ML20215B688
| ML20215B688 | |
| Person / Time | |
|---|---|
| Site: | 05000000, Diablo Canyon |
| Issue date: | 07/01/1985 |
| From: | Bosnak R Office of Nuclear Reactor Regulation |
| To: | Novak T Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML082840462 | List:
|
| References | |
| FOIA-86-197 NUDOCS 8610080334 | |
| Download: ML20215B688 (15) | |
Text
f
- pn merg o UNITED STATES E
NUCLEAR REGULATORY COMMISSION
,g h.
WASHINGTON, D. C. 20555 i
/
JUL 011985 MEMORANDUM FOR: Thomas M. Novak, Assistant Director for Licensing Division of Licensing FROM:
Robert Bosnak, Acting Assistant Director for Components and Structures Engineering Division of Engineering
SUBJECT:
SSER INPUT FOR DIABLO CANYON UNIT 2 Enclosed is the Structural and Geotechnical Engineering Branch SSER input for Diablo Canyon Unit 2 full power license.
The Unit 2 review by SGEB is complete and the full power license can be issued with a required licensing condition. This condition as stated in the enclosure (page 13) requires PG&E to provide a confirmatory seismic analysis of the Unit 2 pipeway structure for the design and double design earthquakes.
This analysis should be provided to the staff before the first refueling outage.
Principal contributors to this input were P. T. Kuo, Harold Polk, and N. Chokshi and our consultants, Brookhaven National Laboratory.
Roberv Sosnak, Acting Assistant Director for Components and Structures Engineering Division of Engineering
Enclosure:
As stated cc:
J. Knight G. Knighton G. Lear H. Schierling P. Kuo H. Polk N. Chokshi D'
6 4
0610090334 860923 PDR FOIA HOLMES86-197 PDR
o STRUCTURAL AND GE0 TECHNICAL ENGINEERING BRANCH SSER INPUT FOR DIABLO CANYON UNIT 2 1.0 TURBINE BUILDING
1.1 INTRODUCTION
The floor slab at elevation 140 in the Turbine Building is a 12-inch thick reinforced concrete slab with a large central cutout for the turbine pedestal and a equipment elevator cutout at column lines C and 19. The slab is supported on a grillage of steel beams.
The diaphragm action of the slab provides torsional stiffness to the turbine building at elevation 140. The diaphragm acts as a one bay Vierendel Truss with the E-W shear walls at column lines 19 and 35 providing the primary support. A critical section of the diaphragm (for in plane forces) occurs along the N-S column line C between the E-W lines 19 and 21 because of the cutout for the equipment elevator shaft at the north end of the slab at column lines C and 19.
It should be noted that the condition is not the same for the Unit 1 turbine building because of the extra bays in the Unit 1 turbine building and a smaller equipment elevator shaft cutout in the elevation 140 floor slab.
During PG&E's stress evaluation of the turbine building for the Hosgri event, the shear stresses in this critical region were found to slightly exceed the criteria (ACI 318-73, Section 11.16). The loads used for this stress evaluation were developed from the three-dimensional model of the entire turbine building. The floor diaphragm was modeled with relatively large, plane strain, elastic elements in this analysis. Concrete cracking was not considered when the diaphragm loads were determined.
PG8E prepared a more detailed model of the floor at elevation 140 to obtain a better estimate of the loads. A finer finite element grid was used to model the concrete floor slab and separate elements were used to model the embedded steel beams and reinforcing steel. The nonlinear computer code FINEL was used so that the cracking of the concrete along planes of principal tensile stresses could be considered. Peak floor nodal accelerations were obtained from the three dimensional dynamic model of the turbine building.
t l *
! These accelerations were used by the computer program FINEL to compute nodal forces. These accelerations were adjusted to produce forces in the i
uncracked FINEL model at the floor slab critical location that matched the i
peak forces computed in the three dimensional model. These adjusted l
accelerations were also used as input to the cracked floor slab FINEL model so that the effect of concrete cracking on the magnitude of the loads on the critical section could be evaluated.
The forces resulting from the FINEL model when concrete cracking was l
l considered were smaller than the forces determined in the uncracked model.
The stress limits of the concrete code (ACI 318-73, Section 11.16) were satisfied using the FINEL cracked concrete model.
The floor response spectra used in the equipment evaluation was based cn l
the uncracked concrete floor slab. The FINEL solution indicates that some l
concrete cracking occurs, but this cracking is restricted to a small region l
of the slab. Since the cracking is localized, one would not expect the l
overall stiffness of the slab to change enough to cause significant changes in l
the floor response spectra.
l 1.2 SCOPE OF REVIEW l
The staff has reviewed the three-dimensional model of the turbine building (PG8E Calculation 64T283) and found to be acceptable as reported in SSER 29.
The shear problem in the elevation 140 floor slab was first discussed with l
PG8E during an audit on January 15-17, 1985. At this time very preliminary calculations were available indicating that a problem might exist. PG&E presented a plan of actinn that would address this matter in detail. Progress
^
on the PG&E work was reviewed at an audit held at Brookhaven National Laboratory on April 16, 1985. While PG&E's calculations were not complete at this time, the preliminary assessment by PG&E was that the problem could be satisfactorily resolved.
A detailed audit of P3&E's calculation was conducted on May 30-31, 1985, at the PG&E offices in San Francisco. The following items were reviewed during this audit:
r
< 4 (a) PG8E Calculation 65T355 The evaluation of shear stresses at the critical section was made in this calculation. The shear, tensile force and moment at the critical section based on the original three dimensional model are:
2780 kips; 2940 kips and 10,400 kip-feet respectively. Using the ACI 318-73 allowable shear for a concrete strength of 6590 psi (the average of the 60-day test results) and a capacity reduction factor "0" of 0.85 and the actual shear capacity was shown to be 2370 kips.
This capacity is less than the demand of 2780 kips.
The shear, tensile force and moment at the critical sections, based on the FINEL model including concrete cracking is:
1850 kips, 1960 and 3700 kips-feet respectively. The shear capacity using the 60-day strength and a O factor of 0.85 is 2370 kips. The shear capacity is larger than the demand of 1850 kips. The moment and tensile capacity is larger than the demand.
(b)PG&ECALCULATION65T318 The FINEL model was developed in this calculation. From the results of these calculations, it is shown that the concrete cracking develops across the critical section and the cracks generally run in the NW to the SE direction. Stresses in the cracked elements indicate that a tensile stress of about 600 psi (close to the modulus of rupture of the concrete) was tho criterion used to decide when cracking occurred. This indicates that at least for the problem of interest FINEL gave reasonable results.
(c) FINEL MANUAL The FINEL manual was reviewed to assess the adequacy of the Code to handle the problem of interest. The Code employs an iterative solution starting with the uncracked solutions. After eech iteration all elements are reviewed to determine whether or not cracking occurs and if so at what angle. The stiffness matrix is modified for the cracked element and the next iteration performed. Convergence is determined when the changes in the
, solution between one iteration and the next are acceptably small.
Six iterations were used for the turbine building analysis.
The code has been verified by Bechtel using comparisons with both analytical and experimental data. The comparisons for a cantilever beam and a deep panel were particularly applicable to the problem at hand.
1.3 FINDINGS AND CONCLUSIONS Based on the staff's audits and reviews of the extensive information provided by PG&E described above, the staff's findings are as follows:
(a) The shear demand based on the elastic model loads are larger than the ACI capacity based on 60-day concrete strength and a O factor of 0.85.
(b) The FINEL code is applicable to the turbine building floor slab and the results for member loads are realistic. When the more realistic model is used including the effects of cracking, the shear loads are j
reduced resulting in the demand being less than the capacity.
(c) Since cracking is restricted to a small region of the slab, one would not expect the overall stiffness of the slab to change enough to cause significant changes in the original floor response spectra.
Based on the FINEL results taking into consideration load redistribution caused by concrete cracking, the staff concludes that the elevation 140 turbine building floor slab is structurally adequate and the floor response spectra are acceptable.
2.0 PIPEWAY STRUCTURE
2.1 INTRODUCTION
Each unit has a steel space frame pipeway structure attached to the outside of O
F
. the containment shell, turbine building, and auxiliary building between elevations 87 feet and 119 feet. The main function of the pipeway structure is to support the main steam and feedwater lines frora their point of exit from the containment to their entry into the auxiliary building.
The seismic Category I pipeway structure is required to satisfy load combinations involving dead, seismic and pipe rupture restraint loads.
The seismic loads include the Hosgri event, the Design Earthquake (DE) and the Double Design Earthquake (DDE). The seismic analysis for the Hosgri event of the Unit 1 pipeway was performed by Westinghouse whereas the Unit 2 pipeway Hosgri analysis was performed by PG&E, The DE and DDE evaluations for both units were performed by PG&E. The pipeway analysis for the pipe rupture loading condition was performed by PG&E for both units of the plant.
2.2 SCOPE OF REVIEW The staff performed three audits of the pipeway civil / structural calculations:
January 16 & 17, February 28, and May 30 & 31, 1985. The first and the third audits took place in San Francisco, California at the PG&E offices whereas the second audit was conducted in Monroeville, Pennsylvania at the Westinghouse offices. Subsequent to each of the above audits, PG&E supplied follow-up information as requested by the staff.
The staff's evaluation of the pipeway structure was reported in SSER 29 (March 1985). The concerns identified in SSER 29 are summarized below:
a.
modeling of the pipeway structural connections to the auxiliary and turbine buildings.
b.
selection of the seismic input for the pipeway seismic analysis, c.
integration time step used in the seismic analysis for the Hosgri
- event, d.
procedure used to account for accidental torsion.
l
, e.
strength capability of the pipeway structure under action of relative motion between containment-auxiliary-turbine buildings.
f.
Items (d) and (e) above were resolved in SSER 29. The remaining items a, b, c, and f are discussed below.
2.2.1 MODELING 0F PIPEWAY CONNECTION TO AUXILIARY TURBINE BUILDINGS The staff audited the PG&E calculations of the Unit 2 pipeway. structure for the Hosgri event (Calculation File No. 52.10.2). The three dimensional dynamic model used in these evaluations incorporates a centilever beam type idealization of the containment exterior shell. The pipeway structure is represented by an assembly of beam elements which represent structural steel members as well as the main steam and feedwater piping. As described in SSER 29, the staff concluded that the connections of the structure to the auxiliary and turbine buildings were allowed to move freely in the horizontal plane. These beams were connected to the turbine building and auxiliary building in the vertical direction only. Since slotted holes were provided to accomodate the ' structural displacements in the axial direction only of the steel beams that frame into the auxiliary building and the turbine building, the in-plane freedom of the corresponding nodes was questionable. The staff requested PG&E to justify the consistency between the modeling of these connections in the three dimensional pipeway model and the as-built condition at the plant.
PG8E provided the details of the slotted holes in both the auxiliary and turbine buildings.
(DWG No. 443375; SKC-PW-01, Sheet No. I to 4).
According to PGAE, the differential displacements between the pipeway structure and the turbine building are 0.17 and t 0.24 inches in the NS and EW directions, respectively (PG&E transmittal of January 1985). These displacements can be accommodated by the clearance provided by the slotted holes. Therefore the staff finds the turbine building connection modeling is acceptable. Since the pertinent differential displacements at the pipeway frame and the auxiliary building connections were not provided in the PG&E
?
. ]
January 1985 transmittal, the staff requested in SSER 29 that these displace-ments should be provided in order to assure that there is sufficient clearance to accommodate these movements.
(See SSER-29, p. 8-2, Item a).
During the May 1985 audit, the pipeway structure to auxiliary building connect-ion was further investigated by the staff. The audit concentrated on the modeling assumptions used in the three dimensional pipeway structural model for a set of three beams which connect the pipeway structure with the auxiliary building. PG&E stated that the as-built condition at the south end connection of these beams use clip angles attached to the beam web and thus the motion of the pipeway structure is not restrained. Thus the relative displacements between the pipeway structure and the auxiliary building can be accommodated as per as-built condition. This condition, however, is not reflected in the three dimensional pipeway structure model.
In the model, these connections are modeled as the moment type. PG&E stated that this difference is due to a modeling necessity in order to avoid instability in the computer program.
Furthermore, these members do not provide structural support for the pipeway structure. The modeling approximation should not effect the overall results of the pipeway structure seismic evaluation performed for the Hosgri event. There-fore the staff considers the issue of modeling of the pipeway connection to the auxiliary building resolved.
2.2.2 SELECTION OF SEISMIC INPUT FOR THE PIPEWAY STRUCTURE MODEL The pipeway structure of both units are attached to the auxiliary and turbine buildings as well as to the containment shell. Thus, the seismic input associated with the pipeway seismic model is not the same at the various attachment points. Both PG&E (Unit 2 pipeway structure Hosgri evaluation) and Westinghouse (Unit 1 pipeway structure Hosgri evaluation) employed a single input motion in their analyses. This motion was taken to be the Hosgri input at the base of the containment structure.
The staff requested PG&E to justify the selection of the input motion. PG&E provided connection details between the auxiliary building and pipeway structure which showed that no input motion can be transmitted between these two structures. Furthermore, for the same reason. no horizontal input is expected from the turbine building to the pipeway structure. To justify the use m._.
~y
. of the vertical input from the containment structure a response spectra comparison was made which showed that the containment spectra enveloped the corresponding. turbine building spectra at the locations of interest. Thus in this case, the input choice is conservative.
The staff, however, requested additional justification of possible transmission of seismic loads from the auxiliary building through the piping systems (see SSER-29, p. 8-3, Item b). According to the methodology used by both PG&E and Westinghouse, a single input motion was applied to the three dimensional pipeway structure models. These are coupled models of the pipeway structure with the major piping systems. Since support conditions were assigned at the nodes of the piping systems in the auxiliary building, input motions are applied at these locations. Thus, the staff questioned the adequacy of the assumption that the piping systems (main s~ team and feedwater lines) incorporated into the three dimensional model were excited at the auxiliary building snubber locations with the containment Hosgri response spectra.
The above concern applies to both units and was discussed during the audit in the Westinghouse office (Monroeville, Pennsylvania) on February 28, 1985.
A local frequency was calculated by the staff using the main steam line cross-sectional properties and the distance between the attachment points i.e.,
snubber locations at the auxiliary building and the pipeway structure.
It was shown that the natural frequency of this segment of the mainstream line is approximately equal to 3 hz. At this frequency the containment spectra envelop the corresponding auxiliary building spectra. Thus, the use of containment input spectra at the snubber locations in the auxiliary building is conservative in this case. Based on the above, the staff concludes that the input trans-mitted through the piping systems into the pipeway structure would not affect
.the results of the seismic analyses for the Hosgri event. This conclusion applies to both units.
2.2.3 INTEGRATION TIME STEP The integration time step used by both Westinghouse and PG&E in their seismic evaluation of the pipeway structures for Unit 1 and Unit 2, respectively, was
.- equal to 0.01 seconds. Since response computations are generally sensitive to the choice of the integration time interval, the staff requested PG&E to justify the time step used in the evaluations. This issue was addressed in SSER 29 (p. 8-3, Item c).
Westinghouse used the computer code WECAN for the Unit 1 pipeway structure Hosgri evaluation. For the corresponding Unit 2 evaluation, PG&E used the Bechtel computer code BSAP.
In these evaluations the time history method was employed. The integration procedure used in the BSAP code was the Nigam-Jennings type. This method is based on the exact solution to the second order differential equation of motion for a linear segment type of input. Experience has shown that this procedure yields accurate results provided that the integration time interval is properly choosen. As a general rule, the time intervals should be taken as a fraction of the period of interest. The staff requested PG&E to show the effect of the time step interval on the results.
In response to the staff's request, PG&E presented comparative studies in which response spectral curves were calculated with smaller time steps.
Floor retponse spectra were computed with low and high damping (i.e., 2%
and 7%). The response spectra curves generated with the time step of 0.01 seconds were found to be in very good agreement with those computed with the smaller time step of 0.003 seconds.
Based on the above, the staff concludes that no significant approximation is introduced in the pipeway structure response evaluations due to the time interval used in the analysis. This conclusion applies to both Units.
2.2.4 DE AND DDE EVALUATION Before the staff's January 16 and 17, 1985 audit, no particular structural evaluations had been performed for the pipeway structure of both Units with respect to the DE and DDE loading conditions. PG&E justified the lack of a DE and DDE evaluation on the results obtained from the Hosgri and pipe rupture loading evaluations. However, load combinations involving DE and I
k
s DDE may control the design of some members of the pipeway struc.ture due to differences in the criteria (i.e., damping values, stress allowables, modeling procedures) for the three earthquake evaluations. Thus the staff requested PG&E to provide calculations in order to assure that the stresses in the pipeway structure satisfy the FSAR design criteria for loading conditions for the DE and DDE.
Subsequent evaluations for the pipeway structures of both Units were performed by PG&E and reviewed by the staff. The Unit 1 pipeway DE and DDE evaluations are documented in the following calculation:
Calculation No. 2151C-2, Diablo Canyon Unit 1, Qualification of Unit 1 Pipeway for DE and DDE The corresponding evaluations done for the Unit 2 pipeway structure are given in the following calculation:
Calculation No. 1149C-1: Diablo Canyon Unit 2, Pipeway Structure Unit 2 Evaluation and Acceptance of Selected Critical Members in the Pipeway Structures due to DE, DDE and Rupture Loading.
In the above calculations the members of the pipeway structure were evaluated for load combinations involving dead load, DE, DDE, and pipe rupture restraint load. The allowable stresses are the same for both Units. The objective of these calculations was to compute stress ratios (i.e. demand divided by capacity) at selected members of the pipeway structure in both Units for a set of load combinations including the DE and DDE loads.
Individual seismic analyses for DE and DDE were not performed to compute stresses in the pipeway structure.
Instead, they were computed by using spectral ratios for DE and DDE to Hosgri spectra and the stress ratios based on information from the detailed Hosgri seismic analyses.
O The Unit 2 pipeway DE and DDE evaluations were performed for a set of 4 load combinations committed in the FSAR. A review of member stresses was performed from the data due to dead load, Hosgri and pipe rupture analyses. As a result of this review two radial bents (28 and 3B) were identified to contain critical members and were used in the DE and DDE evaluation. Radial members of the pipeway structure were found to be generally more critical than tangential or vertical members.
Following the selection of bents 2B and 38, spectral ratios were used to compute stress ratios for several members.
From the Unit 2 pipeway evaluation one structural member (elevation 108' - 3")
in bent 2B reached the shear capacity for the load combination involving dead, DE and DDE and pipe rupture load. However, some of this load should be redistributed to the adjacent members. This situation was not found for bent 3B.
With respect to the DE and DDE evaluations for the Unit 1 pipeway structure, the computation of the stress ratios at six selected bents was based on conversion factors. During the May 1985 audit, the staff requested a justification of th'e screening criteria used to establish these factors.
Furthermore, PG&E was requested to provide bounds with regards to the magnitude of the approximation associated with the conversion factors.
Following the May 1985 audit, PG&E provided additional information which is based on Unit 1 pipeway specific data. The latter calculations were found to be more appropriate than those presented previously. Specifically, a larger number of structural members were incorporated into the DE and DDE l
evaluation procedure. These members were selected on the basis of the Westinghouse results from the Hosgri's evaluation. According to the screening criteria used in the selection process, all memoers with stress ratios equal or greater than 0.90 were used in the DE and DDE evaluation.
l l
l
.- Furthermore, the original conversion factors are not utilized.
Instead new ratios of DE and DDE to Hosgri were computed. The values found for these ratios are generally less than one, although values greater than one were found for three members (1103, 828, and 496). However, the computed stress ratios were less than one for the all members, including the above three members. The maximum stress ratio found in this evaluation was 0.97.
Moreover, the stress ratios associated with members controlled by the pipe rupture load are generally less than those of the members for which the DE and DDE governs. Based on the computed stress ratios the staff concludes that the pipeway members selected in the DE and DDE Unit 1 evaluations satisfy the design allowables.
Based on the above, the staff concludes that the general procedures used by PG&E in the DE and DDE evaluations of the Unit 1 and Unit 2 pipeway structures are acceptable pending a satisfactory confirmatory analysis.
The procedure used by PG&E to identify the critical members in the pipeway structure for the DE and DDE evaluation used the results of the Hosgri seismic analysis. The procedure involves various approximations and the staff is not completely satisfied as to the accuracy of the results.
However, the accuracy is judged to give reasonable interim results.
Furthermore, the pipeway structure is readily accessible and could be quickly modified if necessary. Based on the above the staff concludes that the general procedure used by PG&E in the DE and DDE evaluation of Unit 1 and Unit'.2 pipeway structures are acceptable subject to a confirmatory analysis as indicated in the following section.
2.3 FINDINGS AND CONCLUSIONS Based on the staff's audits and reviews of the extensive information provided by PG&E described above, the staff's findings are summarized as follows:
(1) The modeling of the pipeway structure connections to the auxiliary and turbine buildings is found to be acceptable.
9
l
.- (2) The choice of the input to the pipeway structure model should not affect the response results obtained for the pipeway structure. Any input to the pipeway structure through the piping system is not expected to alter the results of the Hosgri evaluation.
(3) The integration time step used in the Hosgri evaluations has no significant impact on the response computations.
(4) The procedures used in the DE and DDE evaluations are found to be generally acceptable.
It is realized, however, that the computations of the DE and DDE stresses at the pipeway structural members involves various approximations. The degree of the approximation should be further assessed, since results from individual seismic DE and DDE analyses are not presently available. Therefore, the staff will include the following as a condition in the Unit 2 full power license.
Before the first refueling outage, "PG&E shall perform a confirmatory analysis for Unit 2 pipeway structure to further demonstrate the adequacy of the pipeway structure for the load combinations that include the design earthquake and the double design earthquake."
Since the same approach was used in the DE and DDE evaluations of both Unit I and Unit 2 pipeway structures the conclusion from this confirmatory analysis should be applicable to both units.
3.0 MASONRY WALLS During the staff audits in October and November 1984 at the PG&E office in San Francisco and the plant site, the staff reviewed documents related to the masonry wall construction to determine if:
S 1.
the documentation is sufficient to fulfill the requirements of "special inspection" for masonry walls, and
i p 2.
the masonry walls are constructed in accordance with the design documents.
The construction documents included material test reports, photographs and construction logs of five walls.
In addition, the staff also discussed the construction / inspection practices used in the masonry wall construction with the construction personnel involved in those activities. The staff requested PG&E to further undertake an extensive survey of the construction documents and report to the staff its conclusions regarding the above two objectives.
By a letter dated May 23, 1985, the applicant has provided details of its investigation. These details indicate the following:
(1) The construction records provide the details of construction for various walls located 'in the turbine and auxiliary buildings.
They document the various stages of construction, such as:
(a) hauling reinforcing steel and blocks to the construction area (b) placing reinforcing steel and blocks (c) filling block cells with grout F
(d) providing drypack at top of the wall (2) The construction logs indicate that proper communication channels were established between construction and engineering. The
. inspector assi,gned for blockwall construction contacted engineers on the project for approval of changes in design and documentation.
(3) Blocks, in-fill grout, and reinforcing steel were tested for compliance with the appropriate standards during construction of a majority of the block walls. Since the walls were not classified as Class I at the time of construction, test results for some of the walls are not available. However, many test results are available and they indicate that the construction materials met or exceeded specification requirements.
9 F
r o
l 9
. PG&E has further noted that during the implementation of recent wall modifications, it was necessary to locate the reinforcing steel in the masonry walls to avoid cutting the bars. This effort has also confirmed that the reinforcing steel was provided in accordance with the design drawings. During the site visit, the staff also observed the marked locations of the reinforcing bar on faces of walls.
Based on the above findings and results of its own investigation, the staff concludes that the masonry walls at Diablo Canyon plant meet the intent of "special inspection" requirements specified in " Code Requirements for Concrete Masonry Structures" (ACI 531-79), American Concrete Institute and there is a reasonable assurance that they were constructed in accordance with the design documents. The only remaining issue regarding the masonry wall at Diablo Canyon Unit I and Unit 2 is the license condition as discussed in SER Supplement No. 31. The staff will report its activities regarding the license condition in the future supplements.
,_.