ML16341G280
| ML16341G280 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 09/18/1991 |
| From: | Zimmerman R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | Shiffer J PACIFIC GAS & ELECTRIC CO. |
| References | |
| NUDOCS 9109270054 | |
| Download: ML16341G280 (18) | |
Text
qI ~ ~~CO~
0p gt 0
-~ ~a k*py%
UNITED STATES NUCLEAR REGULATORY COMNllSSION REGION V 1450 MARIALANE,SUITE 210 WALNUTCREEK, CALIFORNIA94596 September 18, 1991 Docket Nos. 50-275, 50-323 Pacific Gas 8 Electric Company 77 Beale Street, Room 1451 San Francisco, Cali fornia 94106 Attention:
Nr. J.
D. Shiffer, Senior Vice President and General Manager Nuclear Power Generation Business Unit
SUBJECT:
Diablo Canyon Units 1
8 2 Block Walls
Dear Sir:
Enclosed for your information is a copy of the NRC Safety Evaluation Report regarding the masonry walls at Diablo Canyon.
If you have any questions regarding the enclosed report, please contact Rr. Harry Rood at (301) 492-1352.
No response to this letter is required.
In accordance with 10 CFR 2.790 of the NRC's "Rules of Practice,"
a copy of this letter and the enclosure will be placed in the NRC Public Document Room.
Sincerely, HIP R.
P immerman, Director Div on of Reactor Safety and Projects
Enclosure:
As Stated cc w/enclosure:
J.
A. Sexton,.
PGSE J.
D. Townsend, PGFE (Diablo Canyon)
R.
F. Locke, PGKE D. A. Taggart, PGSE (Diablo Canyon)
R. Weinberg, News Services, PGSE (Diablo Canyon)
T. L. Grebel, PGKE (Diablo Canyon)
State of California Pi09270054 9i09i8 PDR ADOCK 05000275 P
Fp
I
bcc w/enclosure:
J.
Lieberman, OE J.
- Goldberg, OGC J.
- Partlow, NRR H.
Rood NRR M. HcNu/ty Project Inspector Resident Inspector Allegation File RV-A-89-0043 Greg Cook B. Faulkenberry J. Hartin J. p~ A/lk w I/~ pRk
~, ekA.
f <fg i )
pp j))
SRi c ards: jb R2immerman HBlume 09/P~/91 09//3/91 09/ff'91 AJohn on 09/I~/91 eMp w~
Q, P~~
Q//7/9/
ES /
NO
] YES /
NO
]
ES /
NO
] YES /
NO
]
[
E
/
NO
]
]
/
NO
]
SAFETY EVALUATION REPORT DIABLO CANYON NASONRY WALLS BACKGROUND The NRC staff's initial review of safety-related masonry walls at Diablo Canyon consisted of a detailed evaluation of the submittals made by the Pacific Gas and Electric Company (PG&E, or the licensee) in response to NRC
( IE)Bulletin 80-11, "Masonry Mall Design."
By letter dated November 4, 1986 (Reference 1),
the NRC staff transmitted its evaluation to PG&E, including a technical evaluation report (TER-C5506-648) describing the review of masonry walls at Diablo Canyon by staff consultants.
The November 4, 1986 letter stated, in
- part, "Based on our review we have concluded that the energy balance technique as applied to the masonry walls at the Diablo Canyon Plant is acceptable.
On this basis we conclude that the masonry walls are appropriately qualified."
On the basis of that conclusion, the staff considered its review of masonry walls at Diablo Canyon to be complete.
However, in 1989, the NRC staff became aware of an allegation regarding the adequacy of the seismic design of the masonry walls at Diablo Canyon.
On October 26,
- 1989, the NRC staff initiated a conference call with the licensee regarding the allegation.
In this call the staff requested additional information related to the Diablo Canyon masonry walls.
The licensee submitted the requested information on November 10, 1989 (Reference 2).
The staff review of the masonry wall issue included two other requests for additional information on Narch 20, 1990 (Reference
- 4) and April 6, 1990 (Reference 5), as well as meetings with the licensee on June 6, 1990 (Reference 7),
November 27, 1990 (Reference 10),
and January 31, 1991 (Reference 12).
The following is a
description of the staff's detailed evaluation and resolution of the allegation.
EVALUATION After reviewing the information contained in PG&E's letter of November 10,
- 1989, as well as other information on the Diablo Canyon docket relating to masonry walls, the NRC staff identified two issues of concern regarding the adequacy of the walls.
These are (1) the adequacy of the existing top connections between the masonry walls and the slabs or beams
- above, and (2) the acceptability of the use of the energy balance technique (EBT) in conjunction with an allowable displacement ductility ratio (U) of 5, for qualification of masonry walls at Diablo Canyon.
The allowable displacement ductility ratio is the ratio of the allowable displacement of a structure behaving in an inelastic state to the displacement of the structure when the inelastic behavior is just initiated.
Recognizing that both the top connections and the EBT had been previously approved by the NRC staff, the staff held meetings, including extensive discussions within the staff and its consultants regarding the merits and significance of the new findings.
In addition, an outside expert was retained to independently review the. safety significance of these two issues.
The staff consultant's review report (Peference 3} also questioned the adequacy of these
two aspects of the licensee's design.
A total of nine questions had been raised by the staff and its consultant, and were transmitted to the licensee by letters of March 20, 1990 (Reference
- 4) and April 6, 1990 (Reference 5).
It was pointed out in those letters that the lack of ductility and redundancy at the wall top connections and the lack of an engineering similitude study regarding the data (which were obtained from three proprietary wall tests) used in benchmarking the EBT for Diablo Canyon were major concerns.
The licensee submitted its response to the NRC questions on May 21, 1990 (Reference 6).
A meeting between the licensee and its consultants, and the NRC and its consultant was held on June 6, 1990 to discuss the licensee's responses to the NRC questions.
The meeting is summarized in Reference 7.
As a result of the meeting, the licensee submitted additional information on July 13, 1990 (Reference 8).
After reviewing the submittals, the NRC staff and its consultant independently concluded that (I) the top connection details are not acceptable due to the lack of ductility and redundancy and (2) the licensee's method of calculating wall shear reactions at the clip angles may be unconservative.
(The NRC consultant was asked to review only the adequacy of the top connection of masonry walls.)
The licensee used the yield bending moment of the wall cross-section to infer shear reactions.
This formulation suggests that a wall of a certain cross-section would generate a small end wall shear reaction if a small amount of reinforcement is used (hence a low yield bending moment) and is based on shear capacity rather than demand.
The staff finds that this concept and the consequent assumption are inappropriate and unconservative.
Both the NRC staff and its consultant recommend that the wall weight times the average acceleration should be used to determine end wall shear reactions.
At
- present, the end wall shear reaction is transferrea to the top slab or beam through clip angles bearing against dry-pack (unreinforced concrete) poured on top of the masonry walls.
The staff finds it unacceptable to rely on the brittle dry-pack material alone to transfer the end wall shear that would result from the Diablo Canyon reevaluation basis (Hosgri) earthquake.
- Further, the lack of material test records and placement inspection records for the dry-pack makes prediction of its strength unreliable.
Therefore, the staff concludes that the walls should be structurally upgraded to provide a ductile and redundant load path to transfer the end wall shear from the walls to the top slab or beam.
The inadequacy of the dry-pack top connection and the need to upgrade this connection is described in detail by the NRC consultant in Reference 9.
The staff has also re-reviewed the EBT criteria submitted by the licensee.
The staff now concludes that the.EBT for Diablo Canyon masonry walls was established by benchmarking against data from three proprietary wall tests and was applied directly to Diablo Canyon without adequate recognition of major differences between the test walls and Diablo Canyon walls.
The major differences are in physical'dimensions, boundary conditions, and loading conditions.
~ 3 The test walls were about twice as high as the Diablo Canyon walls and the test walls were ungrouted, except for the cells containing reinforcing steel, whereas the Diablo Canyon walls were fully grouted.
The similitude analysis results submitted by the licensee indicate that the U value for the Diablo Canyon walls should be adjusted upward by a factor of four, relative to the test wall, to account for the fact that the Diablo Canyon wall is fully grouted aramet and is half the height of the ungrouted test wall assuming that th th p
ers such as rebar characteristics and concrete block strengths are eo er sho equal.
This large change in U value is greater than is reasonabl d
e, an routin ws that the EBT is overly sensitive to variations in wall heights d
g
- g. It indicates that the U values derived from the test walls may not an in be directly applicable to Diablo Canyon walls due to major physical differences.
those of th The boundary conditions of the test walls also differed substant'l f
e Diablo Canyon walls.
The test walls had steel reinforcing dowels s
n ia y
rom inserted into a concrete
- base, which provided a nearly fixed-end condition, while the connection between the Diablo Canyon wall and the bottom slab was designed as a hinge.
A wall with a hinge connection at the bottom not only behaves differently than a wall with a fixed support at the bottom, but also is weaker.
In the wall tests, the top actuator, which transmitted horizontal motion to the wall simulating motion induced by earthquakes, was attached to a
steel member that was positively. anchored to the op portion of the test walls.
- herefore, the top actuator not only acted as a loading device but also served as a support which stabilized the wall.
In contrast, the Diablo Canyon walls are connected to the top beam by tw 1'
e
, one on each side of the wall, bearing against dry-pack (uiireinforced 0 c ip concrete),
which was poured on top uf the masonry wall.
In this configuration, the enc wall shear reaction due to the horizontal component of ground motion is transferred through the dry-pack to the clip angles.
Also, the vertical component of ground motion will induce a tensile force in the dry-pack due to riction between the clip angles and the dry-pack.
The dry-pack is a brittle material and will fail once its peak elastic shear or tensile stress is reached.
Therefore, the test wall top connection is ductile, but the Diablo Canyon wall top connection is brittle.
A ductile connection can sustain large deformations and thus absorb a large amount of energy during strong earthquakes, while a brittle connection may fail catastrophically during earthquakes.
Consequently, the test wall data cannot be applied to the Diablo Canyon walls unless the top wall connection at Diablo Canyon has been upgraded to behave in a ductile manner.
In addition, the loading conditions for the test walls differed from the actual earthquake loading conditions for the Diablo Canyon walls.
The test walls were loaded only by horizontal actuators placed perpendicular to the walls at the top and bottom of the walls.
The tests did not include any vertical seismic
- motions, and the effect of vertical motion on the EBT has not been addressed b
the licensee.
een a
resse y
't
Based on the considerations discussed
- above, the staff concludes that:
(1) the wall test data did not adequately substantiate the maximum ductility ratio (U=5) for the test walls; (2) the U value is -very sensitive to variations in wall heights and grouting, and-the Diablo Canyon walls are quite different from the test walls in these areas; (3) the wall top connections are ductile for the test walls, but brittle for the Diablo Canyon walls; and (4) the effect of the vertical ground motion on the EBT has not been properl addressed.
roper y The staff concludes that these deficiencies make the EBT, in conjunction with the use of the U=5 criterion, unacceptable for use with the Diablo Canyon walls.
This conclusion was discussed with the licensee in a meeting held on November 27, 1990 (Reference 12).
In a letter dated December 14, 1990 (Reference 11), the licensee stated that it would modify the top connections as well as the lateral bending moment capacity to provide additiona 1 lateral load resistance, proposed a more conservative th method to reevaluate the Diablo Canyon masonry walls and agreed to mod'f e walls if necessary.
On January 31, 1991 the staff met with the licensee to mo i y discuss the method of assigning a priority category to each of the safety-related walls, the schedule for modifying the walls in each priority category for each unit, the details of several optional methods of modifying the walls, and the evaluation criteria to be used to analyze the existing and modified walls.
During the meeting, the staff suggested that the licensee clarify the schedule to indicate that the high-priority modifications would be completed or each unit by the end of the next refueling cycle for that unit.
The staff also suggested some additional evaluation criteria to be used to analyze the walls.
Reference 12 is a
summary of the January 31, 1991 meeting.
As a result of the January 31, 1991 meeting, the licensee submitted a masonry wall reevaluation program in a letter dated February 12, 1991 (Reference 13).
The program may be summarized as follows:
(1) the engineering and modification work for "high" priority Unit 1 walls will be implemented during fuel cycle 5 and will be completed by the end of the fifth refueling outage, which is scheduled for completion in December
- 1992, (2) the engineering and modification work for the remaining
("medium" and "low" priority] Unit 1 walls will be completed by the end of fuel cycle 6, which is scheduled for completion in December
- 1994,
(3) the engineering and modification work for high-oriority Unit 2 walls will be implemented during fuel cycle 5 and will be completed by the end
- 1993, of the fifth refueling outage, which is scheduled for completion in Nay (4) the engineering and modification work for the remaining (medium-and low-priority} Unit 2 walls will be completed by the end of fuel cycle 6, which is scheduled for completion in December
- 1994, (5) a linear elastic analysis coupled with strength design method will be used to evaluate the walls, and (6}
additional evaluation of the walls (by an inelastic analysis th d) ll be erf p
ormed to address the variability of material properties, me o
wi workmanship, and construction tolerances.
On April 4, 1990, the NRC staff inspected the masonry walls at Diablo Canyon believed to be the most susceptib le to earthquakes.
The inspection indicated that there are relatively few safety-related pipes or other safety-related components which could be impacted by the walls should th f '1 h
attern.
On p
em.
n this basis, the staff concludes that immediate modification of the masonry wa11s is not necessary, and finds that the schedule for wall modification proposed by the licensee to be acceptable.
The staff finds the re-analysis method proposed by the licensee (linear elastic analysis coupled with the strength design method) to be acceptable, because the method reflects the state-of-the-art in masonry wall analysis and design.
The licensee states in its description of the proposed method that unreinforced inertia load concrete or masonry sections will not be relied upon to th t d
wi s an seismic loads.
Thi s commitment resolves the staff concern about th d
p nection of masonry walls at Diablo Canyon.
The reanalysis method e
a equacy has also been accepted by the most recent editior. of the Uniform Building Code (Reference
- 14) and is included in the National Earthquake Hazards Reduction Program Provisions (Reference 15).
In order to clarify the proposed method as submitted in the licensee's February 12, 1991 letter, the following two additions should be made:
(I) under masonry flexural design, the words "calculated by linear elastic analysis" should be added to the end of "the moment produced by the applied loads",
and (2} under masonry shear and bearing
- design, add the sentence "Shear and bearing forces shall be calculated based on linear elastic analysis."
The staff does not object to the proposed additional evaluation of the walls to address the variability of material properties, workmanship, and construction tolerances by inelastic analysis; it may provide additional safety insights.
This additional evaluation,
- however, shou ld not be considered to be an alternative to the proposed method of analysis based on linear analyses.
I
CONCLUSION As a result of an alleg'ation, the staff and its consultant re-reviewed the seismic design of the masonry walls at Diablo Canyon.
The staff and its consultant independently concluded that the energy balance technique (EBT),
which had previously been approved for use at Diablo Canyon by the staff and used by the licensee, is not acceptable for the Diablo Canyon masonry walls because the technique has not been appropriate1y benchmarked.
The EBT for Diablo Canyon masonry walls was benchmarked by PGKE against proprietary data from tests of three masonry walls conducted by another licensee.
- However, significant differences exist between the Diablo Canyon masonry walls and the test walls in physical dimensions, boundary conditions, and loading conditions.
These dif erences render the EBT criteria unacceptable for use with the Diablo Canyon walls.
The staff has inspected the masonry walls at Diablo Canyon believed to be the most susceptible to earthquakes.
The inspection indicated that there are a
relatively few safety-related pipes or other safety-related components which could be impacted by the walls, should they fail in the worst pattern.
On this
- basis, the staff concludes that immediate modification of the masonry walls is not necessary.
However, the staff has requested that the licensee reevaluate the walIs using acceptable methods and modify them, if necessary, without undue
- delay, in order to demonstrate full compliance with the requirements of General Design Criterion 2 (protection against natural phenomena) of Appendix A to 10 CFR 50.
The staff finds that the schedule for wall modification proposed by the licensee is acceptable.
The licensee proposes to use a linear-elastic-analysis approach coupled with the strength-design method to reevaluate the masonry walls at Diablo Canyon.
The licensee has indicated that most of the walls will require physical modifications and has presented an acceptable modification schedule to the staff.
The staff finds that the proposed analysis method reflects the state-of-the-art in masonry wall analysis and design and, therefore, is acceptable.
The staff also finds acceptable the proposed schedule for wall modifications, because it will result in modification of all the walls in a reasonably short time, with the most critical walls being modified first.
On the basis of the licensee's commitment to implement the masonry wall reevaluation program and complete necessary wall modifications on an acceptable
- schedule, the staff considers the safety aspects of allegation RV-89-A-0043 to be resolved.
REFERENCES l.
- Varga, S. A.,
NRC to Shiffer J. D.,
PGKE, "License Conditions For Masonry Walls," November 4, 1986.
2.
Shiffer, J. D.,
PGKE to NRC Document Control Desk, "Additional Information On Masonry Wall guaIification," November 10, 1989.
3.
- Atkinson, R. H., "Review of The SONGS-1 Masonry Test Program and Applicability to Diablo Canyon Masonry Walls," Atkinson Noland 5
Associates, March 1990.
4.
5.
7.
8.
Rood, H.,
NRC to Shiffer, J. D.,
PGKE, "Request For Additional Information Relating To NRC Staff Review Of Masonry Wall Issues At Diablo Canyon (TAC Nos.
74142 and 74143)," March 20, 1990.
Rood, H.,
NRC to Shiffer, J. D., PG5E, "Request For Additional Information Relating to NRC Staff Review Of Masonry Wall Issues At Diablo Canyon (TAC Nos.
74142 and 74143)," April 6, 1990.
Shiffer, J. D.,
PGSE to NRC Document Control Desk, "Additional Information Relating to Masonry Walls," May 21, 1990.
Rcod, H.,
- NPC, "Summary of June 6, 1990 Public Meeting On Seismic Design 27, 1990.
of Masonry Walls At Diablo Canyon (TAC Numbers 74142 arid 74143)
" J ri, une
- Brand, D. A.,
PGIEE to NRC Document Control Desk, "Additional Information On Masonry Walls," July 13, 1990.
9.
- Atkinson, R. H., "Evaluation Of Masonry Wall Top Connection Diablo Canyon Nuclear Power Plant," Atkinson-Noland 5 Associates, November 1990.
10.
Rood, H.,
- NRC, "Summary of November 27, 1990 Public Meeting on Seismic Design of Masonry Walls at Diablo Canyon (TAC Numbers 74142 and 74143),"
December 7,
1990.
ll.
Shiffer, J. D.,
PGRE to NRC Document Control Desk, "Reevaluation of Diablo Canyon masonry Walls," December 14, 1990.
12.
Rood, H.,
- NRC, "Summary of January 31, 1991 Public Meeting On Seismic Design of Masonry Walls at Diablo Canyon (TAC Nos.
74142 and 74143),"
February 21, 1991.
13.
Shiffer, J.
D.
PGSE to NRC Document Control Desk, "Masonry Wall Reevaluation Program,"
February 12, 1991.
14.
"Uniform Building Code" by International Conference of Buildin Off 1
May 1, 1988.
ing
>cia s, 15.
"National Earthquake Hazards Reduction Program Recommended Provisions for the Development of Seismic Regulations for New Buildings" by the Building Seismic Safety Council for the Federal Emergency Management
- Agency, 1988.