ML20011E958

Forwards Independent Rept on Differing Prof Opinion,Per 891102 Request.Fuel Failure Extremely Unlikely During First 7 of Design Basis Loca,Contrary to Differing Prof Opinion

ML20011E958
Person / Time
Site:	Zion  File:ZionSolutions icon.png
Issue date:	11/30/1989
From:	Jerome Murphy, Ross D, Soffer L NRC
To:	NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
Shared Package
ML20011D095	List: ML20011D094 ML20011E949 ML20011E952 ML20011E953 ML20011E956 ML20011E958 ML20011E959
References
NUDOCS 9002230197
Download: ML20011E958 (30)

Text

__

I[ \\

UNITED STATES NUCLE AR REGULATORY COMMISSION a

I f

wAswwovow, o. c. rosss

,/

j November 30, 1989 MEMORANDUM FOR:

James M. Taylor q

s Acting Executive Director for Operations l

FROM:

Denwood F. Ross, Jr.

Joseph A. Murphy Leonard Soffer t

Sam Naff

SUBJECT:

INDEPENDENT REVIEW OF OP0

Reference:

Memo from J. Taylor, same subject, dated November 2,1989 i

in Reference 1 you designated Ross, Soffer, and Murphy to serve as an independent review group for the DP0 of Mr. Robert Licciardo of NRR. You i

authurized us to seek the views of an outside individual; we did this, and we called on the services of Mr. Sam Naff of EG&G. Our report, attached, represents the consensus of the four of us.

l As you requested, I am returning the original DP0 which Mr. Licciardo sent you on October 19, 1989. The record and findirgs of our review are as described in our report, attached.

We can summarize our views as follows:

1.

Contrary to the DPO, fuel failure is extremely unlikely during the first seven seconds of a design basis loss of coolant accident.

Further, the concept of specified acceptable fuel design limits does not apply to such an accident (that term is reserved for anticipated transients).

2.

There is little incremental risk associated with operation of a reactor that the valves (or at least one of two redundant valves)gh likelihood such as Zion with the purge valves open, as there is a hi will close on demand.

3.

The dose associated with a conservative iodine " spike" release during the time that the valves are closing is well within Part 100 limits.

4 There seems to be some confusion as to the regulatory requirements for allowable times and conditions for operation of these large valves.

(We noted that the issue may be moot; Zion does not actually envision routine

. operation with valves open, we were informed).

Perhaps the need to routinely purge reflects some basic design deficiency which should be addressed.

In any case, it is clear that the containment leaktight b

9002230197 900131 PDR ADOCK 05000293 P

PNV

i 2

criterion is better served if one does not have to open large butterfly valves from time to time, or leave them open indefinitely.

In our opinion, NRR should reexamine policy in this area, w

Denwood F. Ross, Jr.

seph A. Hurph Leonard fo'Tferr f t.

is ' c f

Sam Naff Attachments:

i As stated

~.-

k' i { h.P. J JP." x,_

.' '- A}

gso'

[ffi

'2 k'

,_.,u i

^

N g; 3 s

3 p.

M.Q L'; -

c-

~

a s

1 O

i M%W ua W l 3...:

~

e bL n< e r z c.

+

qq;f 2_

,.; 7m

/t -

i y

a3...,,.y 4-1'

• I t

• ', o' 'f

, 3-4

...V." 40ffaite Oose. Consequences

@e

^

s

.r F ':y#

s (T 1W VI..

Conclusions 1

.'y.

a e

1 l.w

..q,

~

k.

)

_L.

5 e

=

{ ? f. ~

3 [' z,.-

3

, f. >

i t

K

% o yu r

' )

. i-tr s[m.~d' 1

'h.

a fhl I

(

p

u.

4

-f' t

4

._ Q '?

'1 7

._g

• s r

1 1

l l

l V.'

y 2

' O] r 3

~

1

~

5-

}4r..-

4 i.

f'y~j.k' t 3 '9

,i - '

, n v

.8 i

y e

4 v.

t b.

,,,k-_

\\

t' M).

+

i

~

~.

hc y'l$,3 '

1. {

_E f

79, e.g '.

1

a 3

h",$4._.

N J.V

^

il

.i_

ym....,-v k

- 4 L

\\

p ga

-j;

=x..y A-g...j[')-.I'NJ "

'%,,z_.-

2

,'O7,"'

4

(:

+

--y@5 -$ l v-V g

4..,;

$~

',5 m4?,

p 3

4 r

. ys b

t4

.t..".

5- %' -

5--

L ',..2 ;I _N '...:q I

4

_'3 7

v. 4 3

r

'_'t'.

N y

'2-7

..'et

..t hm

8 1.

Introduction Robtrt Licciardo of NRR submitted a DP0 to the acting EDO, J. Taylor on October 19, 3989.

On November 2,1989, Mr. Taylor establisned a 3 person independent review group, consisting of D

7. Ross (chair), L. Snf fer, 6nd J. fictphy. Our charge was to complete our review within one month.

In our review we made use of the services of Sam Naff, EG&G, inasmuch as he had prior expertise in the PP,F and LOFT programs.

This is the report of the independent review group.

Mr. Licciardo had previously described his concern in several places, most completely in a July 20, 1989 memo to Frank Miraglia of NRR.

The statement of his problem can be condensed to that shown in Table 1.1 The indepe1 dent review group did the following:

1.

Met with Mr. Licciardo 2.

Developed review sssignments:

a) General - Ross b) PRA - Hurphy c) F. P. Behavior - Soffer d) Fuel Failure - Naff 3.

Met with NRR represente.tive (Kudrick) 4.

Wrote this report We will be glad to amplify on this report, orally, if requested.

1 2Mr. Licciardo was shown this table, and he agreed with this condensation.

2

i C

I

~

i I

l TABLE 1 i

STATEMENT OF PROBLEM i

'1.

Containment Purge Valves May Be Open

-2.

A LOCA may occur 3.

During the first 7 seconds fuel will heatup, and may swell and rupture 4.

Following fuel failure, at least Iodine will be released 5.

Radiodine will release during this tinie, along with containment air and water vapor 6.

Transport offsite will result in calculated doses i '

well in excess of Part 100 7.

Thus, purge valves should not be open during operation.

3 L

r 1

(

)).

Regulatory Requirements As noted in Section 1, Mr. Licciardo is concerned that in the first seven seconds after a LOCA, if the containment purge valves were open (and before they close) there will be fuel heatup, failure, and fission produt:t relesse te the environs.

Consequent offsite doses would be well above Part 1001 hence, those purge valves should be closed during operation, in his opinion.

In discussions with Mr. Kudrick of NRR, it appears that regulatory requirements are quite vague.

(One plant (Crystal River) operates with the valves open almost all of the time.) On the other hand, according to Kudrick, Commonwealth really has little interest in routinely having these valves to purge.

They do not actually plan to open them routinely during power operation; they just do not want to be precluded from doing so.

Purge valve opening is useful in reducing containment pressure, humidity, and radiation levels.

The valves are redundant.

The Standard Review Plan mentions closure times of 5 seconds *, and opening intervals of up to 90 hours0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> per year.

In the DPO, Mr. Licciardo equates a violation of thermal margin (i.e., DNBR) to fuel failure.

He cites GDC-10 as the source of the requirement for specified acceptable fuel design limits (SAFDL).

For reference, GDC-10 is quoted herein:

l "11.

Protection by Multiple Fission Product. Barriers Criterion 10 - Reactor design.

The reactor core and associated

{

coolant, control, and protection systems shall be designed with appropriate margin to assure that specified acceptable fuel design limits are not exceeded during any condition of normM operation, including the effects on anticipated operational occurrences."

^Ihe original Zion Technical Specification called for a closing time of 5 seconds, Commonwealth requested that this be changed to 7 seconds to reflect the actual measured closing time and clean up the Tech Spec.

Some other plants have approval to use valves with closing times of up to 15 t,econds.

This information was provided by Mr. Kudrick.

4

i 4-As one can see, GDC-10 applies to anticipated operational occurrences, and not LOCA.

For LOCA behavior, Mr. Licc'ardo provided Section 3 in his July 20, 1989 memo titled "Arpendix K Evaluations, Fuel Failure, and Fission Product Release."

In his Section 3. Mr. Licciardo mistakes SAFDLs in their use for rare events such as LOCA.

We reviewed the regulatory requirement in Appendix K to Part 50 Section 1.B. as follows:

"B.

Swelling and Rupture of the Cladding and fuel Rod Thermal Parameters Each evaluation model shall include a provision for predicting cladding swelling ana rupture from consideration of the axial temperature distribution of the cladding and from the difference in pressure betwe'en the inside and outside of the cladding, both as functions of time.

To be acceptable the swelling and rupture calculations shall be based on applicable data in such a way that the degree of swelling and incidence of rupture are not underestimated.

The degree of swelling and rupture shall be taken into account in calculations of gap conductance, cladding oxidation and embrittlement, and hydrogen generation, c

The calculations of fuel and cladding temperatures as a function of time shall use values for gap conductance and other thermal parameters as functions of temperature and other applicable time-depend?qt variables.

The gap conductance shall be varied in accordance with changes in gap dimensions and any other applicable variables."

5

1 I

Thus, each plant (including Zion), is required by regulation to produce and use a model for fuel swelling and rupture which is conservative in that "the incidence of rupture is not underestimated." Mr. Licciardo st&ted to us that he is not challenging the adequacy of the regulations.

The only residual challenge must be whether the evaluation model used by Commonwealth meets the regulation.

The argument by Mr. Licciardo is vague in this area and, as stated earlier, he seems to confuse GDC-10 SAFDLs (for anticipated events) with fuel performance requirements of Appendix K.

lie any case, we discuss later in this report (Section IV) the extant data for fuel performance for the design basis loss-of-coolant accident.

The independent review group is of the opinion, should fuel failures be predicted during the allowable opening times of the purge valves, Commonwealth would have the burden of showing that Part 100 limits would not be exceeded.

Asdiscussed.in'oLirRiskSection(III), pre-existingopeningsinthecontainment (e.g., open purge valves) can contribute to risk, if not promptly closed.

In Conclusions (Section VI), we offer our views as to the practice of purge valve operations, generically.

y I

i l.

6 l

S i

!!!. Risk Perspectives l

The risk to the puMic associated with operation with containment purge valves open is driven by the possibility that they do not close when called upon to do 50.

In the NUREG-1150 (Draft 2 for Peer Review) analysis of the Zion plant, preexisting containment leakage was treated as a probabilistic distribution, based on generic PWR operating experience.

The largest leakage source considered was a five inch diameter opening with an associated probability of occurrence of 5x10'8 Containment leakage was found to be a 36% contributor to the conditional mean probability of early containment failure, averaged over all accident sequences (0.014).

Larger preexisting leakage paths, such as those associated with the f ailure of the large containment purge valves to close, were not considered (i.e., it was assumed that the plant never operated with the purge valves open).

Situations where the containment failed early in the accident sequence, largely driven by the loads imposed on the containment structure at the t,(gne of meltthrough of the reactor pressure vessel, were found to be the principal contributor to early containment failure.

A rough estimate was made of the potential increase in risk that might be associated with operation with the purge valves open.

(Note:

Information obtained orally from J. Kudrick, NRR, indicates that the plant does not routinely operate with the large purge valves open, although they are allowed to do so by the plant Technical Specifications.) However, for this analysis it was conservatively assumed-that the valves were opened 100% of the time.

The l

purge valves at Zion are large air-operated butterfly valves, environmentally qualified for the differential pressure they might see at containment design basis conditions, and are designed to fail closed on loss of air pressure.

~

Typically, air-operated valves have an unavailability of about 1x10 3 per demand.

However, in view of the large size of the valves, an unavailability of 3x10'3 per demand was assumed for this analysis.

Thus, the independent failure of the two series valves in a purge line would be approximately 9x10'6 per L

demand.

The limited data available on dependent failure of air-operated valves suggests that the dependent failure unavailability should be approximately 3x10'4 Thus, the conditional probability of a preexisting purge valve line being open to the atmosphere would be about 0.0004 (i.e., dominated by i

l dependent failure).

This contrasts with the likelihood of gross failure of l

l 7

e containment of approximately 0.01.

Even if no credit is taken for deposition or other natural removal processes within the containment, such a large leakage path would only increase the overall plant risk by approximately 3% over that estimated in NUREG-11LO (Draf t 2 for Peer Review), and the predicted risk would remain well below the quantitative design objectives of the Commission's Safety Goals. Note that this evaluation assumes that no significant release to the environment occurs after the valves successfully close, based on the analyses presented in Section IV of this report.

We note that 10COR Technical Report 23.12, Section 6.3.3, transmitted to the Commission in 1987, addresses containment leakage through the purge lines at Zion.

This report indicates that purging is done through 10" lines in the i

containment pressure and vacuum release system, with an average usage of 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> per week.

If the same assumptions are made as above, the conditional probability of a preexisting 10" diameter vent path being open would be approximately1x10[,L.

This would have no observable effect on the risk assessments given in NUREG-1150 (Draft 2 for Peer Review).

Based on the foregoing, we conclude that operation with the purge valves open does not have a significant effect on the overall plant risk.

)

8

e r

r l

IV.

Research insights Inasmuch as Mr. Licciardo focused his concerns mostly on fuel behavior during LOCA, we have focused our evaluation on that topic.

Two main conclusions were

[

developed:

(1) The memorandum from G. N. Lauben to M. Wayne Hodges of August 21, 1989, entitled " Comments on a DPV Concerning Early Blowdown cladding Rupture During a Large Break LOCA," was quite thorough and completely adequate to address concerns about the likelihood of early fuel f ailure during a LBLOCA as outlined in the subject DPV.

We agree with the conclusions given in that memorandum.

-(2) Rupture of high burnup fuel rods during the first seven seconds of a LBLOCA, even if in the highest power section of the core, is extremely unlikely.

Rupture of such rods, when in the outer peripheral, lower power regions of thr core, as normal fuel management dictates, is even less likely.

In the following sections we provide the bases for these conclusions.

Reference la supplied the details related to a Differing Professional View (DPV) filed by Robert Licciardo concerning issuance of a SER to Zion allowing full power operation with open 42-inch containment isolation valves.

The DPV is based on the postulation that early rupture of high burnup fuel rods, prior to containment isolation, would occur and result in larger than allowable (i.e., greater than Part 100) offsite thyroid dose.

This section addresses the likelihood of significant fuel cladding failure during a large-break LOCA prior to the seven second closurc time of the 42-inch containment isolation valves.

• Section IV references are at the end of Se-tion IV.

9

r-e i

At the onset (See Section !!) it should be recalled that compliance with 10 CFR

$0 Appendix K effectively precludes significant early cladding rupture.

i Section I.A of Appendix K states:

"A range of power distribution shapes and peaking factors representing power distributions that may occur over the core lifetime shall be studied and the one selected should be that which results in the most severe calculated consequences, for the spectrum of postulated breaks and single failures analyzed."Section I.B states:

"Each evaluation model shall include a provision for predicting cladding swelling and rupture from consideration of the axial temperature distribution of the cladding and the difference in pressure between the inside and outside of the cladding, both as functions of time.

To be acceptable the swelling and rupture calculations shall be based on applicable data in such a way that the degree of swelling and incidence of rupture are not underestimated." As pointed out in Reference 2 Vendor analyses have shown that extensive swelling would occur prior to rupture during blowdown and as a consequence EM calculations which predict early rupture also predict an exceedance of the 2200 F PCT limit, which would require a reductioh in peak linear heat rate.

For this reason, early rupture is not experienced.

(A) Experimentally Measured and Calculated fuel Temperatures.

Extensive large break LOCA experimental and calculated data are available, particularly for the blowdown portion of the transient.

A brief summary

('

of some of the more relevant ones follows:

l (1) Power Burst Facility (PBF)-

The PBF LOCA Test Program (References 3, 4 and 5) was specifically i

1 designed to investigate fuel rod ballooning and rupture as a function of MLHGR*, clad temperature and fuel crystalline structure, pressure i

• Maxium linear heat generation rate usually expressed in KW/ft, and regulated in reactors through allowable peaking factors.

10

i i

1 differential across the cladding, burnup and irradiatioh under thermal-hydraulic conditions similar to those which could exist during a large break LOCA.

The experiments used nuclear fuel rods installed in a special test train within the PBF reactor.

Clad temperatures, internal pressure and fwl rod length were measured during the experiment.

Figure 1 shows the designed, and approximate achieved clad temperatures.

Table I gives a summary of cladding deformation data.

Test LOC-11 is not included because, while cladding collapse and ballooning did occur, no cladding was ruptured.

During the entire 1.0CA Test Program only two cladding ruptures occurred-prior to seven seconds. Rod 7A in LOC-5 ruptured at 2.75 seconds, but under temperatures much higher than one would expect for even the hottest portion of the core during the blowdown portion of a LBLOCA.

Rod 11 in LOC-6 rupture at 5.2 seconds, but water is thought to have leaked into the rod prior to the experiment and flashed during the experiment giving high internal pressure.

(2) FRAP-T6/BALON-2 The FRAP-T6 Code (Ref. 6) with its clad ballooning model (Ref. 7) has been shown to well predict cladding deformation and rupture over a wide spectrum of reactor off-normal conditions including large break LOCAs.

It has been assessed against an extensive data base. m uding the:

German KFK FABIOLA and REBEKA eltetrically heated single-rod and bundle tests, ORNL electrically heated multirod burst tests, PDF and LOFT nuclear rod l

tests.

When comparing data from the various experiments with each other, in general the trends agree well although there is some scatter in the data in particular for irradiated fuel, in comparing FRAP-T6 predictions with the data base the code is seen to somewhat overpredict fuel rod temperatures, in particular after DNB, but the comparisons show that the i

code accurately calculates the extent of cladding deformation and the time 0

of rupture.

As an example, Ref. 8 reports an uncertainty of 145 K (81 F) 11

4 I

when the burst temperature is in the range of 1160 K (1630 f).

Figures 2 and 3 show the FRAP-T5 cladding burst temperature as a function of

. Tangential stress and Hcop stress, respectively.

(3) loss-of-Fluid Test (LOFT) Program During the NRC LOFT Program (Ref. 9) the LOFT Reactor was successfully l

used to study PWR behavior over a wide range of accident conditions including LOCAs, anticipated transients, and ATWS.

During the NRC Program, LOFT accumulated a peak fuel burnup of 3300 Mdd/MTU and 27 simulated accidents which resulted in nine core uncoveries and three occurrences of fast power ramping of the fuel.

During this time, the LOFT fuel never experienced a fuel rod break or rupture even with the adverse effects of densifying fuel.

The LOFT core utilization since 1977 has included three fuel cycles.

The LOTT fuel has~been exposed to the test sequence summarized in the following paragraphs.

The unpressurized Al center fuel module was used for the initial nuclear experiments and was replaced after exposure to three large-break LOCEs, accumulation of 987 mwd /MTU peak fuel burnout, and before a full power (52 kW/m peak) preconditioning status had been achieved.

Poolside examination of the center fuel bundle determined that it had sustained no fuel damage and was reusable even though some of the cladding had exceeded the temperature threshold for Zr recrystallization during LOCE L2-3.

The A2 center fuel module was replaced af ter exposure to two inter-mediate-break LOCEs, six small-break LOCEs, and five operational-transient experiments, one of which (f tre'ne load increase) caused a rapid power increase from 39 to 45 kW/m in 10 sec. at the peak power zone.

During this time there was an accumulation of 2222 mwd /MTV peak fuel burnup and achievement of a full power preconditioned status.

The rapid power increase was sustained after achievement of the full power preconditioned 12

i status.

The final experiment for this fuel module caused some of the cladding to exceed the recrystallization temperature, the fuel module was replaced to improve the core measurement capabilities and is judged to be in a reusable condition.

The F1 center fuel module has prepressurized fuel and was exposed to one large-break LOCE, three small-break LOCEs, and four operational transients of which two (control rod withdrawals) caused rapid power increases (47 to 53 kW/m in 7 s and 47 to 56 kW/m in 58 5 at the core peak-power zone).

It accumulated 765 mwd /MTU peak fuel burnup and achieved a full power preconditioned status.

The rapid power increases were sustained before a full power preconditioned status had been achieved, with some of the cladding in a recrystallized condition and with analysis indicating that up to 4% swelling of the hottest fuel rod's cladding may have occurred in l

LOCE L2-5.

1 The peripheral fuel models were exposed to all of the experiments since September 1977 and accumulated 3317 mwd /MTU peak fuel rod burnup.

Some of the cladding exceeded recrystallization temperatures during the experiments.

Visual inspections of the peripheral fuel module surfaces, exposed by removal of the center fuel module, showed conditions which are judged to be normal for PWR fuel rods in service.

The peak power zone of the peripheral fuel is '83% (43.3 kW/m at full power) of the center fuel bundle peak power zone.

l.

During the OECD-LOFT Program, four additional large-break LOCA Experiments were successfully performed with fuel f ailure occurring in only LP-FP-1 l-(Refs. 8 and 10) which was designed to rupture fuel for fission product measurement purposes.

l Reference 8 reports the best estimate predictions for OECD LOFT Project Experiment LP-FP-1 in which 8 of 22, 6% enriched, prepressurized (2.41 MPa, 350 PSIA) fuel rods were ruptured.

The experiment was designed such that all 22 pressurized rods had a high probability of rupturing.

l l

However, due to unexpected leakage of subcooled water into the upper 13

plenum high cladding temperatures were not reached until late in the experiment and only 8 rods actually ruptured.

Figure 4 shows predicted PCTs and differential pressures across the cladding.

Reference 11 reports the results of a TRAC-PF1/ MOD 1 BE calculation of a i

LBLOCA and the uncertainty bounds on the calculated, PCT.

An extensive data base, of which the LOFT data formed an important part, was used in quantifying the PCT uncertainty.

The best-estimate blowdown PCT was 868 k (1103 F) and with the added uncertainty, the highest blowdown PCT to the

+

95th percentile was 1059 K (1447 F).

Reference 2 showed that this PCT 0

plus uncertainties could be extrapolated down to 989 k (1320 F) for high t

burnup Zion 15 x 15 fuel which would have a MLHGR no greater than 6.4 kW/ft.

Figure 5 shows the calculated best estimate PCT nistory for the CSAU case.

Reference 12 g1ves the results of a study Aprformed to show the proto-typicality of the LOFT LBLOCA results to Zion 1.

Calculations were performed for Zion 1, using Relap 4/ Mod 6, at 70%, 110% and 150% of nominal power (3238 MW(t), PLHGR = 10.9 kW/ft.), and bench-marked against the LOFT Experimental results.

Figure 6 shows PCTs in the central portion of the core and also at the hot spot for the hot pin calculation.

(4) TRAC-PF1/ MOD 1 Calculation Reference 13 reports the results of LBLOCA calculation for Zion under both conservative (with station blackout) and minimum safeguards conditions.

The blowdown PCT's were 950 K (1250 F) and 1100 K (1520 F) for the minimum safeguards and conservative conditions, respectively.

B.

Expected Differential Pressure Across Cladding at 7 Seconds Reference 2 gives an estimated internal cladding pressure of 2500 PSIG (17.2 MPa) for normal operation of 15 x 15 fuel after 50000 MWD /MTV. For a primary coolant operating pressure of 2240 PSIG, this gives a cladding DP of 260 PSID (1.8 MPa) under operating conditions.

Referring to Figure 4, 14

i c

differential pressure across the cladding appears to increase at a rate of about 1 MPa in 7 seconds early in a LBLOCA.

Adding this to the normal operating OP gives an expected DP across the cladding seven seconds after LOCA initiation of about 405 PSID (2.8 MPa).

Reference 1 gives an engineering hoop stress of 4910 PSI (33.9 MPa) for 600 PSID across 15 x 15 fuel cladding.

At 405 PSID, this reduces to a hoop stress of 3310 PSI (22.9 MPa).

Referring to figure 3, the minimum expected rupture temperatures for a hoop stress of 22.9 MPa is seen to be 1150 k (1610 F).

C.

Summary In reviewing the experimental and calculated data for LBLOCA, the highest blowdown PCT found under realistic conditions was 950 K (1250 F).

The CSAU Methodology gives a maximum PCT, after adding uncertainties, of 1059 0

K (1447 f), and calculations performed for Zion under very conservative (station blackout) conditions gave a PCT of 1100 K (1520 f).

According to figure 3, under even the higher 1100 K PCT, cladding rupture would be expected to occur at 35 MPa (5130 PSI) engineering hoop stress.

The expected engineering hoop stress under LBLOCA conditiont at 7 seconds after rupture.is about 22.9 MPa (3310 PSI).

The lowest expected rupture temperature for a hoop stress of 22.9 MPa's is about 1150 K (1610 f).

In sunanary, even under the harshest expected conditions for the core hot spot, fuel cladding rupture for high burnup 15 x 15 Zion fuel should not be expected to occur prior to 7 seconds into a LBLOCA.

l l

1 15 l

.i

REFERENCES 1.

Robert B. A. Licciardo Memorandum to Frank J. Miraglia Differing Professional View (OPV) Concerning Containment Isolation Valves at Zion, July 10, 1989.

2.

G. N. Lauben Memorandum to M. Wayne Hodges, Comments on A DPV Concerning Early Blowdown Cladding Rupture During A large Break LOCA, Aug. 21, 1989.

3.

Jay R. Larson, et al, PBF-LOCA Test Series, Test LOC-11 Test Results Report, NUREG/CR-0618, April, 1979.

4.

James M. Broughton, et al PBF LOCA Test Series Tests LOC-3 and LOC-5 fuel Behavior Report, NUREG/CR-2073, June 1981.

5.

James M. BrouRbton, et al, P8F LOCA Test LOC-6 fuel Behavior Report, NUREG/CR-2244, April 1983.

6.

L. F. Siefken, Developmental Assessment of FRAP-T6, EGG-CDAP-5439, May,

1981, 7.

D. L. Hagrman, Zircaloy Cledding Shape At failure (BALON-2),

EGG-CDAP-5379, July, 1981.

8.

E. W. Coryell, H. G. Glaeser, Best Estimate Prediction for OECO LOFT Project Fission Product Fxperiment LP-FP-1, OECO LOFT-T-3703, Oct.1984.

9.

Charles L. Nalezny, Summary of Nuclear Regulatory Commission's LOFT Program Research Findings, NUREG/CR-3005, June, 1983.

10.

J. P. Adams, et al, Quick-Look Report on OECD LOFT Experiment LP-FP-1, OECO LOFT-T-3704, March 1985.

l 11.

N. Zuber, et al.

Quantifying Reactor Safety Margins:

Application of Code Scaling Applicability, and Uncertainty Evaluation Methodology to a Large Break Loss-of-Coolant Accident, NUREG/CR-5249, Oct., 1989, 16

11.

c; o.

j

.c i

i

-i 12..

Lambert' Winters, Large break Transient Calculations'In A Commercial PWR

.}

g

.and LOFT.Prototypicality Assessment, EGG-LOFT-5093 April 1980.

13.

G. Everett Gruen, James E. Fisher, TRAC-PF DMOD 1 US/ Japanese PWR conservative LOCA Prediction, NUREG/CR-4965, July, 1987.

i

.i e

e

l i

'f I

l

~

i b

i i

e

.I 1

j.

f'.

fl lg.

17 L

L...

i s'

r r

I l

1600 1500 NRC licersin(llm.it 1400 -d Test LOC 5 (1350 K) 4 1300 E

a +d Test LOC 3 (1190 K)

E 1200 13 2a p -

1100 Testa LOC-6 and LOC 11 (1070 K)

- 1000 -a F

]

900 p

c 000 P

700 600 i

e i

i I

I i

g 0

5 10 15 20 25 20 35 40 Tims (s)

INEL A 1T 991 Figure 1.

PBF LOCA program target cladding temperature histories.

t a

18 l

C

• i2 Jgy u-

$b o

st a

=

=

e a

i i

e ya M.

B B

E I

E E

I g ;g it it is in [is it jEa' s

g: jE iE E gE je se

['

.jE e

e r

I[-

P l

l n

n e

9

,1 W a

=

a e

d 3

a W

a e

8 mJ I

?

lhw"9h a

E a

a a

a n

i i

a d

O 6

6 6

6 6

6 6

4 I

s 8

~5 l1 I

i 1

=lx s

1 a

j j

l q

.S 3

3

.E 5

,g E,_

2 2

9 2

9 2

s l

9 2

2

,l E

g "la^~n 6 1 e

e e

a a.

4 g

m3 g

g g

I 3

g a

E g

.g n

El

+

L

.g j 3,g_

r.

s a

=

e

=-

x

~e q

L

.3

'u" I'!

1.

x u

h I

l

}

f5 I

0 1

8*

I Il ji

}

l Ji

=

s t

e e

t $

5

} }

I t

l 7 7 )

E g

Ih 0

S,'

E, E,

R, E

l '-

a a,.

a n

a

(

_ge 4

h

/'

f

. [

1 l

I ~5 l

E i

.c n

m e

e E g

.E ii C

D wI

.I w

J g

M f

l x

?

l2 $N 1 1 x s

r r 0-e o

g 8

8.a 8.J e

4 4 (t

9

>=

a 19

1 x

j

= i :

350 o Chapman 28 K/s in steam o Chapman 9 K/s in steam s Chapman 4 K/s in steam

'00 e Chapman 0 K/s in steam,,,

. Chung 5 K/s.

In vacuum o Hobson 55 K/s in argon NO e Hobson 28 K/s in argon o

a Hobson 14 K/s in argon T 250 e Hobson 6 K/s -

in argon -

g

• Wiehr in air

? Kart in reactor 2

e m E

M n

+

2 200

=

e.

Ee S150 e

c*

a F

\\

2

\\

I e

o i

41M e

o 5

~

e\\ 8 i

50 e

L

\\

o.

O o

W O

O l

0 1000 1100 1200 1300 1400 1500 Burst temperature (K)

INEL 2 0081 J Figure 2. Local tansential stress at failure versus temperature.

I l

i.

I

'O

m i,,

)

1450 1400 a

Constant pressure runs 1350 e

P = 1 HPafs runs 1300 1250 2

v o

1200 3

ro g

1150 E

a- -

--ve s

a100 K/s 1100

/

75 K/s a f

a 50 K/s

_ +,

40 K/s 1,.-

30 Kfs 1050 7

20 K/s 'a

/

1000 f

I 10 K/s 8 K/s

_ _ __ _ __ _ + h.

/_ __

/

o Kf s-4 K/s g gf, 950

/

I I

I' I

i I

i i

1 900 0

25 50

.75 100 125 150 175 Engineering Hep Stress (MPa)

MRST correlations compared with constant pressure and increasin; pressure model calculations.

Figure 3.

r 5

1300 t

1200 coid see

' b:

breet eiesed

'"'i 1100

\\ s

~

,........*.....',,,#D%'..

g to00 N

p..

.,,'3 e

$ 900

...........,,,......***,#I_,,,,,,,,,, _ I ECC on

-Q -

,,q s

800

- p'j~s a

F '700 Center fuel bundle,8% enriched

-1

- Center fuel trundle. 4% enriched

/

-- Hot rod peripheral bundle 600 Q 0.64 m (26.5 inch) elevation 500 O

25 50 75 1CO 12 5 L

Time (s) t,........i.

Maximum Fuel Cladding Temperatures r

6 1300 1200 l......**........

4 110 0

.a g

g3 1000 -

~.y e

900 5

o 2 '-

/

Pressure e

800 st j

~.. Temperature l

l 700 1 -l

'.}

600 0

500

~"

O 20 40 60 80 10 0 Time (s)

L:3.au44.so Cladding Pressure Differential vs. Peak Cladding Temperature Fivre 4 t P-FP-I BE PredictL*N 22

.k m

-c.

'N o

.u.

I 1500 r

i NOMINAL CASE i

I g

4 f f4 GLY

- f '6Loe*D**N PEAet g

1 v.

y i

w 1000 -

4' E

r, i

a to 6-4 l

E

-(pg ro w

l n-i 2

i' s

j W

I-o 500 7

l J

o I

1 O

O 50 10 0 150 Time '(s) l NOMSM4 b

ON vN-wa

~w r-

-r

+,s e

e

-s v.w

--i~w..,erv,.

w-<

,,,nc

~,o

+--ww, mamev--:.w+'

l.

j g

g g

g i

o 70 power A

110% power o

150% po-er x

1000' 7""

b

\\

-=

l-t,

750.

a l

E

</

r e

j 9

=._

e

_ _ __ y -

u e

500.

b 3

0 t,

i 250.

(

0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0' Time after ruptwee ist L

Temperature of hot spot in the core for Zion I hot pin calculations.

\\

i 900.

\\.

L o 70r power 4.I102 power Q

800.

O 1502 power 900. -<

~

^

e i

6 0 0.i

____[

m a

500.

m

%00.

0.0 5.0 10.0 15.0 20.0

-25.0 30.0 35.0 Time after rupture (s)

Cladding temperature l'n middle part of core for Zion 1 calculations.

Fisure c

24

.9 V.

Offsite Dose Consequences

.We decided to evaluate the approximate offsite dose from iodine which might be in the. reactor coolant system, given a conservative iodine " spike. We did not concluded iodine release from ruptured fuel rods, for reasons given in Section IV.

-Offsite dose consequences were independently calculated assuming that a pre-existing iodine spike of 60 uci/gm Iodine-131 equivalent was present in the reactor coolant system, and that the Zion 42-inch purge valves are at a 50-degree position and close within 7 seconds.

Using information taken from R. Licciardo's memo of July 20, 1989 to F. J.

Miraglia, " Differing Professional View (DPV) Concerning Containment Isolation Valves at Zion", the total mass of reactor coolant released into the containment within 7 seconds is approximately 272,000 pounds.

With the purge

~

valves at the 50 degree position, the total mass of reactor coolant system (RCS) inventory discharged to the environment is 5,379 pounds prior to closure.

The total activity released to the environment is, therefore Q = 5379 lbs. x l-454 m x 60 uci = 146.5 curies 1b gm I-131 equivalent L

and the dose to the thyroid becomes Dthy = Q * (X/Q)

• BR

• DCF-l

~4 3

where BR is the breathing rate = 3.47 x 10 m /sec. and DCF is the thyroid dose conversion factor for Iodine-131

-6

= 1.48 x 10 rem / curie inhaled 3

l The value used for X/Q was 5 x 10'4 sec/ m, which is stated to be the value l%

used by the NRC-staff (See Exhibit 2 of Licciardo memo of July 20, 1989) l-l 25

- gm

?

4

- then the dose becomes

~4 Dthy = 146.5 x 5 x 10 x 3.47 x 10'4 x 1,48 x 10-6 = 37 6 rem

.When added to the LOCA containment leakage dose of 123 rem, the total thyroid dose equals about 161 rem, which is within the guideline value of 10 CFR

{

Part 100,-and is, therefore, acceptable.

4 e**P 4

t%

_g-VI.

Conclusion We have examined the governing regulation, the risk aspects, fuel failure data, and fission product release phenomena.

Mr. Licciardo is of the opinion that following the first few (7) seconds of a LOCA there will be fuel failure, fission product release from the fuel to the containment, and while the purge valves are open, substantial release to the environment.

We conciude to the contrary.

Conservative licensing basis experiments which unrealistically produce high temperatures do not show such failure (Section IV).

Commonwealth, as well as other electric utilities, is required by regulation to calculate fuel performance for design basis LOCAs.

Further, a realistic appraisal vs. conservative Appendix K shows much lower clad temperatures.

Further, the operating mode of Zion is such that the plant is not very likely (if at all) to be in a purge mode, and this further reduces the probability of an untoward leak of fission product release from the containment.

It is far from clear that substantial amounts of radio-iodine would be in the vapor phase, and available for release through the vent.

It seemt, to us that Dr. Murley's Standing Review Panel (Congel, Rossi, Miraglia) in their report of August 31, 1989 did a good job, and this should have satisfied Mr. Licciardo.

In particular, we observed (Section IV) that the memo prepared by G. N. Lauben was thorough and adequate.

l In reviewing this Differing Professional Opinion, we have become aware that there may be several plants that operate for a significant period of time with l'

large purge lines (36" in diameter or greater) open between the containment atmosphere and the environment.

(It appears that Zion is not among them, however.) While such an operational practice does not appear to have risk j

significance at Zion, it is our opinion that the original basis for l

L l.

27 l

+

t authorizing routine venting or purging should be reexamined.

It would be more in keeping with the intent of the concept of maintaining multiple barriers to fission product release to limit venting or purging during operation to those situations where it is necessary to reduce the inventory of radioactive gases within the containment to permit needed maintenance and testing.

On a short term basis, we feel containment venting or purging may be authorized to control the environmental conditions inside containment arising from problems associated with inadequate HVAC or equipment insulation, leakage of instrument or plant air systems, or similar faults, to preclude exposing important safety equipment to adverse conditions.

However, in the longer term, we suggest that a root cause analysis may be appropriate to determine if an underlying problem should be fixed to alleviate the need for a continuous vent / purge.

In our opinion, steps should be taken to minimize such use of large purge or vent lines during normal operation on a routine basis.

L l

1 28