ML20071M314

Forwards Response to IE Bulletin 80-11, Masonry Block Walls, Per 820122 Request

ML20071M314
Person / Time
Site:	Three Mile Island
Issue date:	09/20/1982
From:	Hukill H GENERAL PUBLIC UTILITIES CORP.
To:	Stolz J Office of Nuclear Reactor Regulation
Shared Package
ML20071M318	List: ML20071M314 ML20071M320
References
REF-SSINS-6820 5211-82-224, IEB-80-11, IEB-80-1124, NUDOCS 8209270063
Download: ML20071M314 (55)

Text

GPU Nuclear Corporation G U luclear m'aneoss48 Middletown. Pennsylvania 17057 717 944-76?1 TELEX 84 2386 Writer"s Direct Dial Number:

Septanber 20, 1982 5211-82-224 John F. Stolz, Chief Operating Reactors Branch No. 4 Division of Licensing Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Mr. Stolz:

Three Mile Island Nuclear Station, Unit 1 (TMI-1)

Operating License No. DPR-50 Docket No. 50-289 IE Bulletin 80-11 Enclosed is our response to your letter of January 22, 1982 concerning our report on IE Bulletin 80-11, " Masonry Block Walls". Subsequent to the sub-mittal of our original topical report, there have been changes in both our analytical approach and resultant modification in specific cases. Topical Report No. 001 has been revised to incorporate these changes and is also enclosed.

The NRC questions (letter of January 22, 1982) which have been affected by these channes are numbers 3, 5, 8, 10 and 14. With this submittal, we believe the concerns raised in Bulletin 80-11 have been satisfactorily addressed.

Sincerely, mI I. D.

hIcki 1

• Director, TMI-1 IIDil:CJS:vj f Attachment / l cc: R. Jacobs, with Attachments 001 R. Conte, with Attachments 8209270063 820920 PDR ADOCK 05000289 PDR O

I GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporat:on

e. j e.

THREE MILE ISLAND NUCLEAR STATION. UNIT 1 REEVALUATION OF MASONRY WALLS

.=

RESPONSE

TO NUCLEAR REGULATORY COMMISSION REVIEW O

PREPARED BY:

COMPUTECH ENGINEERING SERVICES. INC.

2855 TELEGRAPH AVENUE BERKELEY. CA 94705 -

REVIEWED BY: #M py) f  % 8.3C-82 T.H. CHANG / L GARISIAN DATE ENGINEERING MECHANICS GP NUCLE R C.ORP.

h APPROVALS: - - - -

r- u -r.

ROCHINO DATE MANAGER. ENGINEERING MECHANICS l

G UCLE R CORP.

8.10-&

D.K. CRONEBE ER DATE

~

DIRECTOR - ENGINEERING & DESIGN GPU NUCLEAR CORP.

I k

@ t I

l TABLE OF CONTENTS I SECTION NO. PAGE NO.

1 INTRODUCTION ................................. 1 2 ITEM 1 - MULTI-MODE EFFECTS ....................... 2 2.1 Method of Analysis. ............................ 2 2.2 Description of the Walls. ......................... 3 2.3 Re s u l t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.4 Conclusions. ................................ 4 3 ITEM 2 - THERMAL EFFECTS AND WIND LOADS . . . . . . . . . . . . . . .5 4 ITEM 3 - SINGLE WYTHE VS. MULTl WYTHE WALLS . . . . . . . . . . . . .6 5 ITEM 4 - EQUIPMENT LOADS ......................... 7 5.1 Method of Analysis. ............................ 7 5.2 Evaluation of the Method of Analysis. . . . . . . . . . . . . . . . . . . .7 5.3 Conclusions. ................................ 8 6 ITEM 5 - VERTICAL MOTION .......................... 9 7 ITEM 6 INTERSTORY DRIFT AND PIPING LOADS . . . . . . . . . . . . . . . 10 7.1 In-Plane Effects of interstory Drift . . . . . . . . . . . . . . . . . . . . 10 7.2 Out-of-Plane Effects of interstory Drift . . . . . . . . . . . . . . . . . . 10 7.3 Piping and Equipment Loads ......................11

8 ITEMS 7 and 13 - MODULUS OF ELASTICITY . . . . . . . . . . . . . . . . . 13 l

9 ITEM 8 - IN-PLANE EFFECTS ........................15

10 ITEM 9 - BOUNDARY CONDITIONS ......................16 11 ITEM 10 - STEEL BRACKETS FOR REINFORCING . . . . . . . . . . . . . . . 18 o

9 (i) l

. a SECTION NO. PAGE NO.

12 ITEMS 11 and 15 - IN-PLANE SHEAR STRESSES AND SHEAR STRAING . . 19 12.1 Overview of Test Program .......................19 12.1.1 Applicability of Test Results . . . . . . . . . . . . . . . . . . 20 12.2 Evaluation of Test Data ........................22 12.2.1 Shear Stresses and Strains for OBE and SSE Events . . . . 22 12.2.2 Statistical Analysis of the Stress Data . . . . . . . . . . . . . 22 12.2.3 Discussion of Stress Results . . . . . . . . . . . . . . . . . . 28 12.2.4 Statistical Analysis of the Strain Data . . . . . . . . . . . . . 28 12.2.5 Discussion of Strain Results . . . . . . . . . . . . . . . . . . 31 12.3 C o n clu s io n s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 12.4 References. ...............................32 13 ITEM 12 - ENERGY BALANCE TECHNIQUE . . . . . . . . . . . . . . . . . . . 38 14 ITEM 14 - COLLAR JOINT STRENGTH . . . . . . . . . . . . . . . . . . . . . 39 14.1 Purpose of Collar Joints ........................39 14.2 Collar Joint Tests and Allowable Stresses . . . . . . . . . . . . . . . 39 14.2.1 Reference ...........................40 15 ITEM 16 - SHEAR - UNREINFORCED MASONRY . . . . . . . . . . . . . . . . 41 16 ITEM 17 - TENSION NORMAL TO BED JOINT . . . . . . . . . . . . . . . . . 43 16.1 Overview of Test Programs . . . . . . . . . . . . . . . . . . . . . . . 43 16.2 Applicablilty of Test Results. ......................43 16.3 Evaluation of Test Results .......................44 16.4 Discussion of Results . . . . . . . . . . . . . . . . . . . . . . . . . . 46 16.5 Conclusions ...............................47 16.6 References. ...............................47 17 ITE M 18 - BON D STRESS . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 17.1 Reference ................................49 18 ITEM 19 - WALL AB-14 ............................50 i

19 ITEM 20 - SCHEDULE FOR MODIFICATIONS . . . . . . . . . . . . . . . . . . 51

- 9

- t t

(11)

a g l

}

1 INTRODUCTION l

The Nuclear Regulatory Commission (NRC) review of the criteria and reevaluation I of masonry walls at the Three Mlle Island Nuclear Station. Unit 1 resulted in I a request for additional information on a number of items. The request was submitted in a letter dated January 22, 1962.

In the following sections responses to the requested items are presented.

l b

1

r .

2 ITEM 1 - MULTI-MODE EFFECTS QUESTION ltem 1 of the NRC request for additional Information states:

• The Licensee used a static analysis based on multiplying the mass of the structure by the peak ampilfled response spectra (ARS) Instead of a dynamic selsmic analysis. To account for possible multi-mode effects. an ampilfication factor of 1.05 was used to obtain this equivalent static load. The Standard Review Plan (SRP) (91 accepts an equivalen' static load method if the system is shown to oe realistically represented by a simple model aM the method gives conservative results. Furthermore, it suggests that a factor of 1.5 be applied to the peak ARS of the applicable floor response spectra, in view of this, it is suggested that the Licensee provide information to justify use of an amplification factor of 1.05."

RESPONSE

The analysis method used by GPU was not a purely ' static

• method of analysis, but was a combination of a dynamic and static method of analysis. In the following sections the method of analysis used by GPU will be described and results obtained using this method will be compared with the results obtained using the multi-mode response spectrum method.

2.1 Method of Analysis.

The analysis method utilized by GPU consisted of the following steps:

1

1. A finite element model of the wall was constructed and the fundamental or first mode frequency of the wall was obtained. The finite element model included appropriate boundary conditions. openings and the mass of all attached equipment.

2. Using the fundamental frequency and the response spectra of the floor above the wall. the spectral acceleration at the fundamental frequency of the wall

& was obtained.

~

3. The finite element model of the masonry wall was then subjected to a uniform static load equal to 1.05 times the wall mass times U e spectral acceleration obtained

' in Step 2.

f 2

This method is clearly a combination of a dynamic and static method of analysis since the fundamental frequancy of the wall was obtained from a reasonably accurate finite element model. The results obtained using this method of analysis are compared with the results obtained using the multi-mode response spectrum method for three walls. The computer program SAPSA was used to perform a finite element multi-mode response spectrum method of analysis. The frequencies considered in the dynamic analysis of each wall were the five lowest frequencies less than 100 Hz and the stresses obtained from these five modes were combined using the SRSS method of combination.

2.2 Description of the Walls.

The three walls, that were selected for the analysis. represent typical safety related walls found at the Three Mile Island Nuclear Station. Unit 1.

Wall AB-1 is 168* long. 249' high and 36' thick. It has a small opening at the top of the central portion of the wall. The top edge is considered free but the three other edges are simply supported.

Wall AB-6 is 88.$* long. 93* high and 32' thick. The bottom edge. which rests on the floor, is assumed fixed whereas the other three edges are free.

Wall AB-13 ls 116* long. 248.5* high and 24' thick. It has a door opening and one penetration on each side. The top edge is free, but the other three edges are considered simply supported.

2.3 Flesults

} A summary of the results of the analyses are given in Table 2-1. The table

' compares the maximum stress ratlos for each wall using the two different methods of analysis.

From tha results presented in Table 2-1. It is apparent that the results from the two methods of analysis are very close with a maximum difference of approximately 10%. For two of the walls. AB-6 and AB-13. GPU's method of analysis is conservative. For Wall AB-1 GPU's results are not conservative. but are within approximately 10% of the results obtained using the multi-mode response spectrum method.

h i

4 e

S

TABLE 2-1 Wall No. Basis GUMa)x Ma)y AB-1 Multi-Mode 0.2848 0.0927 GPU 0.2642 0.0805 AB-6 Multi-Mode 0.0353 0.3533 GPU 0.0399 0.3994 AB-13 Multi-Mode 0.1917 0.1187 GPU 0.1935 0.1234 where M is the maximum moment calculated in the wall Ma is the maximum allowable moment equal to c a*S-c'a is the c '.teria specified allowable stress S is the section modulus x denotes a moment on a horizontal strip y denotes a moment on a vertical strip 2.4 Conclusions.

Three typical walls were selected to demonstrate the adequacy of the method of analysis used by GPU in the evaluation of masonry walls. The results obtained from GPU's method of analysis were compared with those obtained using the multi-mode response spectrum method of analysis. For all three walls the results obtained were within approximately 10% and for two of the three walls the results using GPU's method of analysis were conservative.

1

. 9

}

3 ITEM 2 - THEFH4AL EFFECTS AND WIND LOADS QUESTION ltem 2 of the NRC request for additional Information states:

" Appendix 7.2 of Reference 3 provided a summary of stresses, in which load combination included dead load pipe reactions and seismic loads. The thermal effect and wind load were not accounted for, in Reference 2. the Licensee indicated that the thermal effect is insignificant, but did not justify this conclusion. The Licensee should elaborate on this point and also Indicate if wind load was included in the analysis."

RESPONSE

The masonry walls at the Three Mlle island Nuclear Station. Unit 1. are all interior walls and are therefore not subjected to any wind loads. Furthermore, no significant temperature differential was found to exist across any of the walls. Consequently, thermal effects and wind loads were not included in the analysis of the masonry walls at the Three Mlle Island Nuclear Station. Unit 1.

l l

l

=

i 5

l

6 4 ITEM S - SINGLE WYTHE VS. MULTI WYTHE WALLS QUESTION ltem 3 of the NRC request for additional Information states:

'In Reference 2. the Licensee used the single wythe assumption (for out-of-plane loading) for multiple wythe walls. The Licensee should demonstrate that this assumption wiil result in a conservative evaluation.'

RESPONSE

All the masonry walls that were affected by 1.E.Bulletin 80-11. except two (RB-1 and TB-1), are multi-wythe and constructed from solid concrete block masonry units.

Although the reevaluation cr!ieria stated that multi-wythe walls would be analyzed as single-wythe walls, th!4 option was not exercised for the multi-wythe walls at Three Mlle island. Unit 1. As part of the repair procedures for the masonry walls, all the collar joints of all multi-wythe walls were filled with a non-shrink grout called Five Star Grout No.100 from the U.S. Grout Corporation. The grout had a specified compressive strength of 5000 psl at 7 days. Both shear and tensile tests were performed on collar joint test samples and these tests are described in detall In Section 14 of this report. Thus all multi-wythe walls were assumed to have the properties of multi-wythe construction in the analysis.

Therefore the question of the conservatism of the single-wythe assumption is not applicable to the Three Mlle Island Nuclear Station. No.1.

E 1

4 h 6

5 ITEM 4 - EQUIPMENT LDADS QUESTION ltem 4 of the NRC request for additional Information states:

'With regard to stresses resulting from equipment.

a static analysis was used by tr:ultiplying the weight of the equipment by the peak acceleration of the response spectrum of the corresponding floor. The Licensee should clarify whether a multiplication factor was used to obtain the equivalent static load or. If not. explain why.'

RESPONSE

The subsections that follow attempt to clarify the analysis method used and show it to be sufficiently conservative.

5.1 Method of Analysis.

In evaluating the dynamic properties of each masonry wall a finite element grid was constructed. The wall propertles, including the wall thickness, its density, elastic modulus and additional point masses from equipment attached to the walls, were submitted as parameters into the analysis. Using this finite element model the fundamental frequency of each wall was calculated and included the effect of equipment masses attached to the wall.

When evaluating the inertia forces generated in each wall. the wall and

• - the equipment masses were treated separately.

The inertia forces generated in a wall from its own dynamic response were s

calculated by assuming a uniform load. equal to the wall's mass tirr.es the spectral acceleration at the wall's fundamental frequency, multiplied by 1.05 to account for higher mode effects, and distributed over the surface of the wall. This method of analysis is discussed in detall in Section 2.

The forces generated in the wall from the attached equipment were evaluated by mulilplying the equif, ment masses by 1.05 times the peak acceleration of the floor spectrum above the wall. The equivalent static loads resulting from the wall inertla and the equipment were simultaneously applied to The direction of the applied forces the finite element model of the wall.

were selected such that the resulting stresses in the wall would be a maximum for the two sets of loads.

5.2 Evaluation of the Method of Analysis.

. e The method of analysis described in the previous subsection is considered appropriate and conservative for the following reasons.

7

t (a) The walls that did have equipment attached to them were practically rigid. The lowest fundamental .

frequency of s!! the walls was 24.0 Hz. for walls AB-8 and AB-9. This frequency is practically in the ZPA ,

range of the applicable floor spectra, l.a. negligible dynamic tmpitfication occurs in the wall.

(b) The equivalent static load resulting from the equipment was calculated by multiplying 1.05 times the equipment mass by the peak of the floor spectra above the wall. For a rigid wall using this floor spectra is also conservative.

(c) The majority of the equipment mounted on the walls l la rigid. The ratio of peak spot, tral acceleration to the ZPA ranges from about 3 to 5 and thus the static load for the majority of the equipment is very conservatively estimated.

(d) The stress in the walls resulting from equipment loads was less than 10% of the total stress for all but two of the walls in Wall AS-7 the contribution of the equipment was 20% of total stress and in Wall AB-3 Upper. It was 30%.

5.3 Conclusions.

With regard to stresses resulting from equipment, a static analysis was used by multiplying the weight of the equipment by 1.05 times the peak acceleration

. of the response spectrum from the floor above each wall. The stresses thus calculated are low when compared to the stresses resulting from the dynamic response of the walls. In addition, since the majority of the equipment attached to the walls is essentially rigid, the method used to calculate the equipment loads is quite conservative.

In view of the above it is concluded that stresses resulting from equipment are adequately evaluated.

a $

8

I I

l 8 TTEM 5 - VERTICAL MOTION  ;

QUESTION Item 5 of the NRC request for additional information states:

• With regard to the seismic analysis, the Licensee indicated that the vertical component of the motion was not included in the analysis because the positive effect of the dead load on bed jolnt stresses was not included in the evaluation criteria. Although the positive effect is not easily determined. It is suggested that the Licensee attempt to identify this positive effect (from test data or literature) and include It in the analysis or neglect it altogether, which will result in a conservative evaluation. However, it does not seem justifiable to neglect the vertical component of the motion."

RESPONSE

The question posed in this item by the NRO arises from a statement made in the commentary to the reevaluation criteria. The procedure that wat; used by GPU for incorporating vertical acceleration in the reevaluation of the masonry walls was in accordance with the above statements by the NRC. The procedure discussed in the commentary was not used.

The procedure used by GPU for incorporating the effects of vertical acceleration were as folicws:

1. The masonry walls were considered to be rigid (frequency

> 40Hz) in their vertical plane. Therefore, the vertical acceleration for the walls was the vertical ZPA at the appropriate floor level.

2. The dead load of the wall acting on a particular bed joint was calculated by multiplying the appropriate mass of the wall by the acceleration of gravity minus the vertical ZPA acceleration of the floor level above the wall.

3. The compressive dead load on a particular bed joint determined in Step 2 was used, together with the stresses resulting from out-of-plane loads. In calculating the tension stresses normal to the bed [oint in the analysis of the masonry walls.

In summary. the effect of the vertical acceleration was included in the analy' sis of the masonry walls as recommended by the NRC and the question posed in this item is not therefore applicable to the Three Mlle Island Unit 1 Nuc. lear Plant.

7 ITEM 6 -1NTERSTORY DRIFT AND PIPING LOADS QUESTION ltem 6 of the NRC request for additional Information states:

'The Ucensee should describe how the Interstory drift (both In-plane and out-of-plane) during a selsmic event and the loads from piping and/or equipment attached to the masonry walls were accounted for.

Both the local and global effects of piping and/or equipment attached to the masonry walls should be described and examples of the analysis provided."

RESPONSE

The subsections below contain the response to each of the above items.

7.1 In-Plane Effects of interstory Drtft The In-plane interstory drift of the two buildings. Auxillary Building and Reactor Building. Is small. For the Auxillary Building the drift of the floor level where the walls are located is 0.000066 in/in and 0.000132 in/in for OBE and SSE events respectively. For the Reactor Building the largest interstory drift is between floor levels 309' and 321' and is 0.000076 in/In and 0.000152 in/in for OBE and SSE events respectively.

The " Criteria for the Reevaluation of Concrete Masonry Walls - Three Mlle Island Nuclear Station Unit 1* specifies the allowable in-plane story drift as 0.0001 In/in and 0.000167 in/in for unconfined walls for OBE and SSE events respectively and 0.0008 in/In and 0.00134 In/in for confined walls for OBE and SSE events respectively. These allowables are discussed in more detall in Section 12.

A wall is considered confined If. as a minimum. It is bounded on the top and bottom or bounded on three sides. Otherwise the wall is considered unconfined.

The calculated interstory drift given above is in all cases less than the allowable drift for unconfined walls. Interstory drift is therefore not considered

• critical".

7.2 Out-of-Plane Effects of Interstory Drift The impact of interstory drift on the out-of-plane response of the mas,onry walls is dependent on the assumed boundary conditions of the walls. If i

i l

10

the boundary conditions are assumed pinned on all sides, then there will i be no additional stresses in the walls resulting from out-of-plane Interstory drift effects, if the boundary conditions are assumed to be fixed on any of the horizontal bounderles. then the walls will have additional stresses resulting from the out-of-plane interstory drift effects.

A detalled dicussion of the assumed boundary conditions of the wc.ls is t

given in Section 10. For the walls with only one vertical boundary the base is assumed to be flued and the vertical boundary is free. As a consequenco, the wall will be stressed due to out-of-plane drift effects. Table 7-1 below lists the walls affected and Olves the tension normal to the bed joint caused

by the drift and the total tension from all loads, including the drift.

As may be seen from Table 7-1 the stresses caused by the interstory drift range from the smallest values of 6.5 psi and 12.9 psi for wall AB-7 for OBE and SSE events respectively to the largest values of 15.4 psl and 30.9 psi for wall AB-5 for OBE and SSE events respectively. The largest

values are 56.2% and 67.6% of the total allowable stresses of 27.4 psi and 45.7 psi for OBE and SSE events respectively.

For the walls in which the base boundary conditions are assumed to be fixed, it has been shown that the impact of including out-of-plane interstory drift effects can be significant, but in no caso do the stresses from all loads exceed the allowables.

TABLE 7-1 Wall No. Load Tension Normal to the Bed Joint From Drift Total quD (psD AB-3 DL+0BE 8.7 10.6

' 17.4 30.8 Lower DL+SSE AB-4 DL+OBE 8.1 9.0 Lower DL+SSE 16.2 18.3 I

AB-5 DL+0BE 15.4 17.7

! DL+SSE 30.9 42.9 AB-6 DL+OBE 7.9 12.8 DL+SSE 15.9 30.9 AB-7 DL+OBE 6.5 19.0 DL+SSE 12.9 36.8 AB-8 DL+OBE 11.0 25.9 DL+SSE 22.0 34.7

AB-9 DL+0BE 9.8 13.4 DL+SSE 19.5 32.6

, 7.3 Piping and Equipment Loads Piping loads were obtained in one of two ways as follows:

i 11

.~

1. Whenever available, pipe react lon loads from the pipe stress analyses were used.

2. For pipes with diameters of 6 inches or smaller, the maximum spacing as specified in ANSI B31-1 was used in calculating the mass of any piping attached to the walls. The mass of the pipes assumed the pipes were full of water and included the weight of insulation, if applicable. The seismic load from the piping was calculated by multiplying the mass of the pipe by 1.05 times the peak of response spectrum corresponding

'to the floor above the wall.

3. Thermal loads were only considered for the pipe anchors to the walls for the larger pipes. For the smaller pipes that had U-bolts or han0er supports, the thermal loads were ignored.

The determination of seismic equipment loads is discussed in detall In Section 5 of this report i

i t

,t l 12

l 8 ITEMS 7 and 13 - MODULUS OF ELASTICITY OUESTION l

Two of the requests from the NRC review referred to the Modulus of Elasticity as follows*

Item No. 7:

, 'In section 6.1.2 of Enclosure 3 [21, ranges of +/-25%

and +/-20% were given for the modulus of elasticity i

of ungrouted and grouted walls. respectively, to account for uncertaintles in evaluating the frequency of the walls. Tne Licensee stated. *1f the frequency ,

of the walls falls on the low frequency side of the ampilfled region of the response spectrum adequate provlsions are included to ensure that the determination of the strens in the wall is conservative."

The licensee should define and discuss these

" adequate provisions."*

Item No.13:

'The modulus of elasticity for grouted or solid walls was varied from 800 f'm to 1200 f*m. ACI 531-79 recommends a maximum of 1000 f*m. If the Licensee selects 1200 f'm in the analysis, an explanation shou!d be provided.*

RESPONSE

A number of uncertalntles exist in masonry walls with respect to variations in mass, modulus of elasticity and material and section properties.

To account for the effect of structural frequency variations resulting from these uncertaintles, the criteria for the TMI-1 Plant require that the modulus of elasticity of 1000 f*m be varied by an amount of plus or minus 25%. The actual value is selected to cause maximum stress in the wall ano thus is related to position of the fundamental wall frequency with respect to the frequency at which the peak of the appropriate floor response spectrum occurs.

.- If the fundamental frequency of the wall using an E of 1000 f'm is on the

- high frequency side of the amplified region of the response spectrum, then an E value of 800 f'm is used to calculate the fundamental wall frequency

, , and resulting stresses. This lower value of E will account for variations in

structural frequencies and will produce higher stresses in the wall because a higher spectral acceleration will result from the lower modulus. Conversely.

If the fundamental frequency of the wall using an E of 1000 f'm is on ,the low frequency side of the amplified region of the response spectrum then an

. E value of 1200 f*m is used to calculate the fundamental wall frequency and resulting stresses.

13

For TMl-1 a value of E of 800 f'm was used in all cases since the fundamental frequency of walls was on the high frequency side of the amplified region of the response spectrum.

4

%y e

e O 14

9 ITEM S - IN-PLANE EFFECTS QUESTK)N ltem 8 of the NRC request for additional information states:

• With regard to the in-plane effects, the strength of the strut corresponding to a strain at cracking is given in expression (1) of Section 6.5 of Enclosure 3 (2]. The Licensee should provide a complete derivation of this expression and discuss how this expression relates to the permissible strain levels of unconfined and confined walls."

RESPONSE

The expression (1) referred to above was not used when deriving the recommended values for permissible in-plane strain. A full discussion on the dervlation of the permissible strain levels is given in Section 12.

4 [

15

10 ITEM 9 - BOUNDARY CONDITIONS QUESTION ltem 9 of the NRC request for additional information states:

'The Licenses should discuss and justify the boundary conditions used in the analysis of the 14 walls mentioned in Reference 3."

RESPONSE

The analysis procedure used by GPU to assess the out-of-plane response of the masonry walls is described in detall in Section 2 of this report.

To assess the out-of-plane response of the walls, the boundary conditions of the walls were assumed to be as follows:

A. For walls with two vertical boundaries, the two sides and bottom boundaries were assumod to be simply supported and the top was free. The walls in this category were AB-1. AB-2. AB-3U. AB-4U. AB-11. AB-12 and AB-13.

B. For walls with only one vertical boundary, the wall was considered to be free on both vertical boundaries and at the top, and fixed at the base. Walls in this category were AB-3L. AB-4L. AB-5. AB-6. AB-7 AB-8. and AB-9.

C. Wall AB-10 was assumed to span horizontally and therefore was considered to be free on the top and bottom boundaries whereas both vertical boundaries were assumed to be simply supported. '

A summary of the pertinent results corresponding to these assumptions are given in Tsble 10.1. These boundary conditions are considered to be appropriate for the following reasons.

1. The stresses resulting from out-of-plane seismic loads are conservative. I

2. A conservative evaluation of a wall with only one single vertical boundary is to assume that it responds as a free standing cantilever wall with a fixed base.

The Imp!! cation of the assumed boundary condillons with regard to out-of-plane drift effects is addressed in Section 7 of this report. -

t I

B 16

TABLE 13-1 .

TMl-1 CLOCK CALL ANALYSIS ,

STRESS

SUMMARY

STFESS IN P.S.I.

Fundamental W all Frequency Load Tension FlexuraD Compression FlexuraD Shear (Flexuren No. Hz Combination Actual Allowable Actual Allowable Actual Allowable 48.2 D+R+E 10.0 " " 41.1 14.2 313 3.4 34 AB-1 D+R+E' 17.4 = " 68.0 21.4 808 5.7 52 AB-2 55.0 D+R+E 6.3 " " 41.1 20.6 313 2.1 34 D+ R+ E' 10.5 " " 68.0 19.2 808 3.5 52 35.1 D+R+E 10.60" 27.4 23.91 313 1.7 34 AB-3 Lower D+ R+ E' 30.84" 45.7 42.42 800 3.9 52 AB-3 30.1 D+R+E 12.4 * " 41.1 16.2 313 5.6 34 Upper D+R+E' 2 2.5 = " 68.0 25.0 808 9.1 52 AB-4 34.9 D+R+E 9.0 " 27.4 26.32 313 2.2 34 D+R+E' 18.25" 45.7 43.74 808 4.1 52 Lower 50.5 D+R+E 7.3 " " 41.1 24.0 313 1.7 34 AB -4 Upper D+R+E' 1 1. 5 " " 68.0 22.0 808 2.7 52 ,

AB-5 32.0 D+R+E 17.74" 27.4 32.44 313 1.1 34 D+R+E' 42.88* 45.7 55.88 808 2.1 52 32.0 D+R+E 12.82" 27.4 27.34 313 1.4 34 AB-6 D+R+E' 30.88" 45.7 43.58 808 2.4 52

" 30.7 D+R+E 19.00" 27.4 31.97 313 3.9 34 AB-7 D+R+E' 36.77" 45.7 48.24 808 5.9 52 _

24.0 D+R+E 23.73" 27.4 46.4 313 5.5 34 AB-8 D+R+E' 37.90" 45.7 62.3 808 6.4 52 24.0 D+R+E 13.37" 27.4 23.57 313 1.2 34 AB-9 D + R+ E' 32.64" 45.7 45.24 808 3.5 52 D+R+E 6.2 5 " " 41.1 32.0 313 2.1 34 AB-10 20.7 D+R+E' 12.50"" 68.0 41.0 808 3.0 52 66.0 D+R+E 12.26"" 41.1 20.9 315 1.3 34 AB-11 D+R+E' 12.3" 45.7 19.2 808 2.6 52 68.5 D+R+E 3.1 " " 41.1 25.9 313 1.2 34 AB-12 D+R+E' 6.4 " " 68.0 24.7 808 2.1 52 45.0 D+R+E 7.1 " " 41.1 27.0 313 4.09 34 AB-13 D+ R+ E' 1 1.4 " " 68.0 28.3 808 6.76 52 D' = Dead Load including Permanent Equipment Loads l R = Pipe Reactions l E = OBE Loads l E' = SSE Loads -

l

= Tension Normal to Bed Joint

"" = Tension Parallel to Bed Joint

4 a 11 ITEM 10 - STEEL BRACKETS FOR REINFO.CING QUESTION Item 10 of the NRC request for additional Information states:

'The Licensee plans to provide steel brackets to reinforce the end spans of the north and south walls of the elevator shaft. The Licensee should evaluate the out-of-plane drift effects that would result from these brackets.*

RESPONSE

in reference to th3 above item GPUN have decided not to use brackets to support the end walls of the elevator shaft. Instead they have decided to provide supports at the free end of the walls to sufficiently reduce the bending moments resulting from the dynamic response of the walls. For this modification use is made of the existing U-frame inside the elevator shaft. The appropriate pages of GPUN Topical Report No. 001 will be revised and resubmitted to the NRC.

Because the brackets have now been abandoned, the out-of-plane drift problems associated with them do not exist.

e

)

18

12 ITEMS 11 and 15 - IN-PLANE SHEAR STRESSES AND SHEAR STRAINS QUESTION Two of the requests from the NRC review referred to the allowable in-plane shear stresses and shear strains, as follows:

Item No 11:

• With reference to the In-plane effects for factored g

loads, a factor of 1,67 was introduced to the allowable in-plane strain. The Licensee should provide the technical basis for this factor.'

In-plane shear stresses are then addressed in the following questions:

t item No.15:

"With regard to shear for reinforced masonry, the Licensee introduced test results on shear strength of reinforced masonry. Speelfically, Figure 2 of Enclosure 3 121 presented test data for various percentages of reinforcement. For the case in which there is more than 0.3% horizontal reinforcement, there is only one test value for MND = 1.0 and there are two test values for MND = 0.5. For the case in which there is less than 0.3 % horizontal reinforcement, there are no test data for MND = 1.0.

The data presented do not appear to be sufficient to justify use of these values. The Licensee should discuss the technical basis for the applicability of these tests to the walls at TMI-1 Unit 1 with respect to the mortar type, boundary conditions, and nature of the loads (i.e. dynamic, static) and should identify and provide the source of these tests."

RESPONSE

The folicwing subsections present an evaluation of available test data relative to both shear stress and shear straln.

12.1 Overview of Test Program The results from an ongoing masonry test program being performed at the Earthquake Engineering Research Center. University of California, Berkeley.

i are used in this section to evaluate the In-plane shear strength and strain I

of masonry piers. The tests basically consist of subjecting masonry plers I

to an in-plane cyclic shear load with the test setup shown in Figure 12-1.

The results of the research have been reported in references 12.1. 12.2,

+ 12.3, 12.4 and 12.5 The piers are tested by applying three cycles of' load at a specified amplitude. The amplitude is gradually increased as the test 19

a g progresses until the pier is unable to resist any further load. Each test was photographed after each set of three cycles of load, thereby providing detailed records of the crack pattern.

To date over one hundred piers have been tested using three different types of materials. Thirty-five of the piers tested were constructed from hollow concrete block masonry units and of these six had a height-to-width ratio of 0.5. fifteen had a height-to-width ratio of 1, and fourteen had a height-to-width ratio of 2. The plers were constructed from either 6-inch or 8-inch wide hollow concrete block units using Type M mortar. The strength of prisms constructed from the same materials that were used in the piers varied from 1350 to 3500 psi.

The information obtained from each test consisted of a plot of the force-deflection relationship for each cycle of loading. From this set of curves several parameters could be determined, including:

(a) Ultimate Strength (b) Stiffness Degradation (c) Hysterests Envelope (d) Deflection of Pier at each Loading Stage 12.1.1 Applicability of Test Flesults i

The information obtained from the Berkeley test program is valuable in evaluating the in-plane shear performance of masonry plers subjected to selsmic loads. A discussion on the applicabilty of the test results

,j ls discussed separately with respect to the following variables -- loading, 1 size of test specimen, boundary conditions, material strengths and reinforcement.

A. Loadino Although an earthquake type time history was not used as the input motion to the test specimen, the gradually increasing, amplitude dependent cyclic loading was typical of that used in many other test programs on reinforced concrete and steel structural elements.

The most important aspect of loadin0 required to evaluate the seismic performance of structural elements is that the loading be ' yclic or reversed. Other variables such as the rate of loading, sa',uence of loads, etc., may be important but are secondary in comparison to the requirement that the loading be cyclic.

9 20

B. Size of Test Specimen The size of the test specimen used in the Berkeley test program was limited by the capacity of the actuators. The plers with a height-to-width ratio of 0.5 were 3 ft. 4 inches high and 6 ft. 8 inches long, the 1 to 1 plers were 4 ft. 8 inches high and 4 ft.

long whereas the 2 to 1 piers were 5 ft. 4 inches high by 2 ft.

8 inches long. Although these sizes are generaly smaller than the walls found at the Three Mlle Island Nuclear Station Unit 1.

It is assumed that they are of adequate size to represent the behavior of larger sized walls with the same aspect or height-to-width ratio. It should be noted that no experimental evidence is available to validate or refute this assumption.

The aspect or height-to-width ratios included in the test program cover all the walls at the Three Mlle Island Nuclear Station Unit 1.

C. Boundary Conditions The boundary conditions of the plers tested in the Berkeley program were such that moment fixity was forced at both the top and bottom of the piers with no constraints on the vertical edges. Although this set of boundary conditions is different from that of most of the walls at the Three Mlle Island Nuclear Station Unit 1. It is believed that if the walls at the plant are confined either on three or four sides or at the top and bottom, then the performance of the walls will be similar to those tested in the Berkeley program.

Confinement should be provided by either walls or columns capable of resisting the loads imposed by the concrete block walls.

D. Material Strenoths The assumed compressive strength f'm of the walls at the Three Mile Island Nuclear Station Unit 1 was 900 to 950 psi. This is lower than the range of 1350-3500 psi of the prism strength of the piers included in the Berkeley test program, however. It is our opinion that the actual insitu f*m of the walls is within the range tested in the Berkeley program, i

E. Reinforcement The majority of the walls at Three Mile Island Nuclear Station Unit 1 are unreinforced whereas all the piers of the Berkeley test program were reinforced. It is our belief that provided the walls at the l Three Mlle Island Nuclear Station Unit 1 are confined o. three or four sides or at the top and bottom, then cracks le. the unreinforced wall will occur at similar strain levels to the piers tested.

21 J

12.2 Evaluation of Test Data Tre data from thirty-five tests performed on hollow concrete block piers was evaluated on the basis of shear stress and strain. in-plane loads on walls result from both imposed deflections and shear forces % posed by piping and other equipment. ,

Section 12.2.1 explains how the test data was evaluated to determine the permissible strains for both OBE and SSE events. Section 12.2.2 and 12.2.4 provide statistical evaluation of the stress and strain data respectively, whereas sections 12.2.3 and 12.2.5 present discussions of the results.

12.2.1 Shear Stresses and Strains for OBE and SSE Events The test results from the Berkeley program were evaluated to determine in-plane shear stresses and strains appropriate for OBE and SSE events.

The evaluation was performed so that the function of a wall would not be impaired while it was resisting out-of-plane loads. During each pier test. photographs were taken after each set of three cycles of load at a specified amplitude. These photographs in conjunction with thr hysterests envelopes developed for each test were used to determine the appropriate state of cracking due to in-plane loads that could be tolerated from an OBE and SSE event. For an OBE event, the loading stage at which In!tlal visible cracks occurred was used. For an SSE event additional cracking was permitted, however, the loading stage was prior to any significant diagonal cracking. Obviously the evaluation for an SSE event required judgment and photographs shown in Figures 12-2,12-3. and 12-4 show the typical state of cracking used for both an OBE and SSE event for plers with height-to-width ratios of 0.5.

1. and 2. respectively. At each appropriate level of cracking the corresponding shear stress and displacement were determined. The displacement was divided by the wall height to determine a corresponding shear strain. The shear stresses and strains obtained were statistically evaluated and these results are presented in the following subsections.

12.2.2 Statistical Analysis of the Stress Data i

A total of 33 tests were used to evaluate the shear stress for both OBE and SSE events. These were divided as follows:

Masonry Takes the Shear; MNd = 0.25  : 1 Test MNd = 0.5  : 4 Tests MNd = 1.0  : 7 Tests l e 22 l -- - - - _ _ . - _ _ ._.

Reinforcement Takes the Shear:

MNd = 0.25  : 5 Tests MNd = 0.5  : 10 Tests MNd = 1.0  : 6 Tests It is recognized that the number of tests is small in most cases. but these tests are the only ones available.

To begin with, all the test data was normalized by dividing the stress values by the square root of the appropriate f'm. This leaves the constant. C. of the equation:

T=CG This constant is the subject of the statistical analysis that follows herein.

From this data, the following parameters were calculated for the shear stress for all cases. both for OBE and SSE:

I

1. Sample Mean 00

11. Standard Deviation (s)

These statistics were then used as the parameters for the distribution of the population.

The statistical distribution assumed generally applicable for the data is the Gamma distribution. The main reason for this is that the data never takes on negative values. The Gamma distribution is defined by:

) =

)(Ax)k-l e

• fx (x) (k-1)l x>0 and has a mean value of k/A and a coefficient of variation 1/8. It is to be noted though that for k > 15. the Gamma and the Normal distributions aro close and that the Normal distribution is assurr.ed for those cases.

The 95% confidence Interval for the mean of the population m was calculated, assuming that the normalized variable X-M U

is t-distributed, and that the actual population standard deviation 6x.

t is unknown. Here n is the sample size. , ,

l

, 23 l

For the Gamma distribution. confidence intervals on parameters such as m -ka, have no meaning and must be reinterpreted. On the normal curve m-lo" corresponds to a point on the cumulative distribution curve with an ordinate of 0.1587. This means that approximately 16%

of the area under the probability density curve lies to the left of m-la.

Similarly, m -2a and m-30 correspond to points with ordinates 0.02275 and 0.00135. respectively. Based on the confidence Interval for the mean, confidence intervals were calculated for values of the Gamma distribution for which the cumulative distribution function had values of 0.1587. 0.02275 and 0.00135. respectively.

The results from this analysis are presented in Tables 12-2 and 12-3.

The results were then compared with the criteria specified factors on the fm for the allowable shear stress as given in Table 12-1.

TABLE 12-1 I I ME l ME l Masonry Takes Shear MNd) 1 0.9 C. 34 max. 1.5 G 56 max.

MNd = 0 2.0 74 max. 3.4 G 123 max.

Reinforcement Talms Shear MNd ) 1 1.5 G 75 max. 2.5 G 125 max.

MNd = 0 1.0 G 120 max. 3.4 G 180 max.

Probabilities that the criteria specified allowable stress would exceed the stress based on the test results were calculated under two assumptions: firstly. that the population mean was equal to the sample mean, and secondly. that it was at the lower end of the 95% confidence interval.

Finally, safety factors based on the 95% confidence Interval for the mean were calculated for the shear stresses.

These results are presented in Table 12-4.

24

TABLE 12-2 .

8 MASONRY TAKES THE SHEAR h=0.25 h=0.5 h=1.0 OBE 4 7 Sample Size 1 2.471 Sample Mean (m) 3.476 2.465

- 0.703 0.362 Standard Deviation (s) 14.6 %

- 28.5%

Coeff. of Variation 95% Confidence Interval: ~

~

2.136 1 m i 2.806 On (m)

~

- 1.347 1 m 1 3.583 ~

- 0.662 1 m-s 1 2.880 1.774 1 n-s 1 2.444 On GR-s) 1.412 1 5-2s 1 2.082 m

On (ih-2s)

- 0.334 1 m-2s 1 2.177 0'

- 0.125 i Hi-3s 1 1.474 1.050 i m-3s 1 1.720 On (m-3s)

SSE:

4 7 Sample Siza 1 2.594 4.500 2.844 Sample Moan (m) 0.379

- 0.627 Standard D6.!atlon (s) - 22.0 % 14.6%

Coeff. of Variation 95% Confidence Interval: ~

1.846 1

~

m 1 3.842 2.243 1 m 1 2.945 On (E) m-s 1 1.225 1 m-s 1 3.215 1.864 1 2.566 On (E-s) -

1 2.588 1.485 1 E-2s 1 2.187 On (5-2s) - 0.807 1 m-2s m-3s 1 1.808 0.502 1 m-3s 1 1.961 1.106 1 On (5-3s)

TABLE 12-3 .

8 REINFORCEMENT TAKES THE SHEAR M M M g = 0.25 g = 0.5 g = 1.0 l

DEE* 10 (4) 6 Sample Size 5 4.113 3.344 (4.531 ) 3.233 Sample Mean (m) 0.250 0.398 1.251 (0.752 )

Standard Deviation (s) (16.6%) 7.7 %

9.7 % 37.4%

Coeff. of Variation 95% Confidence Interval: 4.239 2.449 i m 1 1 4.607 (3.334 1 m 1 5.728 ) 2.971 1 m 1 3.495 On (m) 3.619 i Tn 1.277 < m-s < 3.007 (2.582 5 Hi-s 5 4.976 ) 2.721 1 m-s 1 3.245 On (m-s) 3.221 1 hi-s 1 4.209 1

0.646 1 m-2s 1 2.122 1_

(1.830 1 hi-2s 1 4.224 ) 2.471 1 m-2s 1 2.995 g On (m-2s) 2.823 1 m-2s 1 3.811 1.440 0.284 1 m_-3s 1 (1.078 1 m-3s 1 3.472 ) 2.221 1 m_-3s 1 2.745 On (m-3s) 2.4 2 5 _< m-3s 1 3.413 SSE: 10 (4) 6 Sample Size 5 4.111 (5.117 ) 3.588 Sample Mean (m) 5.247 l 1.251 (0.415 ) 0.365 Standard Deviation (s) 0.548 t

30.4% (8.1 %) 10.2 %

Coeff. of Variation 10.4 %

95% Confidence Interval: m 1 5.006 3.216 1 1 5.927 (4.457 1 m 1 5.777 ) 3.205 1 m 1 3.971 On (m)

~

4.567 i Tn 3.755 1.988 1 m-s 1 (4.042 1 m-s 1 5.362 ) 2.840 i m-s 1 3.606 4.019 1 in-s

~

On (m-s) 1 5.379 1.212 ' m-2s ' 2.504

~

(3.627 5 m~ 5 4.947 ) 2.475 1 m~-2s 1 3.241 On (m-2s) 3.471 1 rn-2s 1 4.831 1.253 0.683 i m_-3 s i _

2.876 On (m-3s) 2.923 i m-3s 1 4.283 (3.212 i m-3s 1 4.532 ) 2.110 - m-3s Berkeley.

• The numbers in parentheses come from evaluation of yet unpublished data from tests at U.C..

TABLE 12-4 .

s Masonry Takes the Shear Roentorcseont Takse the Sheer h=0.25 h=0.5 h=1.0 h=0.25 h=0.5 h=1.0 DDE Probability of Exceedence: 79/1.000 0*

- 52/1.000 13/1.000.000 0*

l KEY A (108/1.000.000) 320/1,000.000 45/1.000.000 32/100 0*

- 62/100 KEY B (18/1,000) 1.93 - 2.46 1.40 - 2.42 1.98 - 2.33 2.02 0.93 - 2.47 2.37 - 3.12 Range of Safety Factors (1.91 - 3.27)

On the Mean ESE ro N Probability of Exceedence: 178/1.000 14/10.000

' - 26/100 1926/1.000.000 78/1.000.000 KEY A (0 *)

25/1,000 6/1,000 47/100 27/1.000

- 69/100 KEY B (142/1.000.000) 1.44 - 1.87 1.09 - 1.70 1.28 - 1.59 1.54 0.75 - 1.57 1.50 - 1.96 Range of Safety Factors (1.51 - 1.96)

On the Mean

• Probabilities of exceedence less than 1 In 1.000.000.

• • Values in parenthesis come from evaluation of yet unpublished data from tests at U.C. Berkeley.

12.2.3 Discussion of Stress Results The tests performed at U.C. Berkeley were performed under different conditions and on smaller units than generally exist at the TMI-1 plant.

However. because most of the TMI-1 walls have a boundary support of at least three sides, including the bottom, they are considered comparable to the test walls which had only top and bottom boundarles, both cases being within the definttlon of a confined wall. Furthermore, it is our belief that the Berkeley test walls were of a sufficient size to be applicable to a general wall of similar height to width ratios.

In the following discussion. note that due to the top and bottom fixity of the test specimens, the MNd ratio equals half the height-to-width ratio.

Looking at the values of Table 12-4. we see that in general the results for MNd = 0.5 have a much higher probability of exceedence than do the other values. Unfortunately, some of the walls tested in that series were subject to unforeseen problems with the test setup and which had some adverse effects on the results. A relatively low confidence is thus placed on those tests. Four tests whose results have not yet been published have since been performed on walls with MNd

= 0.5. but only for the case where the reinforcement takes the shear.

The results of a statistical analysis of those tests are presented in Tables 12-2.12-3 and 12-4 in parentheses. The results on which we place a lower confidence will be ignorert in the subsequent discussion.

In general, the probabilities of exceedence are low, it can be stated that criteria allowable shear stress will exceed the actual value determined from tests 108 times in 1.000.000 (0.01 %) for OBE events and 1926 times in 1.000.000 (0.19%) for SSE events if the population mean strength is taken at the center of the 95% confidence interval. If one considers the extreme case where the population mean is taken to be at the lowest end of the 95% confidence Interval, then these figures become 18 in 1.000 (1.8%) for OBE events and 27 in 1.000 (2.7%) for SSE events. Given the extreme nature of the assumption on which these second estimates are based. these probabilities of exceedance are deemed satisfactory.

By taking the 95% confidenco intervals on the population mean, the factor of safety associated with the criteria allowable shear stresses for the case of the masonry taking the shear Is 2.02 4 SF 4 3.12 and 1.50 4 SF 4 1.96 for OBE and SSE events. respectively. For the case of reinforcement taking the shear, these values are 1.914 SF 4 3.27 and 1.28 4 SF 4 1.96 for OBE and SSE events, respectively.

12.2.4 Statistical Analysis of the Strain Data A total of 34 and 35 tests were used to evaluate the , shear strair) for the OBE and SSE events respectively. The shear strains were determined by the procedure described in Subsection 12.2.1. From this data the following parameters were calculated for the shear strain for both the 28

OBE and SSE events:

1. Sample Mean 00

11. Standard Deviat:3n (s)

These statistics were then used as the parameters for the distribution of the population. For each case (OBE and SSE). two underlying i

distributions were evaluated, and the effect of the choice of distribution on the results was examined. The more reasonable distribution was then accepted. The two underlying distributions were the normal distribution and the gamma distribut!on, and the gamma distribution was chosen as best representing the test data for reasons given in later In this section.

The interpretation of the 95% confidence Intervals for the Gamma and Normal distributions are explained in Section 12.2.2 and will not be repeated here.

These actual distributions were then compared with the criteria specified allowable shear strains, i.e. 0.0008. respectively for the OBE condition, and 1.67 times this value. for the SSE condition. Probabilities that the criteria specified allowable strain would exceed the actual Strain based on the test results were calculated under two assumptions: firstly, that the population mean was equal to the sample mean, and socondly.

that it was at the lower end of the 95 % confidence interval.

Finally. safety factors based on the 95 % confidence interval for the mean were calculated for the shear strain. The results are presented in Table 12-5 below.

TABLE 12-5 OBE SSE t

Sample Size 34 35 Sample Mean (m) 0.00202 0.00318 Sample Standard Deviation (s) 0.00085 0.00094 j

Coefficient of Verlation 42 % 30 %

. e t

The 95 % confidence intervals on the population mean are: e OBEt 0.00172 4 m 4 0.00232 SSEt 0.00286 4 m 4 0.00350 29

The effect of the assumption of normal distribution versus the assumption of gamma distribution was studied. A plot of the histograms of test data for both the OBE and SSE conditions are shown in Fig.12-5.

Two observations are as follows.

(1) The data never takes on negative values.

(ii) The distribution of data is skewed.

especially for the OBE condition.

Both of these observations Indicate that the gamma distribution is preferable to the normal distribution. The gamma distribution is defined in Section 12.2.2 and in this case the following values of k andfgive best fits to the OBE and SSE data:

Case k A OBE 6 2970.3 SSE 11 3459.1 These curves are also plotted in Fig.12-5.

It should be noted that there is no physicci reason why shear strains should have any particular distribution. However, by suitable adjustment of the parameters k and ) , the gamma distribution can be made to describe the best data far more accurately than can the normal distribution.

The 95 % confidence intervals corresonding to the 10. 26. and 36 levels are as follows:

! (i) Corresponding to cumulative distribution function

= 0.1587 (10' level)

OBE 0.00090 4 X 4 0.00146 SSE 0.00192 4 X 4 0.00257 l

(ll) Corresponding to cumulative distribution function l

= 0.02275 (2a level)

OBE 0.00045 4 X 4 0.00091 SSE 0.00129 4 X 4 0.00189

. (Ill) Corresponding to cumulative distribution function

' = 0.00135 (30 level)

OBE 0.00020 4 X s 0.00053 '

• SSE 0.00081 4 X 4 0.00134 30

Using the allowable strain values from the criteria for confined walls ,

and the above 95 % confidence interval on the mean the following j limits on the factor of safety are established: I 4

OBE: 2.15 4 SF 4 2.90 SSE: 2.13 4 SF 4 2.61 The probabilities that the criteria specified strain values will exceed the available strain capacity based on test results and the gamma distribution

> are then as follows:

Key OBE SSE A 0.034 0.008 B 0.119 0.029 NOTE:

(1) KEY A in the table above gives the probabilities of exceedance assuming the population mean equals the sample mean.

(2) KEY B gives the probabilities of exceedance assuming the population mean is at the lower end of the 95%

confidence Interval.

12.2.5 Discussion of Strain Flesults l

One of the main values of the test data generated in the Berkeley test j

program is that it enables a reasonable estimate of the deflections i

or strains at which various levels of cracking could be expected in a masonry wall.

By taking the 95 % confidence Intervals on the population mean. the factor of safety associated with the allowable strain of 0.0008 for an OBE event is 2.15 4 SF 4 2.90. For an SSE event the corresponding range is 2.13 4 SF 4 2.61 based on an allowable strain of 0.00134.

In terms of probability, it can be stated that code allowable strain will exceed the actual strain obtained from the tests 34 times in 1000 for OBE events and 8 times in 1000 for SSE events if the population

- mean strength is taken at the center of the 95 % confidence intervals.

t if one considers the extreme case where the population mean is taken to be at the lower end of its 95 % confidence Interval, then these figures become 119 times in 1000 for OBE events atid 29 times in

' 1000 for SSE events. Given the extreme nature of the assumption on which these second estimates are based and the self-limiting nature of the load, these probabilities of exceedance are deemed satisfactory.

31

12.3 Conclusions. '

in view of the above discussion of results for both shear stresses and shear strains, it is concluded that the criteria specified increase factors for shear stresses and shear strains of 1.67 - 1.7 and 1.67 respectively for factored loads are reasonable for the reevaluation of the Three Mlle Island Nuclear i Station Unit 1 '

i 12.4 References.

l 12.1 Mayes. R.L. Omote. Y., and Clough. R.W., " Cyclic Shear Tests of Masonry Piers. Volume 1 - Test Results."

EERC Report No. 76-8. May,1976.

12.2 Hidalgo. P.A., et al., " Cyclic Loading Tests of Masonry Single Piers. Volume 1 - Height to Width Ratio of 2.0.* EERC Report No. 78-27. November 1978.

i 12.3 Chen. S.W. et al.. " Cyclic Loading Tests of Masonry .

Single Piers. Volume 2 - Height to Width Ratio of 1.* EERC Report No. 78-28. Dec. 1978.

12.4 Hidalgo. P.A., et al.. " Cyclic Loading Tests of Masonry Single Piers. Volume 3 - Height to Width Ratio of 0.5.* EERC Report No. 79-12. May 1979.

12.5 Sveinsson B.I., et al.. " Evaluation of Seismic Design Provisions for Masonry in the United States." EERC Report 81-10. August 1981.

i l

l l

t l.

32 i

STIFFEN TOP WF BEAM (ADDED1(wl4al5I)

TOP WF BEAM (ORIGINAL)(Wl4a127) 7f gACTUATORS FORCE = P DISPLACEMENT CONTROLLED

/ j

/ Bi 7 -

l l STRONG (MTS SERVO JACKS)

HINGE BACK CONNECTED TO l l l LOAD CELL REACTION FRAME

- l i -'

ACTUATOR SPECIMEN FLANGE l l l Va V(P) s A ACTUATOR V = V(P)

I I I FORCE CONTROLLED HINGE I I I I #I .

CONCRETE BASE BLOCK 8 Y  % \ #-A 7

T

^ ^ -

FLOOR LEVEL

//////// ////////

\/ //// /////////

\ / /////////

STEEL PLATE WITH HEAVY SHEAR KEYS BOTTOM WF BEAM (Wl4 al27)

(TOP AND BOTTOM)

FIGURE 12-1 SCHEMATIC ILLUSTRATION OF SINGLE PIER TEST i

M M M - - m m m H/W RATIO = 0.5 -

- . g i y ,

g . -

4 3 J

.: s ,' * *

't?

S .,

l

. ,/-

. \ / ,/ s' s-s /~

.-_ \ N. ,

42<. -

f*, \ 4

.- /' , g 7 WJ a, .N ./

g. J. /.s J d- ,,

' . i'a*'".. -

... ./ ; , <

, ; ;,'h 4  %

.r , ** '/ '

' N'

'h f .' '

N. j

. \' [ - e d' #f,,

I dM ,; N ,e f ah%'t 2 h ki,Ni .1,, 'PO,'M*/,. #

L

'/* *c - fg - ..'

s . /- . 3.1%>q%g

'.'M

r. 'u t : -* f4,v '

.' Q d.

u y

,h g .

.a

e. I

1. 1i

. t,

.s.

4%*

v' ' * .I , [ '- N t , y',e

t , q.,, ,

,z m3 f y/ ),

3 ,

w b

.a -

N , , ," /

, _ ,t d , f l

o .~

." x h 3,m'?9 '9 K,.3. ;,;% . ' .

r s W d gg y :- -s !.1 s" l'.~-

. , ' .. .s un a. i

,FN ' . Td ,.'.gi, ,f y * ~ ,' _ g, . ? y - '

1

~. . .s- .

, me

- y :v.r i ,7 /ec p >

< b', 7$./. :M .

. , u .- w.,e

. . .'.g rit

-e. ..sp _ ..

. - c,', y .. o

^^

., .cy3v -

4

y , f['s";:. ' '

I" 7 '

. . g ., ,,4.A -_A .. A IA-ev _ " 'i[ .

FIGURE 12-2 SUCCESSIVE CRACK FORMATION AND EXPERIMENTAL RESut.TS, TEST HCBL-12-3 l--- - .. . . . . . . _

I. .

H/W RAT)O : 1.0 l

- U s . ._ ,

, m ,_ _. ,

I 1 .'  ?

g ~

, ,. , _r 4. .

e

'<s..._._, .' ,

t to ,A

,y l y,,~8'.

e 6

m. m .,

. M ,a A-s

_- u- -

_ h. .b I

I -

8 . y-  ? ,- -

v ~.)

ry;. x ,- gy ;) .

l

s -

g r- -)-r y xg

'\ ,

,-:r r s, 1 <,,

\p 7 4:

. 15. A  ?- a I

',,.. i y. , 15. .

~ ' ~

. g 's -

> g.

I s.

. , l sg

s. 2-ya N, ,/ s a

'g.

u, y , .g I

. > j.s =* Ja 1,

• _b ? ,

.* ., . ./.! , L_ e_ .. -.

H

-~ g g ~ - 3,g- ..-

g u.*. _ ia .. ._

I FIGURE 12-3 SUCCESSIVE CRACK FORMATION AND EXPERIMENTAL RESULTS, TEST H C B L- 11 -4 I as

H/W RATIO = 2.0 g .

i meme

~

..- 1 g .

s s ' .

g I

.< g ,

.L. .N C.

~ '

(. '

g

• 7'1  : , .

,\

[I f I e + vt DIRECTION SHEAR ORCEA SHE AR STRESS

- vE L ATE R AL oiSPL CoRRES E .?

I SHE AR FORCE AND SME AR STRESS O.16"; 23.O KIPS O.29"; 18.8 KlPS O.32"; 19.9 KIPS

. 12O PSI 9 8 PSI 104 PSI O.13"; 20.1 KIPS O.18"; 21.7 KIPS O.24"; 19.1 KIPS -

150 PSI Il3 PSI 99 PSI 1 -_ ,, , .. 3 3

" '" 1

) m]

" r

. . . . ., ., . j ,_. ,

I

%:-. 'j., .

/ {

ll . ,

\ U

.i-lk ]}ll'

~

g l \ . l lL 1 N,.f.O.. ,e j

.,_f\.. - , . l l yQ -

.g '

q ,

s' '. _ _

{

.  ; . .....

_ /,  ;

l e .e

--ib r I 0. 37"; 19.9 KlPS 104 PSI 0.51"; 10. 5 KI P S 55 PSI 0.7 "

s -

I"

- ;- 9

{

O.28"; 19.7 KIP S O.41"; 13.6 KIPS O.7" I" 102 PSI. 71 PSI FIOURE 12-4 SUCCESSIVE CRACK FORMATION AND EXPERIMENTAL RESULTS. TEST HCBL-21-6 j1 ~36 -

, 00mmo Distributi:nc: CBE & SSE l

l l

Ak '

Frequency mean 0.00202 '

std. dev. 0.000825 10 <

• ,g l r

' \ Gamma (k .6 A.2970.3) e- j \/

/

\

8

/ \

r

\'

s N

1. l* \

4 / 's a l's '

2 ./

i N!%

i/ \

/) N'N g--

j 0 10 5 15 20 25 30 35 40 45 -4 Strain to x:0.00202 OBE STR AIN D ATA A -

, Frequency ,

10 Lmean 0.00318 l l

f std. dev. 0.000959 8

[

G amna(k .11. 'A3 34 59.1)

,s' '% j/

6 e s

,/

/ \ %

1 4

i

/ 's N y

/ s 2- '

l  %

N ,'

, * ~~. w

a. r O 5 10 15 20 25 30 35 40 45- 50 5$ 16' 7.0.00318 SSE STRAIN DATA FIGURE 12-5 DISTRIBUTIONS OF OBE AND SSE STRAIN DATA 37

13 ITEM 12 - ENERGY BALANCE TECHNIQUE QUESTION ltem 12 of the NRC request for additional information states: l

• With regard to the

• Energy Balance Technique" and the ' Arching Theory' the Licenses should not resort to these approaches. If possible."

RESPONSE

This item was addressed in a letter from GPUN to NRC 5211-82-085 dated April

15. 1982.

9 t

G

. ,t 5

38

14 ITEM 14 - COLLAR JOINT STRENGTH QUESTION ltem 14 of the NRC request for additional information states:

- *With regard to the collar joint strength, the Licensee used the same test value that was used for the Trojan plant. The Licenses should discuss the applicability of this test to the TMl-1 masonry walls. In Reference

3. the Licunsee proposed that the collar joints of multiple wythe block walls be filled with non-shrink Portland cement grout. The Licensee should provide technical data to support the use of this grout and indicate how this repair will strengthen the collar joint. Furthermore the LlCensee should clarify whe;her the auxiliary building has any multiple wythe block walls and, if not, explain why this proposed modification was introduced."

RESPONSE

14.1 Purpose of Collar Joints All the masonry walls in the Auxillary Building that were affected by IE Bulletin 80-11 are multi-wythe walls. In Reference 14.1. Section 4.1. It is stated

• that during the field survey, no collar joints were able to be verified in any of the multi-wythe block walls.* As a consequence. GPU had the option of analyzing all multi-wythe walls as a set of single wythe walls or provide modifications so that each multi-wythe wall had verifiable collar joints. The latter option was exercised to ensure that the multi-wythe walls were capable of meeting the reevaluation criteria.

All the collar joints of all multi-wythe walls were filled with a non-shrink Portland Cement grout. The grout used was Five Star Grout No.100 from U.S. Grout Corporation. The grout had a specified compressive strength of 5000 psi after 7 days. The shear and tensile strength of the repaired collar joints were tested and the tests are described in Section 14.2 14.2 Collar Joint Tests and Allowable Stresses The allowable tenslie and shear stressiss for collar joints included in the j reevaluation criteria for factored loads was 12 psl. This was based on  :

the values used for the Trojan Plant provided there was a verifiable and )

well-constructed collar joint. The strengh of the collar joint was verifled by tests. Core samples were taken from the walls and tested for shgar j and tenslie strength.

The shear strength was determined by firmly securing one end of the core i specimen while applying a load across 3 Inches of the unsupported end 39

until failure. The tenslie strength was determined in accordance with ASTM C-496. " Standard Test Method for Splitting Tensile Strength of Cylindrical Concrete Specimens.*

The results of these tests were as follows. The average shear strength of 7 test specimen was determined to be 192 psi with a coefficient of variation equal to 6.4%. The average tensile strength of 7 test specimen was determined to be 182 psl with a coefficient of variation equal to 6.9%.

Despite these high average test values for the strengthened collar joint.

the criterla specified allowable value of 12 psl is still considered applicable.

This implies an average factor of safety equal to 15.8. which is very satisfactory.

14.2.1 Reference 14.1 *Three Mlle Island Unit 1 - Reevaluation of Safety-Related Concrete Masonry Walls - NRC IE Bulletin 80-11. Topical Report No. 001, Rev 0.*, GPUN Report dated July 1981.

. W t .

1 1

c 40

h 15 ITEM 18 - SHEAR - UNREINFORCED MASONRY OUESTION ltem 16 of the NRC request for additional information states:

"With regard to shear for unreinforced masonry, a factor of 1.5 was introduced for allowable shear for factored loads. SEB criteria (41 suggest a factor of 1.3. The Licensea should provide any !Iterature or test data to support the use of a factor of 1.5.*

RESPONSE

! The response to this item is presented in the paragraphs below.

The in-plane loads generated in the masonry walls of the TMl-1 plant are very small. Table 15-1 below lists the in-plane shear stresses and the ratios to the allowable stresses both for OBE and SSE events. A study of the effect of using an increase factor. (LF), of 1.3 for factored loads is also shown in Table 15-1.

TABLE 15-1 l IN-PLANE SHEAR STRESSES IN TMi-1 MASONRY WALLS WALL OBE SSE NO. v (wh ) v (wh ) (wh )

(psD (psD LF=1.5 LF=1.3 AB-1 0.24 0.009 0.4 0.010 0.011 AB-2 None .- None .- .-

AB-3g 1.13 0.041 2.26 0.054 0.063 0.34 0.012 0.62 0.015 0.017 AB-SU AB-4L 0.94 0.034 1.88 0.045 0.052 l, 0.15 0.005 0.30 0.007 0.008 AB-40 AB-5 0.98 0.035 1.96 0.047 0.054 AB-6 1.13 0.041 2.03 0.049 0.058 AB-7 1.85 0.067 2.91 0.070 0.081 AB-8 1.88 0.068 2.26 0.054 0.063 AB-9 1.13 0.041 2.26 0.054 0.063

. AB-10 None .- None .- .-

AB-11 0.40 0.014 0.83 0.02 0.023

{

AB-12 0.40 0.014 0.83 0.02 0.023 AB-13 0.60 0.022 1.00 0.024 0.028 I i

41

As can be seen from the above table, the in-plane stresses generated in the TMI-1 walls are very low. The maximum stress for both OBE and SSE events is in well AB-7 and is only 6.7% and 7.0% of the allowable stresses for OBE and SSE events respectively. Using a stress increase factor of 1.3 for factored loads, the maximum stress ratio for SSE events only increases to 8.1% with the remainder of the walls being less than 7.0% of the allowable.

Because the highest value of in-plane shear stress is only 8.1% of the SEB suggested allowable stress for factored loads, the SEB suggested increase factor of 1.3 is accepted for TMl-1.

4 u

I I

I

[

l 1

i A I

42

18 ITEM 17 - TENSION NORMAL TO BED JOINT QUESTION ltem 17 of the NRC request for additional Information states:

"With regard to allowable tensile stresses normal to bed joints. SEB criteria (4) suggests a factor of 1.3 for factored loads. The Licenses should discuss and

.P lustify the use of a factor of 1.5. The Licensee

}

should also discuss the applicability of those tests mentioned in Section 5.1.6 of Enclosure 3 (2) to the TMI-1 masonry walls.*

FESPONSE First, it should be noted that the increase in stress allowables for factored loads was specified in the criteria as 1.67 not 1.5 as stated in item 17 of the NRC request for additional information.

In the subsectior i below the available test data is evaluated statistically to justify the stress increase factor of 1.67 for factored loads for tension normal to the bed jolnt.

16.1 Ovurview of Test Programs Test programs in general have not covered the out-of-plane behavior of masonry walls constructed of solid masonry units very extensively. There Is therefore very limited data available for the analysis of the tension capacity of solid masonry normal to the bed joint. The tests referenced in (16.11 are the only available test results related to solid concrete masonry walls, it should be noted that these tests have all been performed on multi-wythe walls with solid concrete units forming one of the wythes, in all, results for 6 unreinforced test specimen involving only type S mortar as specified by proportion In ASTM C270, are available. All the tests were performed with a uniform pressure (alr bag) loading. This produces a parabolic moment d!stribution over the height of the wall with the maximum moment at center of the span.

9

?

16.2 Applicability of Test Results.

f

• The test results described in Section 16.1 are used in the following sections g

to justify the allowable stress factors for tension normal to the bed joint for solid concrete masonry for factcred loads.

In our opinion the test results from static, monotonic tests are applicable in determining allowable tension stresses for the following reasons:

43

1. An unreinforced masonry wall responds elastically to selsmic loads provided it is not cracked.

2. There are no test results available indicating that dynamic loading reduces the tenslie strength normal to the bed joint. In fact the only test data available for any type of cyclic loading on masonry structural elements indicates that the in-plane shear strength of masonry shear walls tested pseudostatically is 8-23%

less than that of a 3 cps equivalent dynamic test (Reference 16.2).

3. Cyclic or shake table tests are essential to determine the post-cracked or inelastic performance of structural elements. However, they are not essential to determine the ultimate or cracking strength of structural elements.

4. Points 1. 2 and 3 above indicate that the uniform load tests are reasonable methods to determine the cracking or tensile strength of an unreinforced masonry wall subjected to out-of-plane loads, ideally. only tests with similar mortar types to the type N used at Three Mlle Island Nuclear Station. Unit 1 would be selected for statistical analysis.

Due to the limited data, only tests involving type S mortar were available.

However, the specil'ed allowable values use a non-dimensionalized stress allowable that is a function of the mortar strength m o, i.e. 1,67 q.

The test results with type S mortar are normalized by @ for the statistical evaluation that follows and this normalized value is used to justify the specified allowable.

16.3 Evaluation of Test Results l

The results from several monotonic tests on the tensile strength of mortar l

l normal to the bed joint in solid multi-wythe concrete masonry walls form j

the basis of the statistical analysis presented herein. In all, data from l

6 tests were available. Involving one mortar type, namely type S. The test data was normalized prior to the statistical evaluation by dividing the calculated modulus of rupture by the square root of the actual mortar strength present in the test specimen. That leaves the constant C. of the equation:

r-Cag

. This constant is the subject of the statistical analysis that follows herein.

The test values of the modulus of rupture were all modified prior to normalization by reducing the calculated modulus of rupture by the actual compression stress acting on the eventual fallure plane during the gests.

The test results thus reduced and subsequently normafized 6 in all, were evaluated statistically using the Gamma distribution. A full description of 44 l -- -- . - - _ _ _ - . - _ _. . _ _ _ _ _ _

the details of this analysis. and the reasons for selecting this particular distribution. Is given in Section 12 of this response. A summary of the results of this statistical analysis is given in Table 16-1 and Table 16-2 below.

TABl.E 16-1 Results of Statistical Analysis of Solid Masonry Data l

{ Sample Size 6 Sample Mean (m) 2.849 Standard Deviation (s) 0.452 Coefficient of Variation 15.9%

95% Confidence Intervals:

On (m) 2.375 4 m 4 3.323 On (m-s) 1.923 4 m-s 4 2.871

• On (m-2s) 1.471 4 m-2s 4 2.419 On (m-3s) 1.019 4 m-3s 4 1.967 The test results of Table 16-1 were then compared with the criteria specified factor, multiplying the @. for stresses normal to the bed joint, namely 1.0 and 1.67 for OBE and SSE events respectively.

• Probabilities that the criteria specified factors for /ffg would exceed the C factor based on the test results were calculated under two assumptions; firstly, that the population mean was equal to the sample mean, and secondly, that it was at the lower end of the 95% confidence interval.

Finally. safety factor based on the 95% confidence Interval for the mean were calculated for the @ factors.

These results are presented in Table 16-2.

e 45

4 ,

i' i

TABLE 16-2 i 1

Exceedance Probabilities and Safety , Factors l l OBE l SSE Probabilities of Exceedance KEY A 0.000027 0.0046 KEY B 0.00117 0.0598 Range of Safety Factors on Mean 2.38 4 SF 4 3.32 1.42 4 SF 4 1.99 __

NOTE:

(1) KEY A in the table above gives the probabilltles of exceedance assuming the population mean equals the sample mean.

(2) KEY B gives the probabilities of exceedance assuming the population mean is at the lower end of the 95%

confidence interval.

Discussion of Results 16.4 The key results for the confidence intervals are plotted in Figure 16-1, together with the OBE and SSE q factors from the reevaluation criteria. The confidence intervals for the data reflect the small sample size. It is seen that the OBE factor.1.0. lies below the "mean minus three standard deviations

• confidence interval whereas the SSE factor,1.67.1:es within the "mean minus three standard deviations

• confidence Interval, it can be stated that criteria specified allowable stresses will exceed the actual tensile strength of the mortar normal to the bed joint in solid masonry walls about 27 times in 1.000.000 for OBE events and about 46 times in 10.000 for SSE events if the population mean strength is taken at the center of the 95% confidence interval. If one considers the extreme case where the population mean is taken to be at the lower end of its 95% confidence Interval. then these probabilities change to 12 times and 598 times in 10.000 for OBE and SSE events respectively. Based on the extreme nature of this

.- second assumption these probabilities are deemed very satisfactory.

Alternatively. Instead of calculating probabilities of exceedence, one may take the same data and calculate factors of safety based on the mean. If this is done for the OBE events. using the full range of the 95% confidence interval for the population mean, the safety factor lies in the range of '2.38 4 SF 4 3.32. Similarly, for SSE events, the range is 1.42 4 SF 4 1.99.

j 46

16.5 Conclusions in view of the above discussion of results, it is concluded that the criteria specified increase factor of 1.67 for tensile stresses normal to the bed joint for solid concrete masonry walls for factored loads is reasonable for the re-evaluation of the Three Mlle Island Nuclear Station. Unit 1.

16.6 References.

16.1 'Research Data and Discussion relating to " Specification for the Design and Construction of Loadbearing Concrete Masonry". NCMA,1970.

16.2 Mayes. R.L. Omote. Y., and Clough, R.W., " Cyclic Shear Tests of Masonry Piers, Volume 1 : Test Results". EERC Report No. 76-8, May,1976.

G m

- 0 47

n ~

1 I

. I I

2 I I I I 1

E O ,' e$

o . .

1 I

• i I

$ $$o k l l 95% CONFIDENCE INTERVALS

'a i I 8

o

~s e .

t I

i i e i i i O.5 1.0 1.5 2.0 2.6 3.0 f, Re-evaluation Criteria 1.0 1.67

, FIGURE 16-1 CONFIDENd INTERVALS FOR POPULATION STATI8 TICS SOLID CONCRi!TE MASONRY - MULTI-WYTHE

a. O 17 ITEM 18 - BOND STRESS QUESTION Item 18 of the NRC request for additional information states:

• With regard to bond stress, the Licensee should discuss and justify an increase of 33-1/3% for factored loads."

RESPONSE

The specified increase of 1.33 for reinforcement bond for factored loads is in accordance with the 1979 UBC. Section 2303(d). This section is partly in reference to Table 24-H and items 9 and 10 of that table specify the same allowable values for reinforcement bono at a working stress level as does the

" Criteria for the Ee-evaluation of Concrete Masonry Walls - Three Mlle Island Nuclear Station. Unit l'. Furthermore, the UBC permits an increase of 1.33 in this allowable stress for seismic loads. Thus the specified increase of 1.33 for factored loads is in accordance with the 1979 UBC.

17.1 Reference 17.1 Uniform Building Code.1979ed., international Conference of Building Officials. Whittler. California.1979.

i l,

t O

49

18 ITEM 19 - WALL AS-14 l I

QUESTION Item 19 of the NRC request for additional information states:

" Indicate the Intended action to evaluate wall AB-14.*

RESPONSE

This item was addressed in a letter from GPUN to NRC. 5211-82-085. dated April 15.1982.

t I

f

- 0 50

• d s l 19 ITEM 20 - SCHEDULE FOR MODIFICATIONS' QUESTON ltem 20 of the NRC request for additional Information states:

" Provide the schedule for the proposed modification specified in Reference S."

l RESPONSE i

This item was addressed in a letter from GPUN to NRC. 5211-82-085. dated April 15,1982.

f.

A 9

I I

t i

51

, l

..