment of the FNP core-melt accident at an estuarine or riverine site, the staff concluded that a direct release of radioactive material to such areas would resul; in unacceptacle 4 consequences to the environment. As such, the s:sff, in consulta icn si h tne U.S. EPA, has concluded that applicants wno wish to site RIPS at specific loca-tions (including sites in estuaries and rivers) must comply with certain environ-mental siting requirements including specific mitigative actions to limit the environmental consequences of a core-melt accident at an estuarine / riverine sited R4P. I FES, Part II 2FES, Part II Addendum 3FES, Part III (NUR~G-0502)

1. FES, Part III (NUREG-0502) p. xiv

, BOD l04 I

43 The proposed installation of the core ladle in the FNP did not change the NRC staff's position regarding the acceptability of FNP siting in estuarines, rivers or near barrier islands. Environmental siting requirement 1.3 reproduced belcw from the FES, Par: ::I must be complied with by an applicant wno wishes to locate an F:iP at a specific site in an estuary, river or near a barrier island and since i: rela:es to specific site conditions it was not imposed as a c:nci:icn of :ne manufacturing license application. Environmental Sitino Recuirement 1.3 " Proposed FNP sites in estuaries, river or near barrier islands must be appr:- priately modified in an environmentally acceptable manner sucn : hat in -he event of a core-melt accident, the release of radicau.ive material into the surrounding water body shall be limited to levels that will not resul in undue impact to man or the ecosystem.d With respect to actions and time periods considered practical to isniate the core for river and estuary sites, the staff concluded that total isolation of racio-active core-debris frcm open estuarine / riverine waters, following a cere-mel: accident would be very difficult to achieve. Furthemore, :he s:aff concluded that total isolation wculd not be necessary,providec the ccmbination Of site characteristics, FNP design features and interdiction methods could provice adequate assurance that a core-melt type accident would not pr cuce risks any worse than a typical land-based plant at a river or estuary site. Thus the staff required (siting requirement 1.3) that an R{P site in such areas usc be modified to restrict the potentially widespread and chronic release of radio-activity in the event of a core-melt accicent. Siting requirement 1.3 is stipulated independently of manufacturing license concition flo. A wnica requires 1366 105'

44 that the FNP be redesigned to incorporated a core ladle. The core-ladle design would provide additional delay before potential melt-through beneath the reactor vessel in order to provide additional time to incorporate interdictive measures, but in the event of an actual melt-through, radio-active debris would undoubtedly be released to the ambient estuarinet riverine environment. This would, in the staff's view, produce unacceptable environmental impacts. Environmental siting requirements 1.3 is intended to prevent waterborne contaminants resulting from core-melt type accidents from spreading offsite in an uncontrolled manner. The bases for the requirement included considera-tion of mitigation and interdiction techniques that could be employed at both land-based and FNP sites to limit the offsite migration of activity into the estuary or river and reduce the long-tenn environmental consequences of such releases. The environmental consequences in most estuary and river siting situation were judged likely to produce both acute and chronic effects on biota due to the generally very cilw naturai pollutant flushing capability of such water bodies. Classes of aquatic biota might be destroyed, therefore impacting the ecosystem for years. A direct result of such chronic conditions upon biota would be an indirect effect upon man due to relatively long-term public restriction of water resource related activities on a large scale. So as to implement this environmental siting recuirement, the aaplicant has pro-posed an additional plant-site interface criterion in their FNP Ccre Lacle Topical 6 Report. The criterion requires that site modifications be made at procosed IFES, Part III, p. xy 0 FNP Core Ladle Design and Safety Evaluation, Offshore Power Systems, Topical Report No. 36A59. April 1979, p. VI-2, 1366 106

45 specific FNP sites in estuaries and rivers to ensure that the envircnmental con-sequences of an FNP core-melt accident in :nese areas wouic be no <.cese tnan those for estuary sited typical land based plants consicered in one LPGS Report. The staff has accepted this criterion, noting that the consequences of core-melt type accidents should be assessed for any proposed estuary or river :NP site. The assessment will consider specific FSP site and plant cesign informa- -tion for ccmparison with typical land based reactor sites in estuaries and rivers. Thus, at this time no specific sections in a given tire period have been specified. e e 9 i 366 107

46 4. Discuss the NRC Staff's position that the RIP Core Ladle is considered an environmental issue and not a safety issue. STAFF POSITION The following background is needed in order to place the response to this ACRS request in perspective. Under its present mandate the NRC assesses the imp 11-5 cations of licensing nuclear powar plants under two Acts: the Atomic Energy Act of 1954 as amended (.f.e., protection of the public health and safety) and the National Environmental Policy Act of 1969 (NEPA) (i.e., protection of the environment and overall cost-benefit balancing). Pursuant to the Atomic Energy Act,10 CFR Part 50 of the Commission's regulations was issued in the mid 1950's and formed the basis for the staff's analysis of the safety of proposed nuclear power plants. Subsequently, in August 1974, the Ccmmission issued an interim statement of policy in the Federal Register concerning the treat-ment of postulated accidents in the staff's safety reviews. The Commission stated: "In the approach to safety reflected

• in the Conmission's regulations, postulated accidents, for purposes of analysis, are divided into two categories - " credible" and " incredible." The former (" credible")

are considered to be within the category of design basis accidents. Protective measures are required and provided for all those postulated accidents falling within that category, and proposed sites are evaluated by taking into acenunt the conservatively calculated con-sequences of a spectrum of severe postulated accidents. Those accidents falling within the " incredible" category are considered to be so improbable that no such protective measures are required." Using this statement as a basis, the staff judged that a core-melt accident fell into the " incredible" category and therefore would not be considered in its safety ' evaluations prepared for the licensing of nuclear plants, including the FNP. This reasoning is based upon the staff's perception that the RIP nuclear system design is similar to that of land-based plants, and thus the probability of occurrence of acore-melt accident was viewed as equivalent (i.e., incredible) 1366 108 ,._m. ~ = = - ~ - ~ ~ ' '

47 for both types of siting options. The scope of the staff's safety review relative to the core ladle centers only upon the question of whether incorporation of the core ladle in the FNP would alter previous staff conclusions regarding the overall safety of the FNP design. The Comission's implementing regulations for NEPA are contained in 10 CFR Part 51 but there is no specifically approved Ccmission regulation for the consideration of accidents under NEPA. The NRC has historically censidered the potential environmental consequences of plant accidents in the manner prescribed in the proposed Annex A to 10 CFR Part 50, Appendix 0. The proposed Annex is a part of an AEC proposed regulation to implement ;; EPA. The Annex was issued for public coment in December 1971, but no final Annex has been prepared. Subsequently, the Comission replaced Appendix 0 to 10 CFR Part 50 with 10 CFR Part 51 wnich specifically addresses the MRC censideration of NEPA issues. Technically, the rulemaking proceeding for the Annex is still p;nding before the NRC, and while the Ccmission has never fornally adopted the Annex, it authori:ed its use as guidance. The proposed Annex divided radiological accidents into nine classes for NEPA evaluation purposes. With respect to the ninth class (Class 9 accidents), the Annex concluded that applicants would not be required to discuss such accidents in their Environmental Reports since the probability of occurrence was so low as to make the risk negligible. With regard to the OPS application for FNP's, the staff found that the FNP design offered a departure frem land-based siting, and the potential environmental 1366 109 s

r..

48 consequences frcm a care-melt type accident could differ in type and magnitude from those ascribed to land-based plants. Therefore the environmental review of the core-melt accident at the FNP need not be guideu by the proposed Annex. ' With this information, one can specifically respond to the ACRS request. Since the Commission Policy Statement 1recluded the review of the FNP-core-melt accident pursuant to the Atomic Energy Act, the staff decided to evaluate such an accident under i.e., within the bounds of the FNP environmental review, pursuant to 10 CFR Part 51 of the Commission's regulations which imple-ments NEPA. Various factors weighed heavily in the staff's decision to consider this accident under NEPA. These included: (a) The novel siting option of the FNP led to FNP core-melt consequences different from LBP core-melt consequences; (b) the staff position that the proposed Annex A to 10 CFR Part 50, Appendix 0, did not apply to FNPs; and (c) NEPA's mandate to disclose to the fullest extent possible the consequences of major federal actions. (a) FNP core-melt consecuences different from LBP core-melt consecuences. Early in its review of the FNP. concept the staff perceived that the core-melt accident at an FNP could result in environmental consequences different from thosa for a similar accident at an L3P. From a core melt accident viewpoint, the FNP did not offer the same degree of natural isolation as a LSP, i.e., a core-me_lt accident at the FNP could probably result in a prcmpt release of radio-active debris into the water which then could be diffused by currents and tides. In the L3P, such an accident would probably result in the retention of core-debrfs in the earth with significantly different liquid pathway impacts. This staff perception together with the ACRS concerns expressed in their letter of O 1366 110

49 November 1972, to the AEC, prompted tra NRC to initiate the LPG 5 in order to ecmpare the design basis and core-me' c liquid pathway risks for FNPs and LSPs. (b) Procosed Annex A to 10 CFR Pa~t 50, Accendix 0 did not acoly to FNPs* N staff reasoned that the FNP rguld produce core-melt environmental consequences different in k1ne ""m LBPs ';.e., liquid pathway consequences) and undoubtedly different from those considered when the proposed Annex was being developed. Further, the FNP concept was not specifically considered by the Ccemission when it issued the proposed Annex in 1971 for public comment. The staff concluded, therefore, that the policies set forth in the proposed Annex were not applicable to RIPS and that an evaluation of the environmental impacts fr:m core-melt acci-dents was a proper topic for staff consideration in the generic environmental impact statement for FNPs. (c) NEPA Mandate of Full Disclosure The staff's position is that NEPA requires a federal agency to fully disclose all pertinent environmental infornation including controversial or opposing views, within an enviremental impact statement such that the decision-makers and public are fully informed. The staff, having cencluded in the LPG 5 Report that the liquid pathway consequences at an FNP differed significantly frem the L3P counter-part, was obligated under the intent of NEPA to fully assess the environmental implications of such a finding in the staff's FES, Part III.

• This position is supported by the Ccmmission's Memorandum and Order of September 14, 1979 concerning the certified question: "Are Class 9 accidents a proper subject for consideration in the staff's enviromental statement on the floating nuclear power plant manufacturing license application?"

1366 111

50 In summary, the proposed core ladle requirement is a direct result of the staff's decision to consider " class 9" or core-melt type accidents as part of the environmental review for the FNP manufacturing license applications The staff views the proposed FNP core ladle design requirement set forth in the FES, Part III (NUREG 0502) as both an environmental and safety issue. The genesis and imposition of the core ladle requirement, however, is based solely on the staff's environmental assessment (FES, Part III) and the LPGS Report (NUREG-0440). The overall safety implications of incorporating such a feature into the FNP is currentiy under evaluation by the staff. 9 I 1366 112 9

51 0. Additional Information Recuests From The NRC Staff 1.~ Provide available information on the 'Sandia 100 plant liquid pathway study STAFF RESPONSE The Liquid Pathway Study at Sandia has not yet produced useful results. If and when the study produces results, we will make them available. 'de anticipate some early results in the next few months; however, the numerical uncertainties are expected to be so large as to make the results useful only for directing areas of further study. 2. Provide available information on the WASH-1400 type study of the ice con-denser type plant, along with a comparison for non-ice condenser type plants. STAFF RESPONSE A draft Sandia report on the ice condenser exists at this time and was provided to D. Okrent of the ACRS in July 1979. A comparison study with othEr non-ice condenser plants in underway but is currently stopped because of higher priority work. That report (comparison study) is not expected to be 1.9 draft form before about June 1980. 1366 113 \\

APPENDIX A ~ "o* UNITED STATES + ~ 5 ?.. NUCLEAR REGULATORY COMMISSION { ADVISORY COMMITTEE ON REACTOR SAFEGUARDS t f WASMNGToN. o. C. 20555 y. July 25, 1979 Harold R. Denton Director, office of Nuclear Regulatory Regulations SUBJEC1': AGS RE7IIW CF IHE ECATDG NUCI. EAR PLANT CIRE IADI.E DESIGN At the June 27,1979 ACRS *W4ttee Meeting on the Floating Nuclear Plant, :mmobers of your staff requested that the ACRS :neet at an early date to discuss the proposed !NP Core Iadle Design and to write a letter to Mr. Gossick canmenting on that preliminary design prior to the NRC Staff's ~ issuance of its safety evaluation. The Acting ACRS 91W4ttee Chair: nan infor:ned your staff and representatives of offshore Power Systens that the stqgestion to bold an early AGS :neeting would be considered at the July 1979 AGS :necting. The proposal to hold an early AGS review of the conceptual design of the ENP core ladie was discussed at the July 1979 AGS :neetirg. It was decided that additional infor: nation, as indicated below, is necery be-fore the Ceaunittee can proceed with its review of the INP. a. Items Related to the Imoact that the Core Eadle Will Have on Other Centainment Structures 1. Calculate the fraction of decay heat radiated frce the pool for'the proposed design. 2. Calculate the effects of heat radiation in Itan 1 on the rate of: (a) disintegration and collapse of exp: sed concrete (b) disintegratien and collapse or melting of cencrete behind the 6 in-b :nagnesite brick wall (c) collapse of steel frca the reactor cavity. 3. Discuss the consequences of Item 2 with respect to: (a) loss of integrity of superstructures (b) loss of hearth capacity ...m.m.

ACRS Review of FNP Core Ladle Design Jt21y 25,1979 (c) impact resistance of the hearth and its supports (d) integrity of structural steel members. 4. Discuss the stability of the 6 inch magnesite brick wall above the hearth level with respect to: (a) loss of brick by e=1mg (b) differential motion with respect to the hearth, concrete walls, and anchors (c) loss of concrete behind the wall by spallirg, disintegration, and melting at calculated taperatures, or at taperatures indicated in Fig IV-6 of CPS Topical Report No. 36A59 (d) slagging reaction between the brick walls and melted concrete. 5. Discuss the fluxing of synesite brick by siliceous material falling into the hearth. 6. Discuss the properties, and merits of basalt as a concrete aggregate. 7. Discuss the possibility of the beat flux beirg higher on the sides of the molten mass than on the bottom (FIC conclusion for concrete melt) with melting goirs hori::entally faster than vertically. b. Items Related to 'Miree Mile Island Accident 1. Discuss the possibility of the Upper Head Injection System re-leasing nitrogen into the primary system and impedirg the ability to establish or maintain natural circulation. 2. Discuss the acceptability of the single failure criterion. 3. Discuss the timed sequence of events upon the loss of all AC power before core damage will result. 4. Discuss the reliability of the auxiliary feedwater system. ? n the ice condenser con *=4-rit is dealt i 5. Discuss how M41 A with fell a M event and following a core melt. 6. Discuss bow the PNP : -{--rates for the difficulty, due to the re-mtd' location and the lack of space available, in improvising new systems and techniques in case of an accident. 7. Discuss how one faces lack of flexibility for design charges due to the w _ess and lack of available space on the PNP. })Ob

ACRSRadriewofCoreLadleDesign 3-July 25, 1979 c. Items Concernire the Effects of Chanaire Base Mat Materials 1. Discuss the effects of changing the base mat fr a concrete to magnesim oxide on the probability of a major air release during a core melt accident. Discuss the comparisons of probabilities and dose levels for air releases associated with concrete and magnesi m oxide during a core melt accident. 2. Discuss the consideration given to the use of a vented contairment. Discuss the consideration given to the use of sea water for venting and/or cooling a molten core. 3. Discuss the change in position for allowing the RIP to be placed on riverine and estuarine sites. Has the propssed installation of the core ladle changed the NRC Staff's position on this matter, if so, why? What actions and in what time period, are censidered practical to isolate the core for a riverine or estuarine site? 4. Discuss the NRC Staff's position that the ENP Core I.adle is con-sidered an environmental issue and not a safety issue. d. Additional Information Recuested Frem the NRC Staff 1. Provide available information on the Sandia 100 plant liquid path-way study. 2. Provide available information en the WASE-1400 type study of the ice condenser type plant, along with a emparison for non-ice condenser type plants. Following receipt of Offshore Power System's response to the items listed above and a written evaluation by the NRC Staff, another ACRS 9 m=4 ttee meeting will be held. Please advise us of the date by which you believe the above information will be available so we can *'"le related ACRS activities. R. F. Fraley Executive Director cc: D. Muller, IlSE E. Case, NRC D. Vassallo, DPM F. Schroeder, IlSS } 3bb, b 54 -T t ,n s

a [ FNP-PAL-050 APPENDIX B .iffshore Power Systems wx)0,wm.n ws.. '-24 :4-7 oc Ei:x ?,000. ;.o.tsonvine. > '.<icta 22211 t- %240G Septemcer 14, 1979 Mr. Robert L. Baer, Chief Light Water Reactors Branch No. 2 Division of Project Management U.S. Nuclear Regulatory Comission 7920 Norfolk Avenue Bethesda, Maryland 20852 3.Haga Re: Docket STN 50-437; ACRS Questions on Core Ladle and TMI-2 ,t , l$ 3 L'* Clifj

Dear Mr. Baer:

Transmitted herewith are 20 copies of the Offshore Power Systems responses to the ACRS Subcomittee questions contained in R. F. Fraley's letter to H. R. Denton dated July 25, 1979. Please note that we have not offered responses to part d. of Mr. Fraley's letter as these requests were made specifically to the NRC Staff. By copy of this letter, 20 copies of our responses are being transmitted directly to Mr. Fraley for distribution within ACRS. Certain material in the attached resconses reflects modifi-cation to the design presented in OPS Report 36A59, "FNP Core Ladle Design and Safety Evaluation". The principal changes are increased ladle volume and increased refractory insulation on the walls of the reactor cavity. Both of these enanges resulted from our ongoing evaluation of radiant upheating from the pool surface. The analyses of radiant upneating, which are described in the attached res::enses, are believed to be adequately conservative to show feasibility and therefore to support the issuance of the Manufacturing License. Following NRC Staff review those responses wh.ch affect the present content of Report 36A59 will be retrans-mitted in the fann of a revision to that report. We ask that these responses be reviewed on an expedited basis leading to an ACRS Subcomittee meeting as early as October 1366 117 __=

Page Two September 14, 1979 1979. To this end, we are prepared to offer any assistance the Staff may require. Ve truly y urs, 'v P. B. Haga /lel Attachments CC: R. F. Fraley (ACRS) V. W. Campbell A. R. Collier 4 e 56 \\\\% ,,6 \\30

APPENDIX C .e D**D

• D wo w

m NEAR TERM REOUIREMENTS FOR IMPROVING EMERGENCY PREPAREDNESS While the emergency plans of all power reactor licensees have been reviewd in the past for confurmance to :he general provisions of Appencix E :o 10 CFR Part 50, the most recent guicance on emergency planning, primarily that given in Regulatory Guide 1.101, " Emergency Planning for Nuclear Power Plants", has not yet been fully implemented by mos: reactor licensees. Further, there are some additional areas where improvements in emergency plan,ning have teen highlighted as particularly significant by the TMI-2 accicent. We plan to undertake an intensive effort over about the nex: year to" improve licensee preparedness at all operating power reac:ces anc tnose reac ors scheduled for an ocerating license decision within :he nex: year. Thi s~ effort-will be closely coordinated with a similar effor: by the Of fice of State ~ Programs to improve State and local response plans througn :he concurrence process and the efforts of the Office of Inspection anc Enforcement to verify proper implementation of licensee emergency preparedness activities. Further, the Commission has initiated a rulemaking procedure, now 'scheculed for completion in January 1980, in the area of Emergency Planning and Preparedness. Additional requirements are to be expected when this rulemaking is completed and some modifications to the emergency preparedness requirements contained in this letter may be necessary. Our near term requirements in this effort are as follows: (1) Upgrade licensee emergency plans to satisfy Regula:ory Guide 1.101, with special attention to the development of uniform action level criteria based on plant parameters. (2) Assure the implementation of the rela:ed reccmencations of the Lessc.s Learned Task Force involving instementation :o follos the course of an acciden: and relate the information proviced by :nis instrunentation to the emergency plan action levels. This will incluce ins ementation for post-accident sampling, high range radioactivi:y monitors, ann improved jn-plant. radioicdine instroentation. The implementation of the Lessons Learned Task Force's recccinendations on instrunentation for cclection of inadequate core cooling will also be factored into the emergency plan action level criteria. (3 ) Determine that an emergency operrtions center for Federal, State and local personnel has been estaclished with suitable ex.unications to the plant, and tnat upgrading of the facility in accordance with the Lessons Learned Task Force's reccmmendation for an in-plan technical support center is underway. (4 ) Assure that improved licensee offsite monitoring capao:lities (inclucing accitional :nermoluninescent dosimeters or the equivalent) have :een r proviced for all sites. l 5 1 57 1366 119 i

.J... 2-(5 ) Assess the relationship of State / local plans to the licensees' anc Feceral plans so as to assure the capability to take accrepriate emergency actions. Assure that this capacility will se extencec to a distance of ten miles. This item will be performed in conjunction with th'e Of fice of State Prograns and the Office of Inspection and Enforcament. (6 ) Require test eleccises of approved emergency plans (Federal, State, local and licensees), review plans for such exercises, anc participate in a limited number of joint exercises. Tests of licensee pl ans will,be required to be conducted as soon as practical fc all facilities and before reacter startup for new licensees. Exer:ises of State plans will be cerformec.in conjunction with the concurrence reviews of the Of fice of State Progrens. As a preliminary planning bases, assume tnat joint test exercises involving Faderal, Scate, local and licensees will be. conducted at the rate of about ten per year, which would result in all sites being exercised once each five years. Revised clanning guidance may result from the ongoing rulemaking. D**]D ]D ' 3 Y @ o Ju 2. \\ W2 o w Ju 58 1366 I20

, = >. APPENDIX 0 cn nn D D c> D 'D0 M J[]}/fb k REFERENCES ,eo a 1. R. D. Gasser and W. T. Pratt, "MELSAC - A Computer Code to Determine the Thermal Response of a Sacrificial Sed and Surrounding Structures to a Core Melt Event," to be published as a BNL Report. 2. W. T. Pratt and R. D. Gasser, " Thermal Analysis of an FNP Sacrificial Bed," to be published as a BNL Report. 3. F. A. Kulacki and R. J. Goldstein, " Thermal Convection in a Horizontal Fluid Layer with Uniform Volumetric Energy Sources," J. Fluid Mech., Vol. 55, Pt. 2, pp. 271-287, (1972). 4 R. Faradieh and L. Baker, Jr., " Heat Transfer Phenomenology of a Hydrodynamically Unstable Melting System," J. Heat Transfer, Vol.100, Pt. 2, pp. 305-310, (1978). - 5. Letter from P. 8. Haga (OPS) to R. L. Baer (NRC), " Docket STN 50-437; ACRS Questions on Core Ladle and TMI-2," September 14, 1979. 6. For a description of the original GROWS code, see Chapter V of ANL/ RAS 74-29, October 1974 - 7. W. T. Pratt, "Tri'> Report on GROWS Code Meeting held at ANL on July 10, 1979," i memorandum to R. A. Bari, Brookhaven National Laboratory, August 9,1979. 59 1366 121 %}}