View source for 3F1185-23, Responds to 850826 Request for Addl Info Re IE Bulletin 80-11 Re Masonry Wall Design.Justification for Fixed Boundary Conditions Along Horizontal Edges in Walls 1,2 & 3 Unnecessary

{{Adams
| number = ML20137H448
| issue date = 11/25/1985
| title = Responds to 850826 Request for Addl Info Re IE Bulletin 80-11 Re Masonry Wall Design.Justification for Fixed Boundary Conditions Along Horizontal Edges in Walls 1,2 & 3 Unnecessary
| author name = Westafer G
| author affiliation = FLORIDA POWER CORP.
| addressee name = Stolz J
| addressee affiliation = NRC OFFICE OF NUCLEAR REACTOR REGULATION (NRR)
| docket = 05000302
| license number = 
| contact person = 
| document report number = 3F1185-23, 821125, IEB-80-11, NUDOCS 8512020394
| document type = CORRESPONDENCE-LETTERS, INCOMING CORRESPONDENCE, UTILITY TO NRC
| page count = 6
}}

=Text=
{{#Wiki_filter:'N O
:           n N
   *'I I.&
o*..eoe  ce
                                                                                  -e Power CORPO A AT 6ON November 25,1985 3Fil85-23 Director of Nuclear Reactor Regulation Attention: Mr. John F. Stolz, Chief Operating Reactors Branch #4 Division of Licensing U.S. Nuclear Regulatory Commission Washington, DC 20555

==Subject:==
Crystal River Unit 3 Docket No. 50-302 Operating License No. DPR-72 IE Bulletin 80-11, Masonry Wall Design Request For AdditionalInformation
                                                                                           ~ '

==Dear Sir:==
Your {{letter dated|date=August 26, 1985|text=letter dated August 26, 1985}}, requested additional information in order to complete your review of Florida Power Corporation's (FPC's) responses dated November 7 and 17,1980 to IE Bulletin 80-11. In a {{letter dated|date=September 24, 1985|text=letter dated September 24, 1985}}, FPC indicated this information would be provided by November 27,1985. By this letter, FPC fulfills that commitment. Review of the identified concerns and their incorporation into the analysis where applicable has not changed any initial conclusions on the capability of the subject walls to withstand the postulated loadings.
Sincerely, n
o      /
G. R. Westafer Manager, Nuclear Operations Licensing and Fuel Management GMF/feb 8512O20394 e51125 PDR Attachment                   G     ADOCK 05000302
,                                                              PDR i
;          xc:     Dr. 3. Nelson Grace
!                  Regional Administrator, Region II Office of Inspection & Enforcement U.S. Nuclear Regulatory Commission dl    i 101 Marietta Street N.W., Suite 2900                                                  JP l                  Atlanta, GA 30323                                                                         '
l h
l GENERAL OFFICE 3201 TNrty fourth Street South e P.O. Box 14042, St. Petersburg, Florida 33733 e 813-8865151

ITEM 1:      According to Section 1.2.3 of Reference 1, a portion of wall I has been removed to eliminate the wall's possible effect on safety-related equipment (see Figure 4, Ref.1). Explain how this modification removes the safety concern of wall collapse on equipment near wall 1.
RESPONSE 1:  Florida Power Nuclear Engineerirg reviewed the effects of the collapse of wall 1 on safety-related equipment. The only safety-related equipment which could be adversely affccted by the collapse of the wall was the control complex ventilation system                      '
main supply air duct to the lower elevations. The collapse of the wall could damage the vertical section of this duct where it turns and runs down the west wall of the control complex. Damage to this portion of the duct would cut off cooling air to critical electrical components lesding to higher than rated temperatures for continuous duty and subsequent failure of the equipment.
Two possible resolutions to the problem were available. The first i
was to stiffen the wall using a complex arrangement of an I beam with through-bolting and attachments to the floor and ceiling. The other available option was to remove the portion of the wall which would Impact the duct if it fell. Nuclear Engineering evaluated these two options and determined that the removal of the wall section would be the quickest, most effective resolution.
G/CI concurred with this decision stating that the removal of a
,                portion of wall I was an acceptable alternative to reinforcing the wall to withstand the postulated seismic accelerations.
ITEM 2:      According to Reference 1, walls 2 and 3 exceed ACI 531-79 allowable stresses by 50% to 60%; however, according to the cover letter of Reference 1, a recent inspection has shown that the collapse of walls 2 and 3 would not have an adverse effect on safety-related equipment in the area. Indicate the recent findings that have permitted these walls to be considered differently than in the original evaluatio- (Reference 2) which presented wall 2 and 3 as safety-related. ... so, provide the documentation of the evaluation of these walls that, according to the cover letter of Reference 1, was due by April 30, 1984 but has not yet been provided.
RESPONSE 2:  Florida Power Nuclear Engineering conducted a review of walls 2 and 3 to determine the exact locations'of all safety-related equipment in their vicinity. This review concluded that the only safety-related equipment located in the vicinity of the walls was electrical conduits located d the top of the walls. At the request of Florida Pouer Corporation, G/CI evaluated the potential impact of postulated masonry wall failure on these electrical conduits. The conduits in question penetrated the walls at the boundary between the top of the masonry walls and the concrete floor system above. G/CI's evaluation concluded that as long as the conduits were not grouted into the masonry walls or physically attached to the masonry walls, the conduits would not be damaged
               , by a postulated wall failure. This was bx3ed on probable failure modes of the wall being:

a)    Rotation of the wall panel as a unit about the base joint; b)    Brittle rupture of the masonry panel at boundary limits and internal areas of maximum flexural moments.
Either failure mode would result in the masonry walls moving away from the conduits in question. Impact of masonry wall           '
segments on the conduits is not possible due to their location at the top boundary of the walls. Slight modifications to the walls were accomplished to ensure that they were not physically in contact with the conduits. Therefore, >o damage to the conduits was postulated.
FPC, in our {{letter dated|date=February 29, 1984|text=February 29, 1984 letter}}, did not commit to an evaluation of the subject walls, but rather to evaluate alternatives to determine the best controls to assure that these walls will not adversely affect any safety-related equipment which may be placed in this area in the future. FPC submitted tha results of that evaluation in the {{letter dated|date=April 30, 1984|text=April 30, 1984 letter}} (3F0484-2 31 ITEM 3:     The response to item 4 in Reference 1 indicates that for the five control complex walls, a " fixed" boundary condition is assumed for mortared joints at masonry wall bases, top edges along a support slab, and side edges where the corner is integral with another masonry wall.         No dowels, anchors, or other boundary reinforcements are present. Provide the justific~ati~on' for the assumption of fixed conditions 'for mortared joints by examining a typical connection in detail.           Indicate the effect of joint flexibility on calculated results.
RESPONSE 3: The five control complex walls were modeled as completely free along the top edges as no supporting slab exists at these wall locations. The instrument room walls support a concrete slab which was cast-in-place on top of the walls, and have a fixed boundary condition modeled along the top.
Justification for fixed boundary conditions along horizontal edges in walls 1, 2, and 3 is unnecessary as these walls are already overstressed. Since overstresses occur at regions both near and away from the clamped edges, relaxing the fixity would only produce a more severe stress condition away from the edges.
For wall 4, the base fixity is justified since dead load compression due to wall self-weight exceeds seismic bending. Due to the running bond arrangement at the corners, some fixity will exist along the sides. To evaluate alternate boundary conditions, an approximate comparison can be made between a fixed and hinged side condition. Using plate solutions for a uniformly loaded plate having a length-to-height ratio of one, and being free along the top and fixed along the remaining three sides, a maximum positive moment of .043 pb2 results, where p is the applied uniform load and b is the wall height. If two plate edges are hinged, the maximum positive moment becomes .094 pb2.                  Since the maximum stress from the fixed boundary computer model is about

one-third the allowable, multiplying it by the hinged / fixed moment ratio will stil: not cause an overstress.
(.094 pb2/.043 pb 2X5.8) - 2.5 = 10.2 psi <(1.3) (13.7 psi)
For wall 5, seismic flexure exceeds dead load compression by 5 psi. This occurs near the center of the base only. Since a 13.7 psi     '
tensils allowable is permitted, mortar bond will resist flexure in this region, and some fixity will be achieved. Full fixity will be assured where dead load compression exceeds flexure. Using similar plate solutions as was done for wall 4 with a length-to-height ratio of two, a maximum positive moment of .095.pb 2 results when three sides are fixed, and the top is free. Changing the side conditions to hinged results in a maximum positive moment of 0.117 pb2 . Multiplying the maximum mid-region stress generated in the fixed boundary computer model by the hinged / fixed moment ratio still does not produce an overstress.
(.117 pb2/.095 pb 2X19.5) = 24.0 psl<(1.5X27.4 psi)
Use of the preceding approach results in the maximum edge
,             shears increasing up to 60% for walls 4 and 5.                 However, maximum shear in the actual computer model is less than 15% of the allowable, and will not cause a problem if increased accordingly.
ITEM 4:     With respect to Section 3.1.2 of Reference 1, walls 1,4, and 5 do not meet the ACI minimum requ[rements for thickness, so the ACI 531-79 formula, Fa = 0.255 f'a J-(h/40t)3), for allowable axial compression becomes negative. These walls were qualified using the Southern Standard Building Code (SSBC) compression stress allowable of 70 psi. Also, fixed side boundary conditions in walls 4 and 5 were relaxed to allow for a redistribution of moment, which helped to keep stresses within allowable limits.
Provide and justify the reasons for accepting these walls rs meeting the SGED criteria.
RESPONSE 4: Walls 4 and 5 intersect other masonry walls with running bond at corners and are therefore integral and continuous. This condition permitted the use of fixed boundary conditions at the sides.
These intersecting walls do not provide complete rigidity so that some redistribution will occur. Although not specified in ACI 531-79 or the Southern Standard Building Code, redistribution of negative moments is implied if one inspects the approximate moment formulas provided in Section 9.5.1 of ACI 331-79. These formulas are identical (except for one case) to those formulas presented in Section 8.3.3 of ACI 318-83 for reinforced concrete.
Furthermore, Section 8.4 of ACI 318-83 states that except where approximate values of momer.ts are used, negative moments calculated by elastic theory at supports of continuous flexural members may be redistributed up to 20%.
The analysis of wall 4 which assumed fixity on the sides and base, showed maximum stresses in both the positive and negative moment regions to be less than one-third of the ACI 331-79 tension allowable. With this margin, a 20% redistribution will still not exceed the code limits.

l Results of the analysis of wall 5 are based upon fixed sides and a fixed base and show that stresses due to the maximum positive bending moment (near wall center region) are less than 50% of the ACI 531-79 code allowable. The maximum negative moment                         ,
region occurs near the top left and right sides of the wall. For the OBE condition, the maximum stress exceeds the 27.4 psi allowable by 10% for a fixed wall condition. The SSE allowable of 41.1 psi is exceeded by 21%. Using a 20% redistribution will lower edge stresses such that the OBE condition will be within limits and the SSE condition within 1% of the allowable. The corresponding increase in the positive moment region will still result in stresses which remain within acceptable limits.
Shears are well below the code allowab!e for both walls 4 and 5 (less than 25% of the allowable) and will still remain within code limits if boundary conditions are relaxed (see previous item).
The ACI 531-79 formula Fa = 0.225 f'm (1 - (h/40t)3) is for allowable compressive stress of a load-bearing masonry wall. This formula considers the P-A effect of an axially loaded (end loaded) column and has a factor of safety of approximately 9.             As mentioned in 3.1.1.4 of the February 10, 1984 Masonry Wall                     ,
Report, using the theoretical buckling equation of a vertical cantilever wall under uniformly distributed self-weight with a                 ;
safety factor of 9 results in an allowable stress at the base of an              i unreinforced, non-load bearing wall of approximately 90 psi (for walls I,4, and 5). Therefe e, the Southern Standardhilding Code allowable of 70 psi appeau reasonable.                                           ,
i ITEM 5:     With respect to the seismic analysis of the instrument room walls                )
in the turbine building, Section 3.2 in Reference 1 Indicates that the effect of the first mode of vibration was increased by 5% to account for high modes. Indicate why these ivalls were not checked with an eight mode analysis as were the five walls in the control complex. Indicate whether there are other differences in the criteria between the control complex walls and the instrument room walls. Also, Figures 22 and 23 show " knife edge" boundary conditions for walls in the turbine building. Indicate whether these conditions were changed to " fixed" or " free" as in the control complex walls, and specify the physical boundary connections. The staff does not accept the use of arching or wedging action at boundary connections in qualifying masonry walls (the staff position is attached).
RESPONSE 5: The instrument room walls were done in a separate analysis prior to the control complex wall analysis. The " knife-edge" condition represents a pinned condition, with free rotation permitted about the plane of the wall. Because of the geometry of the structure, the lowest frequency was determined to be 36 cycles /second, high enough to justify additional mode determination as unnecessary.
The interaction equation fa/Fa + im/Fm Produces a maximum value of only 0.23. Modification of the boundary conditions to fix               l the base and free the wall edges will not cause sufficient change to result in overstress, since such a large margin exists between

i computer model stresses and allowable stresses. The geometry of the structure is such that the walls and supported slab behave as a unit, reducing the importance of side boundary conditions.

==References:==

Item numbers refer to items enclosed in the {{letter dated|date=August 26, 1985|text=August 26, 1985 letter}} from John F.
Stolz (NRC) to Walter S. Wilgus (FPC), Docket No. 50-302.
Wall numbers, figures, and analytical discussions upon which the items are based are presented in the Gilbert Associates Report " Masonry Wall Re evaluation -
Program Including Response to the NRC Questions", dated February 10, 1984 (identified as Reference 1 in the NRC letter referenced above).
Plate solutions are taken from the U.S. Dept. of the Interior Bureau of Reclamation Engineering Monograph No. 27, " Moments and Reactions for Rectangular Plates", July 1970.
O e
                                                                                      ,-}}