ML20003A264

Evaluation of Defective Spot Welds in Calvert Cliffs Nuclear Plant High-Density Spent Fuel Storage Racks

ML20003A264
Person / Time
Site:	Calvert Cliffs
Issue date:	01/12/1981
From:	Milos R NUCLEAR ENERGY SERVICES, INC.
To:
Shared Package
ML20003A263	List: ML20003A262 ML20003A264
References
81A0824, 81A824, NUDOCS 8102030277
Download: ML20003A264 (29)

Text

Enclosure 2 O

I DOCUMENT NO. 81A0824 REV.

mr 1

29 NUCLEAR ENERGY SERVICES. INC.

PAGE OF EVALUATION OF DEFECTIVE SPOT WELDS IN THE CALVERT CLIFFS UNIT 1 NUCLEAR PLAN T HIGH DENSITY SPENT FUEL STORAGE RACK 5 CDW{hfb prem Acciication Preoareo By Date 5134 R. Miles 1/19/oi APPROVALS TITLE / CEPT.

SIGNATU9E i

CATE NM l

Structural Engineering Engineered Products d2 3374

//

./

/-/2-A /

V. P. Engineering p1/)4 A T4

[-/2-8!

Projact Manager

/

/ "- / 2 - P/

g1

).Pi i-i.r-si

a. a. Man Eer 1

1 l. '

81A0824 REVISION LOG

. g AG:

OF NUCLEAR ENERGY SERVICES, INC.

l",*[ f Qf DEscalPTiON l

APACvAL CATE l

See CRA No. 1677 1

1/21/S1 Minor revisions on ces. 1,2,3,7,16,21 i

l l

l l

i i

I i

l I

\\'

i i

I l

I I

l t

i l

i i

l i

i l

l l

l i

I i

l I

i l

l l

l l

l l

I I

l l

l i

ii i

i i

e i

j i

l l

l l

l l

l l

l l

i i

l i

t l

ii l

I I

l

.o....,m,,

I

DOCUMENT NO.

PAGE 05 NUCLEAR ENERGY SERVICES, INC.

TABLE OF CONTENTS Page 1.

INTRODUCTION 4

2.

SUMMARY

5 3.

WELDING PROCESS CONTROL 6

3.1 2 x 2 Cell Module Fabrication 6

3.2 Welding Process 6

3.3 Pull Tests 7

3.4 Welding Parameter Adjustments 7

4 NONCONFORMING 2 x ? CELL MODULE 9

4.1 Description 9

4.2 Structural Analysis and Load Test 11 l

4.3 Ultrasonic Testing 14 4.4 Shielding Gas Composition Test 15 4.5 Frequency of Occurrence 16 5.

RANDOM UNFUSED WELDS

S 6.

UNFUSED WELD SENSITIVITY ANALYSIS 19 6.1 Method of Evaluation 19 6.2 Results of Evaluation 20 6.3 Conclusions 24 7.

WELD PARAMETER VARIATION TEST 25 8.

REFERENCES 26 l

l FCRM = NES 2C5 2/80

81A0824 DOCUMENT NO.

PAGE OF NUCLEAR ENERGY SERVICES, INC.

1. INTROGdCTION During the fabrication of the Calvert Cliffs Unit 1 Nuclear Plant high density spent fuel storage racks, a 2 x 2 cell module was f abricated with four of its 40 vertical rows of spot welds exhibiting defective (unfused) welds. Originally, the assumption was made that approximately 50 of the 200 modules could be similarly aff ected. Based on subsequent tests and investigations, however, it is now estimated that the number of these nonconforming modules is significantly less than 30.

A structural analysis and a simulated load test were performed to evaluate the effect of these nonconforming modules. The load test and analysis demonstrated that even if 13 of the 25 modules in a rack structure were nonconforming, the structure still meets the requirements for Seismic Category I structures and is therefore structurally adequate to store fuel in a safe geometry during and af ter all anticipated loading conditions.

The results of the analysis, tests, and other investigations concerning the nonconforming 2 x 2 cell module were discussed during a meeting at Calvert Cliffs on November 14, 1930, with BG&E, NRC, MPC, and NES personnel. At the conclusion of that meeting, the NRC requested that a written submittal be prepared to discuss various aspects of the fuel rack welding problems and subsequent investigations. This report summarizes the weld-related topics discussed at the November meeting and provides additional information developed since that meeting.

l l

l l

l l

l 4

l l

l l

i l

i l

l FCAM 8 NES 205 2/80

DOCUMENT NO.

11 m

5 29 PAGE 08 NUCLEAR ENERGY SERVICES. INC.

2. SUMM ARY Since the discovery of a 2 x 2 cell module with four rows of unfused spot welds, many and varied tests, structural analyses, and investigations were performed to determine the extent of the problem and its effect on fuel rack integrity. Although it was originally suspected that up to 50 modules could have similar defects, it is now believed tha no more than three modules were aff ected.

Structural analysis and load tests were performed to demonstrate that, even under the assumption that 50 nonconforming modules actually were made, the fuel racks would still be capable of storing fuel in a safe geometry during and af ter a!! anticipated loading conditions.

An unfused weld sensitiv:ty analysis was also made to demonstrate that the Calvert Cliffs fuel racks could tolerate a large number of unfused rows of welds or randomly unfused welds and still meet Seismic Category I structural requirements.

l l

l 1

i FCRM

• NES 2C5 2/80

DOCUMENT NO.

e W 824 11

=

NUCLEAR ENERGY SERVICES, INC.

3. WELDING PROCESS CONTROL 3.1 2 x 2 CELL MODULE FABRICATION The basic structural element of the Calvert Cliffs fuel storage rack is the 2 x 2 cell module. The module walls are fabricated f rom 0.060 inch thick stainless steel. Single cells were first assembled by welding identical halves sogether (Parts A and B in Figure 3.1) along the seams. This welding was accomplished by manually spot welding every three inches along the two vertical seams, using a template to control the three-inch spacing.

The neutron absorbing poison was then placed on two sides of the previously completed single cell, and an L-shaped wrapper (Part C in Figure 3.1) was then welded to the single cell. The welding on this assembly was made automatically, two simultaneous welds per side, with the welding machine stopning and welding at three-inch intervals along the cell length until 56 locations had been traversed.

Finally, four of the latter sub-assemblies were fixtured into a 2 x 2 assembly.

Poison material was placed on each of the top two cells and was covered by an outer sheet (Part D in Figure 3.1). The outer sheet was automatically welded to the single cells by making four simultaneous welds per side, with the welding machine stopping and welding at three-inch intervals along its length. The assembsy was rotated 900 and the process was repeated until all four sides of the module had been welded. The 2 x 2 cell module was then complete.

3.2 WELDING PROCESS The maximum production rate for single cells was 10 per five-hour shif t and for 2 x 2 cell modules the maximum rate was three per five-hour shif t.

Prior to the start of each five-hour shift, the tungsten electrode of each welding gun was inspected and sharpened, and the electrode distance to workpiece was checked and set.

Before starting production welding, each welding gun was fired l

i FCRM

• NES 2C5 2/80

81A0824 DOCUMENT NO.

PAGE OF NUCLEAR ENERGY SERVICES. INC.

a sufficient number of times (10 to 20) to warm up t!'e machine and stabilize conditions. At that point, a test sample was welded by exh gun and the samples were pull tested to demonstrate that the minimum pull test requirement (1300 lbs) was exceeded. Pull test results usually exceeded the minimum by 700 to 300 lbs. If excessive penetration (melt through) occurred during production welding, the welding was stopped, adjustment to current or arc time was made, and then the pull test requirement was verified with additional samples. At no time was current or weld time adjumd during production welding without pull test verification.

3.3 PULL TESTS In addition to qualifiying the welding process on a per shif t basis, the welding parameters for the arc spot welding process shown on the Welding Procedure Specification (WPS)I were established by parametric evaluation of hundreds of pull test results. Pull test specimens consisted of two 1 x 5 inch samples of production stainless steel material, joined by lapping one piece for a distance of two inches over the other, and then spot welded.

3.4 WELDING PARAMETER AD3USTMENTS 1

1 Three parameters on the WPS were varied during production welding. bs rated l

earlier, current and arc time were adjusted as required to assure that rP imum pull test strength was exceeded and that excessive melt through dH nt occur.

i The third WPS parameter varied was the shielding gas cyNetion.

Under various conditorn the shielding gas used was either 100o6 Aryr, or 98% Argon l

l plus 2% Helium. Although this minor variation is consid Jr a have little or no effect on spot weld strength, shielding gas comp < tHM is an essential variable and WPS allowed only 100% Argon and, similarly,,m 100% Argon was qualified on the Procedure Qualification Record (PQR).

No other we4utng yarameter specified on % + S/PQR was varied.

m.v.-

FCAM 8 NES 205 2/80

81A0824 DOCUMENT NO.

PAGE OF NUC1. EAR ENERGY SERVICES, INC.

~

l FART PART A OR B C

FART D 9

I I

.y.

.y.

".y.

~

n m

r I

l, i

r a

l t

l i

I t

V

(

4srM.

i.

, 6 i

(1- _

,-e.

y

..L!r -

M i

+.

.g.

l l\\

\\

/l

'i 1

-A i

B r

(

w

'h u

\\

'C

-D 1

1 i

2X2 CELL MODULE ASSEMBLY FIGURE 3.1 l

FORM s NES 205 2/80

DOCUMENT NO.

31A08:4 Nr,E C5 NUCLEAR ENERGY SERVICES. INC.

4. NONCONFORMING 2 x 2 CELL MODULE

4.1 DESCRIPTION

During the fabr. cation of the Calvert Cliffs Unit 1 Nuclear Plant high density spent fuel storage racks, a 2 x 2 cell module was damaged by an accidental drop during handling. The 2 x 2 cell module was set aside and later dismantled to recover the associated poison material. Eight of the nine high density fuel racks had been assembled at the time the damaged 2 x 2 cell module was dismantled.

An examination of the removed outer stainless steel sheets indicated that the third vertical row of spot welds on each side (see Figure 4-1) was defective. The vast majority of welds in these rows (# 90%) indicated a complete lack of fusion.

The spot welds in each of the other 36 anical welds of the 2 x 2 cell module were found ta be adequately fused.

As described earlier, the outer stainless steel sheet was welded to single rells to

.i form a 2 x 2 module employing automatic welcing equipment which mtained i

four welding guns attached to a movable carriage. The carriagemoved : ang the length of the module in three-inch increments, making four simulta.+,os welds at er.i of 36 locations, for a total of 224 spot welds per side.

An investigation was made to determine the cause of the unfused welds in the l

l nonconforming module. Initially, it was determined that the only obvious change I

to the welding process that could have had an effect on weld quality was the variation in shielding gas composition. The nonconforming module was made in a series of 50 modules,in which 100% Argon was used for the shielding gas. It was conservatively assumed that all 50 modules in this series should be suspected of having an unfused thiro vertical row of spot welds.

j The maximum number of suspect 2 x 2 cell modules in any one rack was 13.

l Consequently, it was concluded that the ability of a fuel rack with 13 nonconforming modules to accomodate the imposed seismic loads must be evaluated.

i FCAM *NES 205 2/80

81A0824 DOCUMENT NO.

DAGE OF NUCLEAR ENEAC' 7 RVICES. INC.

^

y,

-.u.a 7

O-M N '3

.)

i I

L a-_l_

l 6[ v i

1 Unfused Row HF-of Spot Welds JJ (Typ) m' tr -

v-.

.p!- -

n r

t X

't i

- M 14-.

.=E.

.i J

_QUJ V ^-

4 O

1. ~

y

.K

+.,

j q

+.

+

.t

.g

.i

.e

..i

.I

..i

.I

.I

.I

.I

.I

.I

.I

.I

..i 3

a e

i*.

s

. +

I 8

l1 1

1 I

i 8

9 g

• l

'l

• 1 1

I e

i g

i

' i

,6 i

i !

I l

I I

I i

i 1

i 4

i

,t a

a i

a e

i i

I l 0 1

e l

.i

,1

,1

)

1 q

9 o

+

.

• g og q

g q

.i

..i 6

4, 4

. -i NONCONF0FFING 2X2 CELL MODIN.E FIGURE 4.1 FORM

• NES 205 2/80

DOCUMENT NO.

RIA08:4 PAGE II OF 29 NUCLEAR ENERG) SERVICES. INC.

Two evaluations were determined to be necessary:

1.

A structural analysis to evalu " ?,e eff ect that the def ective third row spot welds have on the results of the seisr ' '"tyses previously performed f or the Calver; ' tiffs Unit i fuel racks.

2.

A load test to determine the effect that the def ective spot welds have on the ability of the 2 x 2 cell module to accommodate the local loads applied by the fuel assemblies during a seismic event. A load test was concluded to be necessary since sufficiently accurate local stress calculations could not be perf ormed to make definitive conclusions.

4.2 STRUCTURAL ANALYSIS AND LOAD TEST If the third row of spot welds were unfused in any 2 x 2 cell module, natural frequencies and ultimately the seismic loads could be affected. The first step in the structural evaluation was to determine the effect of the defective third row welds on the module stiffness and, hence, the natural f requencies of the non-conforming rack. The reduced natural f requencies, in turn, established the seismic acceleration values to be applied to the module / rack base configuration.

The maximum cell bending stresses (occurring at the base of the 2 x 2 cell module) and the associated combined stress ratios were calculated and shown to be within acceptable values, in addition, calculations were perf ormed to determine the potential for cell wall buck!. g (at the base of the cells) due to the missing spot welds, and to evaluate the stress levels in the welds bet.veen the 2 x 2 cell walls and the 2 x 2 cell module base plate and between the module base plate and the rack base structure. It was concluded that the overall stress levels in the fuel rack were acceptable for the nonconforming racks when compared with allowable stress values based on the actual (certified) tensile strength values of the cell and base materials.

it was also concluded that sufficiently accurate stress calculations could not be performed to determine the effect that the missing spot welds would have on the ability of the 2 x 2 cell module to accomodate the local loads applied by fuel FCRM = NES 205 2/80

81 A082I.

DOCUMENT NO.

11 m

NUCLEAR ENERGY SERVICES. INC.

assemblies during a seismic event. Consequently, NES recommended that load tests be performed to establish the ability of the nonconf orming 2 x 2 cell module to handle local seismic loads.

Three local stress / loading conditions are of concern because of the missing vertical welds:

1.

The bending stress in the outer cell wall at the top of the cell module 2.

The bending stress in the outer cell wa!! at an intermediate cell elevation corresponding to a spacer elevation for the stored fuel assemblies 3.

The combined shear / tension loads on the cell wall spot welds.

Three basic tests have been devised to evaluate the above local conditions for the Operating Basis Earthquake (OBE) and Design Basis Earthquake (DBE) seismic loadings associated with fuel assembly and" loose pin" storage. For these tests, a 2 x 2 cell module of half-height has been f abricated using standard production spot welds at all locations, except f or.he third vertical row location on each outer sheet, containing no welds. A ft.ll-height module could not be constructed because of the lack of sufficient f ormed pieces.

The module, however, contains all of the principal items comprising a full size module except f or the center spacer assembly. The 2 x 2 module,in turn,is welded to a carbon steel structure that supports the module in a horizontal attitude entirely f rom the module base plate (see Figure 4.2). Steel blocks were distributed inside the storage cells in accordance with the loading requirements specified f cr the three tests. Details of the test plan applied loads, configurations, and acceptance criteria are provided in Ref erence 2.

l The structural analysis details and load test results ar presented in Ref erence 3.

Based on the results of the structural analysis and local loading effects evaluations, the following conclusions have been made.

1 1

I l

FCRM 8 NES 2C5 2/80 L

81A0824 DOCUMENT NO.

my 11 13 PAGE op 2 9,,,,,,

NUCLEAR ENERGY SERVICES. INC.

A f -Q T 1

j.

+

t t

I-M l

l I

i l

me i

!d l

3 e

w

.=

U 1

cc m

I m

e cn in m

l l

W-22 3

C 4

z 4

1 4

+

+

+

E

+

4 s

f

+

9

+

4 Y

7 y

\\

N 4

N 4 ? t t 4 + +

+e 4

\\

C

[

m O

m w

3 s +..

4 l

TEST C0hTICURATION FIGUPE 4.2 FORM *NES 205 2/80

-~

w -

• MMw

DOCUMENT NO.

81A0824 my PAGE 14 OF 29 NUCLEAR ENERGY SERVICES. INC.

l__

1.

The results of the structural analysis indicated that the stresses in the non-conforming storage rack structure, resulting f rom the loadings associated with the normal and abnormal conditions, are within the a!!owable stress limits f or the actual material used in the rack fabrication. Consequently, the nonconf orming rack is structurally adequate to meet the requirements for Seismic Category I structures.

2.

The results of the nonconforming 2 x 2 module local loading effects testing indicated that there will not be any permanent deformation of the 2 x 2 module under OBE loading conditions.

3.

The results of the nonconforming 2 x 2 module local loading ef fects testing indicated that the small permanent deformation in some local areas of the 2 x 2 module resulting f rom DBE loading conditions will not adversely affect the structural and functionalintegrity of the storage rack.

4.

The Calvert Cliffs Unit I nt onforming high density fuel storage rack is structutally adequate to perform its function (the storage of fuel in a safe geometry) during and af ter all anticipated loading conditions.

l 4.3 ULTRASONIC TESTING In order to determine the extent of unfused welds in other 2 x 2 cell modules, NES tested spot welds in the second and third rows of modules still available at the f abricator.

The determination of weld fusion was made by ultrasonic methods. The use of ultrasonics in this type of application is not new; it has l

l been used f or years to verify fusion in spot welds. Rows I and 4 were not i

ultrasonically tested because they are readily inspected visually or by feeler gauge. Test samples of fusM and unfused spot welds were used as standards to verif y the testing technique. This type of ultrasonic testing is self-checking in i

that a fused weld is represented as a double thickness of base material. As the 1

ultrasord,- transducer approaches a spot weld, the indication is one thickness of base material. When the transducer is directly over a fused weld, the indication i

l 1

l t

FC=N = NES 2CS 2/80

DOCUMENT NO.

9 m 824 pt.G E c5 NUCLEAR ENERGY SERVICES, INC.

becomes approximately twice the thickness of base material. The examiner, therefore, sees an indication change for every fused spot weld and no change in indication for an unfused spot weld.

Spot welds in rows 2 and 3 of 17 modules were examined, in accordance with a sampling plan.

Approximately 293 spot welds were tested, with all welds indicating adequate fusion. One module was available for testing from the series of 50 suspect modules. This module had been assembled into a fuel rack with only one side available for testing. In addition to a random sample of row 2,32 of the 56 welds in row 3 on this module were tested, with positive indication of fusion on each weld. The test sampling plan and test results are provided in Reference 4 4.4 SHIELDING GAS COMPOSITION TEST The fabricator had shown that, when test sample spot welds were made for pull tests, the addition of 2% Helium to the Argon shielding gas had little or no effect. Spot welds of adequate strength were repeatedly made with either gas composition.

In order to verify that the automatic welding process of making four simultaneous spot welds in a production-type setup would also show a lack of sensitivity to the addition of 2% Helium, the fabricator ran the following test. A test module was made using two full scale single cells assembled into a 1 x 2 cell I

module. The process of welding the outer sheet, covering the two side-by-side cells, was identical to that used in making a 2 x 2 module. Telds were made using 100% Argon for one half of the test module length and 2% Helium was added to the Argon for the remaining half of the test module. The outer sheets were then removed by pulling them with the overhead crane until they tore apart at the spot welds. Investigation showed that all spot welds were completely I

fused and no significant difference was observed in the welds made with different gas mixtures.

1 FCRM = NES 205 2/80

81A0824 DOCUMENT NO.

PAGE OF NUCLEAR ENERGY SERVICES, INC.

4.5 FREQUENCY OF OCCURRENCE It was originally assumed, for conservative reasons, that 50 noncenforming 2 x 2 cell modules could have been produced.

Structural analysis and load tests demonstrated that even if 30 nonconforming modules existed, the Calvert Cliffs fuel racks are structurally adequate to store loose pin fue; during and af ter all anticipated loading conditions.

We are now confident that the previous assurnption of 50 nonconforming modules was grossly conservative. Based on subsequent tests and evaluations, a more realistic assumption is that no more than three nonconforming modules were made.

The bases for this revised estimate are as follows:

1.

The shielding gas variation does not significantly affect weld quality; therefore, the original basis for selecting 50 modules is no longer valid.

2.

Ultrasonic tests on 17 other modules showed not a single unfused weld of l

those tested in rows 2 and 3 (one of these modules came from the original suspect group of 50).

l 3.

Test samples made for pul testing during production consistently showed l

adequate fusion. A minimum of 2 test samples per welding gun were made l

dady.

l l

4 Rows of unfused weids were never found in rows 1 and 4. The latter rows were inspected af ter welding by visual and feeler guage methods. Since all four welding guns are essentially identical and were set up and used at the same time, it is highly improbable for them to behave differently.

5.

The most probable cause of an unfused row of welds occurring is that the particular tungsten electrode became contaminated. This has to be an extremely rare occurrence since the fabricator checks each electrode prior FCAM

• NES 205 2/80

DOCUMENT NO.

I PAGE OF NUCLEAR ENERGY SERVICES. INC.

to a production shif t and verifies its onerability by making test welds.

Since only a maximum of three 2 x 2.odules were ever made prior to rechecking the electrode and making more pull test samples, that is the maximum number of modules that could have been nonconforming. An electrode contaminated to the point where it cannot produce satisf actory welds is an extremely unlikely event; this is verified by the f act that, for this fuel rack design, approximately 5,000 vertical rows of spot welds were made - which were all inspectable af ter welding (e.g., rows 1 and 4 on the 2 x 2 cell module)and not a single case of an unfused row of welds has been detected.

t

,s FORM z NES 205 2/80

DOCUMENT NO.

81A082'.

mar OE M

9 NUCLEAR ENERGY SERVICES. INC.

5. RANDOM UNFUSED WELDS Although the probability of a complete unfuseo inw of welds is extremely low, and virtually impossible in an inspectable seam, a small nun ber of randomly unfused spot welds can be expected to occur because of local irregularities around a potential weld site. When such a random unfused weld is detected during in-process inspection, the unfused weld is repaired by a manual spot weld. Although records of these unfused welds are not kept by the f abricator, the latter estimates that approximately 14 unfused welds per 2 x 2 cell module were detected and repaired.

If this type of random unfused weld is not detected until final rack inspection, NCR documentation is processed and then the welds are similarly repaired. A review of NCR's by the f abricator reveal a total of 33 unfused welds have been discovered during final rack inspection.

It would not be realistic to expect that the fuel racks are completely f ree of this type of random weld f ailure but the percentage is extremely low In order to put these s

f allure numbers in perspective, it must be pointed out that a 10 x 10 fuel rack contains approximately 56,000 spot welds along its vertical seams in the 2 x.2 cell modules.

The complete pool array of 330 storage locations contains approximately 465,000 spot welds.

j l

I i

l t

I t

l i

FCRM :NES 205 2/30

DOCUMENT NO.

.91 A083 PAGE 19 OF 29 NUCLEAR ENERGY SERVICES. INC.

6. UNFUSED WELD SENSITIVITY ANALYSIS There is one case of a 2 x 2 cell module with four vertical rows of 90% unfused welds, and there is evidence that a small number (less than lo6) of random unfused welds have occurred. Based on the above, it was decided to perform an unfused weld sensitivity analysis to determine what eff ect various combinations of unfused welds could have on the structuralintegrity of the Calvert Cliffs fuel racks.

6.1 METHOD OF EVALUATION The first step in the sensitivity analysis was to determine the eff ect of the unf used welds on the module stiffness (moment of inertia) and, hence, the natural f requency of the module and the rack. The reduced natural frequencies, in turn, established the seismic acceleration values to be applied to the module / rack base configuration.

The natural frequency of the module with defective weld configuration was calculated by pro-rating the f requency of the nonconf orming cell module with the square root of the ratio for the respective moment of inertia. If the effect on the natural frequency of the module is small, the effect on the natural frequency of the storage rack will be.still smaller. If the seismic acceleration values for the defective weld configuration and nonconforming module are the same, the seismic anertaa loads will be the same.

The maximum bending stresses in the storage cell wall, with and without considering the locas buckling effects, can be calculated by pro-rating the results of the nonconf orming module analysis.

The structural analysis has been performed for both the normal fuel storage condition and the loose pin storage condition. Analysis has been perf ormed for the DBE event only since this seismic loading condition was found to be more critical than that of the OBE loading condition.3 i

The structural evaluation has been perf ormed by comparing the results of the 4

structural analysis for the defective weld configurations to that of the nonconforming module and to the allowable stress values. Results of the local i

1 FORM = NES 205 2/80

DOCUMENT NO. M AO S M 11 mr NUCLEAR ENERGY SERVICES. INC.

Ioading ef f ects testing f or the nonconf orming module were also consicered in this evaluation. Details of the analysis are presented in Reference 3 and the results are summarized in the f ollowing paragraphs.

A.2 RESULTS CF EVALUATION It is possible to propose a large number of combinations and permutations of unfused welds; this analysis was theref ore limited to five representative cases, which include not only what has been observed during f abrication but also unfused weld combinations so extensive they could not exist undetected.

Case 1 Four vertical rows of seam welds around the periphery of the 2 x 2 cell module are unfused. Two variations (I A and IB) with different locatians of the unfused welds were considered, as shown in Figure 6.1.

This case duplicates the nonconf orming module discussed in Section 4 Some minor deformation (separation of the outer stainless steel sheet approximately 1/16 to 1/3 inch) could occur uncer DBE conditions with loose pin storage.

These minor def ormations will not adversely aff ect the structural and functionalintegrity of the storage rack.

Case 2 Four vertical rows of seam welds near the center of the 2 x 2 cell module are unfused.

Two variations (2A and 28) with diff erent locations of unfused welds were considered, as shown in Figure 6.2.

Unfused seam welds near the center of the 2 x 2 cell module have an insignificant effect on stiffness and load carrying capacity of the module. Case 2 is stronger than Case 1, and no def ormation is expected under DBE conditions with loose pin storage.

Note: Cases I and 2 cover the eight vertical rows of seam relds in locations of the 2 x 2 cell module that are not accessible for inspection by visual or feeler guage methods af ter welding. The I

remaining 12 vertical rows are completely inspected after welding and it is not plausible that a significant number of unf used welds in these rows could exist undetected.

i i

FC M = NES 2C5 2/80

~

81,10824 DOCUMENT NO.

'9 21 PAGE OF NUCLEAR ENERGY SERVICES. INC.

n Y

D l

[

(

.e.

l 1

'l I

L b

v CASE lA h

f 0

n

-'i

[e 1

l Xnfused Row of Spot l

)

Velds (Typ) db" 4,ff

%:{i '

y!)

T O

h f

CASE 13 LT TSED WELD CONFIGURATION - CASE i FIGURE 6.1 l

l l

l FORM # NES 205 2/SO

I.

81't0824 DOCUMENT NO, my 11 2'

'o PAGE OF g

NUCLEAR ENERGY SERVICES. INC.

~

i

\\

\\

I i

l i

I i

[

)

.r-I l

I i

~b*6 CASE 2A i

.[

T[.

a

,O

/,

s I

v l

4 i

l Unfused Row of

[

Spot Welds (Typ) h n

v i

I

'~

_i CASE 2B UNFUSED WELD CONFIGURATION - CASE 2 FIGURE 6.2 FORM sNES 205 2/80

..m 81A0824 DOCUMENT NO.

23 PAGE OF NUCLEAR ENERGY SERVICES. INC.

Case 3 This case represents a clearly unrealistic situation, but was analyzed to demonstrate the extent of unfused rows of welds that can be tolerated. In this case, f our vertical rows of seam welds around ths periphery, three rows of seam welds in the interior cell walls near the peripnery, and two rows of seam welds near the center of the module were unfused. This represents a loss of nine rows of seam welds per 2 x 2 cell trc'ule. Three variations (3A, 3B, and 3C) with diff erent locations of unfused welds were considered, as shown in Figure 6.3.

Under DBE conditions with loose pin storage, some of the cell walls could be separated f rom adjacent walls and some minor deformation is expected but the cell inside dimension will not decrease and the cell pitch will not change. The structural and functional integrity of the fuel rack will not be adversely affected.

Case 4 This case assumes that 25% of all spot welds in the vertical seams of a 2 x 2 cell module are unfused in a miformly distributed pattern, as shown in Figure 6.4 Twenty-five percer.i

he spot welds in a 2 x 2 cell module e

represents 553 unfused welds (this represents 14,000 unfused welds in a 10 x 10 rack). There could be minor deformation under DBE conditions with loose pin storage, however, the structural and functional integrity of the storage rack is unaffected.

Case 5 This case represents a combination of four vertical rows of seam welds around the periphery of a 2 x 2 cell module to be unfused (Case !), plus 10% of the remaining spot welds to be unfused in a uniformly distributed pattern. Two variations (5A and 5B) with different locations of the unfused rows were considered, as shown in Figure 6.5.

Some local deformations could occur under DBE conditions with loose pin storage.

This local deformation will not adversely aff ect the structural or fur'ctional integrity of the storage rack.

eCCM s NES 2C5 2/80

DOCUMENT NO.

1'1 PAGE OF NUCLEAR ENERGY SERVICES, INC.

m n

'f.

I j

l l

l l

l y,

9 C.

J v

--R-M'

~ }%

• I

?f':

?

w.

m llv l.

J n

y.

qt q f-tj) if{

T l

i l

I l l C

~_b M.

b

_ __2 v

v CASE 3A CASE 3B i

D

\\

T

• r 3

l l

l t.sh l

I

' s v ',

U Unfused Row of

,g Spot Welds (Typ)

. J, '..

4 p l

l f

c.-

9 2r -

~W/

CASE 3C UNFUSED WELD CONFIGURATION - CASE 3 FIGURE 6.3 FCAM 2 NES 205 2/80

OIA00?'

DOCUMENT NO.

au FAGE OF NUCLEAR ENERGY SERVICES. INC.

A e

o

^

f

(

l f

I i,

I k

x r

x n

N l

I 1

l 4

I I

h Unfused Spot Welds 25%

y Uniform Distri-4-

bution (Typ) h D

a a

a a

I ll l

l Il l

I Il I

I il I

J l

l I

[

l 8

bI.2 f.)

bb d h.) bI !

I

.s.

. 6e..

e./. E.

e._5 n

A s.

+

i 2

a -.,

y s.

e.@

62, s.6

. G.e l

a 2

3 G.

v.

G.$. 6

6..

6*

g *,

1 1

g

2.. s..

.. G y.

a m

l UNFUSED WED CONFIGURATION - CASE 4 FIGURE 6.4 l

l I

l l

FCRM # NES 205 2/80

~

81A0824 DOCUMENT NO.

26 29 PAGE gp NUCLEAR ENERGY SERVICES, INC.

m a

n m

_y- _- _

'h..

b,,

s i

t t

l l l ll l

(J o

^

x x

-c W

'h

[

l il i

I i

I i

J A

Q y

v v

b j

s u 6 CASE 5A

~

m n

n n

%,[

g A. ' _.

4 N".

l p

Unfused Rev o f i

Spot Welds (Typ)

It

_d 6

)

se Spot Welds.

Distribution (Typ)l 10% Uniform l

l i

l 1

d I

y v-v CASE SB UNFUSED WELD CONFIGURATION - CASE 5 FIGURE 6.5 FCAM 8 NES 2CS 2180

DOCUMENT NO.

81A0824 m

GE 27 05,,29 NUCLEAR ENERGY SERVICES. INC.

6.3 CONCLUSION

S The results of this structural analysis demonstrated that the stresses in the fuel storage rack structure, with the unfused weld configurations (Cases I through 5) resulting f rom normal and abnormal loading conditions, are within the allowable stress limits for the actual material used in the rack fabrication.

These nonconf orming storage rack configurations are capable of storing fuel in a saf e geometry during and af ter all anticipated loading conditions.

Consequently, these analy::ed racks are structurally adequate to meet the requirements for Seismic Category I st uctures.

t l

l l

l l

l i

l l

l FC:M = NE5 205 2/S0 l

81A0821.

DOCUMENT NO.

29 st.G E Oc NUCLEAR ENERGY SERVICES. INC.

7. W".l.D PARAMETER VARIATION TEST Details of the weld parameter variation test will be supplied at a later date. It is anticipated that the test will be completed in January,1981. The following steps must be accomplished:

1.

NES to develop the test procedure and schedule 2.

BG&E to approve 3.

MPC to perform testing 3

NES and MPC to evaluate test results 5

NES to prepare the test report.

cCa'.5 = NES 205 2. 80

~

DOCUMENT NO.

PAGE OF NUCLEAR ENERGY SERVICES. INC.

3. REFERENCES 1.

" Process Procedure Specification QC-PPS-254-1 f or Arc Spot Welding (GTAW)of Staialess Steel," Metal Products Corporation (March 11, 1930) 2.

" Test Plan for Nonconforming 2 x 2 Storage Cell Modules for the Calvert Cliffs Unit 1 Nuclear Plant High Density Spent Fuel Storage Racks," Nuclear Energy Services, Inc., Document No. 30 A3652, Revision 1 (October 22, 1930) 3.

" Evaluation Report for Nonconforming 2 x 2 Storage Cell Module for the Calvert Cliffs Unit 1 Nuclear Plant High Density Spent Fuel Storage Racks," Nuclear Energy Services, Inc., Document No. 31 A0569, Revision 2 (November 11, 1980) 4

" Ultrasonic Testing of Module Spot Welds in the Calvert Cliffs Unit 1 Spent Fuel Racks," Nuclear Energy Services, Inc., (October 22, 1980) 5.

" Sensitivity Analysis of Defective Spot Welds in the Calvert Cliffs Unit 1 Nuclear Plant High Density Spent Fuct Storage Racks", Document No. SI A0570, Revision 0 (January 12, 1980) l l

FC:M :NES 205 2/80