ML20138K006

Forwards Comments Re Rept Entitled, Lower Head Integrity Under In-Vessel Steam Explosion Loads, to Serve as Agenda Items for Currently Unscheduled Telcon

ML20138K006
Person / Time
Site:	05200003
Issue date:	03/25/1997
From:	Joseph Sebrosky NRC (Affiliation Not Assigned)
To:	Liparulo N WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
References
NUDOCS 9705090234
Download: ML20138K006 (11)

Text

', . Tkttel NP

{ perog ,-

a y t UNITED STATES l- g j

t NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20eeH001

{

\...../ March 25,1997 b

1 Nr. Nicholas J. Liparulo, Nanager Nuclear Safety and Regulatory Analysis i Nuclear and Advanced Technology Division j

Westinghouse Electric Corporation 3

P.O. Box 355 Pittsburgh, PA 15230 i

SUBJECT:

COWiENTS REGARDING REPORT ENTITLED LOWER HEAD INTEGRITY UNDERi IN-VESSEL STEAM EXPLOSION LOADS FOR THE AP600, I

Dear Mr. Liparulo:

l 1

As a result of its review of the June 1992, application for design certifica- '

j tion of the AP600, the staff has determined that it needs additional informa-i tion. Specifically, the staff has reviewed the report entitled " Lower Head '

j:

i Integrity Under In-Vessel Steam Explosion Loads," (DOE /ID-10541). The report.  !

was' transmitted for NRC staff review in support of the AP600 in-vessel  ;

4 retention topic. We propose that the enclosed comments serve as agenda items 3

for a currently unscheduled teleconference on the matter. During this  !

i teleconference the staff will determine which of the enclosed comments need to i be formally addressed by Westinghouse. The staff also requests that the peer

! review comments for the report (DOE /ID-10541) be made available to the staff  !

l for its review. The staff understands that responses to the comments may not have been developed at this time, but because of the aggressive schedule for -

1 the AP600 review the staff feels it would be valuable to have the review l

comments without the responses as soon as possible.

l You have requested that portions of the information submitted in the  !

) June 1992, application for design certification be exempt from mandatory t public disclosure. While the staff has not completed its review of your l 1

4 request in accordance with the requirements of 10 CFR 2.790, that portion of

} the submitted information is being withheld from public disclosure pending the i staff's final determination. The staff concludes that these followon ques-i tions do not contain those portions of the information for which exemption is sought. However, the staff will withhold this letter from public disclosure i

for 30 calendar days from the date of this letter to allow Westinghouse the j opportunity to verify the staff's conclusions. If, after that time, you do )

not request that all or portions of the information in the enclosures be withheld from public disclosure in accordance with 10 CFR 2.790, this letter l will be placed in the Nuclear Regulatory Commission Public Document Room. l

'J l 1 l i

3 9705090234 970325 pf0b PDR ADOCK 0520 3 gg {g g I

l

. l L

Mr. Nicholas J. Liparulo . March 25, 1997 If you have any questions regarding this matter, you may contact me at (301) 415-1132.

Sincerely, original signed by:

Joseph M. Sebrosky, Project Manager Standardization Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation Docket No.52-003

Enclosure:

As stated cc w/ enclosure:

See next page DISTRIBUTION: _ Enclosure to be held for 30 days

• DocketsFile !

PDST R/F TMartin

• PUBLIC MSlosson TQuay TKenyon DJackson BHuffman JSebrosky WDean, 0-17 G21 ACRS (11)  ;

JMoore, 0-15 B18 MSnodderly, 0-8 H7 JKudrick, 0-8 H7 RPalla, 0-8 H7 CAder, T-10 K8 SBasu, T-10 K8 ,

ARubin, T-10 K8 '

i i

DOCUMENT NAME: A: IVSE W.COM  ?(

T2 becalve a copy of this desunient,indeeTo in the ben: *C' s Copy without ettechmentsqoure, "E* == Copy with attachment / enclosure "N* = No copy 0FFICE PM:PDST:DRPM l- SCSB:DSSA E SC:SCSSI D (Sh\ D:PDST:DRPM l NAME JSebrosky:sg M C MSnodderly /RS JKudric h iM \ TQuay 1 "

DATE 03/o '/97 V 03/2//97 03/ty97 \\\V ' 03/f/97 0FFiCIAL RECORD COPT

l i

l l

Mr. Nicholas J. Liparulo Docket No.52-003 Westinghouse Electric Corporation AP600 1

l cc: Mr. B. A McIntyre Ms. Cindy L. Haag Advanced Plant Safety & Licensing l Westinghouse Electric Corporation Advanced Plant Safety & Licensing '

i Energy Systems Business Unit Westinghouse Electric Corporation Energy Systems Business Unit

! P.O. Box 355 Box 355 j Pittsburgh, PA 15230 Pittsburgh, PA 15230 l

Mr. M. D. Beaumont Mr. S. M. Modro l

Nuclear and Advanced Technology Division Nuclear Systems Analysis Technologies  !

Westinghouse Electric Corporation Lockheed Idaho Technologies Company

! One Montrose Metro

Post Office Box 1625 11921 Rockville Pike Idaho Falls, ID 83415

' Suite 350 Rockville, MD 20852 l

t I Enclosure to be distributed to the following addressees after the result of the proprietary evaluation is received from Westinghouse:

Mr. Ronald Simard, Director Ms. Lynn Connor l Advanced Reactor Programs DOC-Search Associates Nuclear Energy Institute Post Office Box 34 i 1776 Eye Street, N.W. Cabin John, MD 20818 Suite 300 Washington, DC 20006-3706 Mr. Robert H. Buchholz GE Nuclear Energy Mr. James E. Quinn, Projects Manager 175 Curtner Avenue, MC-781 LMR and SBWR Programs San Jose, CA 95125 GE Nuclear Energy 175 Curtner Avenue, M/C 165 Mr. Sterling Franks San Jose, CA 95125 U.S. Department of Energy NE-50 Barton Z. Cowan, Esq. 19901 Germantown Road Eckert Seamans Cherin & Mellott Germantown, MD 20874 600 Grant Street 42nd Floor Pittsburgh, PA 15219 Mr. Charles Thompson, Nuclear Engineer AP600 Certification Mr. Frank A. Ross NE-50 U.S. Department of Energy, NE-42 19901 Germantown Road Office of LWR Safety and Technology Germantown, MD 20874 19901 Germantown Road Germantown, MD 20874 Mr. Ed Rodwell, Manager PWR Design Certification Electric Power Research Institute 1 3412 Hillview Avenue Palo Alto, CA 94303

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

i j Review of " Lower Head Integrity Under In-Vessel Steam Explosion Loads" l (D0E/ID-10541)

1. Overall Anoroach The overall approach taken by the authors of the subject report (henceforth

, denoted as the IVSE report) to demonstrate the integrity of an externally 4

flooded lower head under in-vessel steam explosion loads involves decomposi-j tion of the problem into a set of " causal relations" and a second set of

" intangible parameters", quantification of the causal relations by best 1

estimate physics, attribution of probability measures to intangible parame-i ters, and finally, convolution of the best estimate quantification and the i probability measures within the framework of Risk Oriented Accident Analysis

! Methodology (ROAAM). The effectiveness of the approach and that of the j solution depends on adequate representation of the controlling physics in causal relations and on the confidence levels associated with the probability j' measures for intangible parameters. For reference, the approach was proven to

be effective in resolving both the Mark I liner issue and the DCH issue, and j in assessing the in-vessel retention (IVR) strategy for the AP600 design.

l For the issue of lower head integrity of AP600 under in-vessel steam explosion loads, the intangible parameters, identified in the report, are~: '(1) the location and size of the failure, (2) melt characteristic length scale and its evolution, (3) initial size of melt particles and breakup rate, and (4) trigger strength and timing. The causal relations are: (1) premixing (given by PM-ALPHA), (2) propagation (given by ESPROSE.m), and (3) structural loads (given by ABAQUS). The report claims to have taken a bounding" approach with regard to the location of failure. Specifically, the failure location is predicated upon a melt relocation scenario that leads to a stable blockage formation at the lowest region of the active fuel thus making the downward relocation path unavailable, and leading to a sideways release of melt as a consequence of reflector and core barrel meltthrough. For all other intangible parameters (failure size, melt length scale, initial particle size, breakup rate, and trigger timing), either " conservative values" or ranges of values are specified. With this rationale, the causal relations allow calculations of premixture composition, explosion energetics, and structural loads in a " bounding" manner. Next, the concept of fragility is employed to determine if the lower head integrity would be maintained under the imposed impulse loading from explosions. These various steps and/or processes have been reviewed in detail and the discussions are presented below.

2. Quantific'ation of Nelt Relocation Characteristics The melt relocation scenario for the AP600 design is built upon three key processes: (1) blockage formation on top of the core plate, (2) meltpool formation on top of the blockage, and (3) sideways meltthrough of reflector and core barrel leading to melt release.

Enclosure

j l

i i

, 2 j Blockaae Formation i

j The IVSE report argues that as the core materials melt, mainly metallic

zircalloy at about 2000 K will relocate first followed gradually by oxidic J melts, i .e. , 00,-Zro, at about 3000 K. As the melt relocates to the lowest i 25 percent of the core (still submerged in water and thus acting as a " cold  ;

i trap"), it freezes rather rapidly (illustrated in the report in terms of i spacer grid plugging time of 10 seconds or less) and forms a stable blockage

! of 10 cm thickness (metallic) or 25 cm thickness (oxidic). The blockage

remains cool 1.e., its temperature remains about 1200 K (metallic) or 1500 K '

l (oxidic)'as long as it is in contact with water underneath, and the duration is calculated to be about 100 minutes. The plugging time calculation is based

on an energy balance and a crust growth rate formulation. The stable blockage

thickness and temperature calculation is based on a heat transfer formulation.

The duration of stable blockage is also based on a heat transfer formulation.

l The freezing (plugging) time is dependent on the crust growth rate or equiva-4 lently, on the crust growth constant, A. The report mentions that A is

! evaluated incrementally, but does not provide a value (or a set of values) for j it. In the absence of such information, reasonableness of the plugging time i cannot be ascertained. Once the blockage is formed, however, the heat '

j transfer calculation clearly demonstrates that for the given set of values of i radiative heat fluxes, emissivities, etc., the blockage will have a coolable configuration (i.e., minimum thickness and a maximum surface temperature).

1 The duration that the blockage is in contact with water depends on the l downward radiative heat flux, taken to be 0.2 MW/m2 maximum in the report for

] a metallic blockage. For an oxidic blockage, the downward heat flux will be

less than that for a metallic blockage and correspondingly, the contact

! duration will be longer. The heat flux depends, among other factors, on the i conductivity of the blockage material and the power density of the melt within i

the crust. Uncertainties in these factors were not considered in the heat i flux calculations and we recommend the report address such uncertainties.

i i Meltocol Formation i

i The core continues to heat up past the zircalloy melting temperature and in

! about 27 to 42 minutes (after core uncovery), according to the report, the j fuel melting temperature (~ 3000 K) is reached. The molten fuel (UO, and

remaining Zr02 ) begins to slump or drain downward at this point, relocates on

! top of the blockage, and continues to heat up towards a fully-developed molten

! pool. The report estimates that an additional 15 minutes (i.e., a total of j 42 to 57 minutes after core uncovery) is needed for the molten pool to fully

develop. During this process, the reflector and the core barrel also continue to heat up by radiation.

Reflector and Core Barrel Meltthrouah i

i The report estimates that an additional 34 to 38 minutes is needed for the 2 reflector and core barrel meltthrough. This means meltthrough occurs between

} 76 and 95 minutes after core uncovery, still less than 100 minutes required j for the water underneath the blockage to evaporate, according to the report.

i

1 1 -

! 3 i

The calculation of evaporation time (100 minutes) is based on a downward heat i flux of 0.2 MW/m" maximum and, as already mentioned previously, it is not clear if uncertainties in heat flux are accounted for in the estimate. The sideways meltthrough time may exceed the evaporation time for the lower plenum water if uncertainties in either time is accounted for. On the other hand, j

the AP600 design has a flatter radial power profile and a higher aspect ratio,

' which make the sideways meltthrough more likely to precede the core plate meltthrough. Also, the core plate is much thicker in the AP600 design (about i l twice that of operating reactors) so that it acts as a substantial heat sink.

It takes a significant amount of time for the core plate to melt, even under a j

dry lower plenum condition. All these factors combined work to ensure that the reflector / core barrel meltthrough occurs before the blockage / core plate

meltthrough.

Quantification of Melt Release Conditions 4

Given the melt relocation scenario postulated in the IVSE report, melt release rates considered (100, 200, and 400 kg/s) are comparable to the THI-2 scenario

' (~ 160 kg/s). These rates were calculated based on an exit velocity of 1 m/s under gravity draining and exit hole sizes of 10 cm x 10 cm, 10.cm x 20 cm, and 10 cm x 40 cm, respectively. The report claims these numbers to form a reasonable range to bound the release rates, but provides no evidence to l support the claim. For reference, the hole size in TMI-2 was much larger

' (60 cm x 150 cm). Also, in steam explosion studies, it is not uncommon to consider much higher release rates in order to determine the envelope of 4

energetic interactions. It is recommended that higher release rates be considered and parametric calculations be made to determine the release rate

' that would breach the lower head, given all other input conditions the same as in the base run, in order to determine the sensitivity to melt release.

3. Quantification of Premixtures Quantification of premixtures involves specifications of distributions of the intangible parameters - melt length scale and melt breakup rate as well as their evolution with time - and formulation of a causal relation for the  !

quantity of fuel mass in a premixture. The length scale is a characteristic '

melt parameter which represents the initial state of the broken up melt stream and is used for the fragmentation calculation. This parameter and a melt breakup parameter (characterizing the breakup rate), B, are essential in describing the melt breakup and fragmentation processes. The report considers I a single value of melt length scale (20 mm), but it considers a range of '

values of the breakup parameter, p. The causal relation is based on the PM-ALPHA formulation.

The basis for the choice of a 20 mm melt length scale for AP600 is not stated.

Elsewhere (DOE report DOE /ID-10504), different length scales are chosen for

' MIXA06 (6 mm) and FARO L-14 (40 mm) assessment. It appears from the PM-ALPHA documentation (DOE /ID-10502) that one can virtually construct any number of

, combinations of the melt length scale and the breakup parameter, #, which will produce code results that are comparable to experimental data. In fact, by varying the breakup parameter, p, one can arguably simulate a variable length

_ _ . _ . _ _ _ . . _ _ _ _ _ . _ _ ~ _ _ _ . . _ . _ _ _ . _ _ _ _ _ _ _ . _ . _ _ . _ _ . . _ . _ .

4 1 .

}

I

)

l 4

• \

scale as long as the product of the two is kept constant. However, the

} relative influence of such variations on the quantity of fuel premixed may be different and, as such, should be investigated. For this reason, PM-ALPHA j calculations with a range of values of melt length scale is recommended.

4. Quantification of Ernlosion Loads Quantification of explosion loads involves specifications of distributions of the intangible parameters - trigger strength and trigger timing - and formula-tion of a causal relation for the impulse loading from steam explosions.

Triggers of sufficient strength (~ 100 bar) to initiate explosions are considered in the report. Several values of trigger timing are considered, and a total of 24 loading calculations performed as shown in Table I below.

Of the 24 calculations, only 7 cases produced a peak pressure in the range between 200 MPa and 1000 MPa (an indication of the degree of severity).

However, the calculated impulse loads in these seven cases were between 90 kPa-s and 190 kPa-s (the maximum 1000 MPa peak pressure).

value sponding corre(see to the Tab' The report concludes that impulses do not depend strongly on the size of the mixing zone (expressed in terms of 200 kg/s and 400 kg/s release rates). This I conclusion is not generally supported by the results in Table 2. In fact, it  !

does not appear that a clear trend between the impulse load or the peak j pressure and the mixing zone size can be readily established from the above  !

table. To do so and to generalize the conclusion in the report for a wider spectrum of release rates (mixing zone sizes), calculations with higher release rates are recommended. '

l The ESPROSE.m calculations of impulse loads in the KROT0S-38 expe'riment (see DOE /ID-10503) make use of a melt participation factor to account for the l

uncertainty in melt freezing. A participation factor of 1 means the entire amount of melt participates in FCI. It appears from the calculations of the l

KROTOS-38 experiment (see DOE /ID-10503) that good agreement between the calculated and the experimental data (pressure amplitude and pulse width) was observed for a melt participation factor equal to 0.65, meaning that 65 percent of the melt was considered to participate in FCI. What value of '

the melt participation factor was used in the AP600 calculations? For l conservative calculations, a participation factor of I should be considered. i

5. Structural Failure Criteria Equivalent plastic strains were calculated using ABAQUS to determine the response of the lower head for a given impulse load (equivalently, a pressure peak with a corresponding pulse width). The calculated values were then compared to some failure criteria (e.g., percentage of cross section exceeding certain strain values) to determine if the lower head is going to survive (expressed in the report in terms of the failure probability). For impulse loads of interest (90 kPa-s to 190 kPa-s from the quantification of explosion loads), the equivalent plastic strain was calculated to be within 10 percent (Figures 3.4 and 3.8).

l O I

i Table 1. Explosion Load Calculations for AP600 Calc. ID Melt Release Breakup Parameter, # Trigger Timing mummmmmmmmmmmmma Cl-10(0.05) 200 10 (rapid breakup) 0.05 l C1-10(0.11) 200 10 (rapid breakup) 0.11 Cl-10(0.19) 200 10 (rapid breakup) 0.19 l C2-10(0.05) 400 10 (rapid breakup) 0.05 C2-10(0.11) 400 10 (rapid breakup) 0.11 C2-10(0.19) 400 10 (rapid breakup) 0.19 i Cl-20(0.04) 200 20 (intermed. breakup) 0.04 Cl-20(0.08) 200 20 (intermed. breakup) 0.08 I Cl-20(0.095) 200 20 (intermed. breakup) 0.095 Cl-20(0.115) 200 20 (intermed. breakup) 0.115 Cl-20(0.155) 200 20 (intermed. breakup) 0.155 C2-20(0.04) 400 20 (intermed. breakup) 0.04 C2-20(0.08) 400 20 (intermed. breakup) 0.08 C2-20(0.095) 400 20 (intermed. breakup) 0.095 l C2-20(0.115) 400 20 (intermed. breakup) 0.115 C2-20(0.155) 400 20 (intermed. breakup) 0.155 l Cl-nb(0.05) 200 m (slow breakup) 0.05 Cl-nb(0.10) 200 m (slow breakup) 0.10 i

Cl-nb(0.20) 200 m (slow breakup) 0.20 Cl-nb(l.00) 200 m (slow breakup) 1.00 C2-nb(0.05) 400 m (slow breakup) 0.05 C2-nb(0.10) 400 m (slow breakup) 0.10 C2-nb(0.20) 400 m (slow breakup) 0.20 i C2-nb(1.00) 400 m (slow breakup) 1.00

A 4 6 l

a Table 2. Calculated Explosion Loads Indicating Degree of Severity Calc. ID Release Rate Breakup # Peak Pressure Impulse Load t

mummmmmmmmmmum nummmmmmmmmmmmuu l Cl-10(0.05) 200 10 500 90 C1-20(0.08) 200 20 260 90  !

Cl-20(0.095) 200 20 530 130 C2-10(0.05) 400 10 550 130 C2-20(0.08) 400 20 200 110 C2-20(0.095) 400 20 900 130 C2-20(0.12) 400 20 1000 190 The authors assigned failure probabilities (see Table 3.3 and F'igure 3.10) to percentages of lower head cross section exceeding 11 percent strain for given impulse loads and for different loading patterns. This form of probabilistic l quantification is reasonable, in particular, if uncertainties or sensitivities (e.g., material properties) are to be accounted for. It is noted from the fragility plots (Figure 3.11) that the failure probability would be close to unity for impulse loads in excess of 300 kPa-s. This is because the plastic ,

l strain corresponding to such loads can be an order of magnitude higher, which  ;

may not be sustained by the structure. It is recommended that the report '

provide some discussion on how the case of intersecting load and fragility curves (in other words, if the calculated impulse load is 300 kPa-s or more) ,

would be treated within the ROAAM framework.

6. Intearation and Assessment The usual ROAAM approach, i.e., consideration of splinter scenarios, assign- .

ment of probability distributions to intangibles, and convolution of causal l relations with the probability distribution (illustrated in Figure 2.3) was not followed. The reasons cited for this are: (1) a unique melt relocation scenario, (2) bounding approach taken with regard to premixing and explosion calculations, and (3) non-intersecting load and fragility curves. Moreover, it is argued that the bounding approach obviated any parametric and sensitivi- l ty calculations. To confirm this, the report presents ABAQUS calculations from the two most energetic explosions (C2-10(.05) producing a 130 kPa-s impulse and C2-20(.12) producing a 190 kPa-s impulse), and concludes that the equivalent plastic strains are very small in these cases.

Table 3 below summarizes the treatment of intangible parameters identified in the report.

2

+ l

! 7 l i

3- i l Table 3. Treatment of Intangible Parameters for AP600 l

Intangible Parameter PrescriptionintheReport

Location of Failure Single location argued' based on melt reloca-  !

j tion scenario (sideways failure)

! Release Rates

' Release rates of 100, 200, and 400 kg/s used; rates reasonable and comparable to TMI-2;

Melt Length Scale Single value of 20 mm considered; indirectly varied through breakup rate variation 5-i Breakup Rate (Parameter) Three breakup rates (rapid, intermediate, and

slow) considered 3

Trigger Strength Sufficient strength (~ 100 bar) considered for ,

j triggering an explosion  !

Trigger Timing Several trigger timing considered; different i ranges for different breakup rates

! i i

The relocation scenario-dependent treatment (i.e., formation and stability of  :

! lower core plate blockage resulting in a sideways failure) warrants some  !

i uncertainty (sensitivity) analysis, in particular, with respect to blockage .

I properties (e.g., thermal conductivity, viscosity, etc.). Moreover, the claim j in the report that the melt release rate is bounding has not been demonstrated  ;

l through any supporting evidence. Therefore, additional parametric calcula- '

! tions involving this parameter are recommended. Finally, sensitivity studies I are recommended with regard to the melt length scale and melt participation l factor.

d The report presents an argument about reflood FCIs, namely, that a stratified i- explosion of sufficient magnitude (350 MPa peak pressure or equivalently, l 35 kPa-s impulse) to cause structural damage to the lower head is not " physi-i cally possible." The argument is based on the assertion that a molten upper

surface of a metallic melt layer cannot coexist with a water layer on top that is more than 10 cm deep. The water depth calculation is based on the direct vessel injection (DVI) design parameters. Also, the 35 kPa-s impulse limit is

.! based on a 20 percent equivalent plastic' strain, presumably used as a failure l- criterion. Why is the 11 percent plastic strain criterion as in Chapter 3 not 1, used in this case?

{ 7. Review F ry As in other cases of ROAAM applications (e.g., Mark I Liner Meltthrough, DCH, 1 and In-Vessel Retention), the problem of in-vessel steam explosions is well

[ formulated within the structure of ROAAM. The analysis, presented in the i

i

,i

. 1 d

i 8

report, demonstrates that in-vessel steam explosions of sufficient magnitude to challenge the structural integrity of the lower head is " physically unreasonable," but provides no supporting evidence that the analysis is bounding with respect to the melt release rate. In the absence of such evidence, additional parametric analysis using higher release rates is recommended. Also, the analysis does not consider uncertainties in blockage i properties and downward heat flux, and additional sensitivity studies involv-ing these parameters are recommended. Finally, sensitivity studies involving melt length scale and melt participation factor are recommended so that the

.: conclusions in the report can be generalized for a reasonable range of values 4

of these parameters.

1 i

i f

e