Text

0 to Updated Final Safety Analysis Report, Chapter 3, Appendix 3G, Nuclear Island Seismic Analyses
	ML21179A125
Person / Time
Site:	Vogtle
Issue date:	06/15/2021
From:	; Southern Nuclear Operating Co
To:	; Office of Nuclear Reactor Regulation
Shared Package
ML21179A130	List: ML21179A093; ML21179A094; ML21179A095; ML21179A096; ML21179A097; ML21179A098; ML21179A099; ML21179A100; ML21179A101; ML21179A102; ML21179A103; ML21179A104; ML21179A105; ML21179A106; ML21179A107; ML21179A108; ML21179A109; ML21179A110; ML21179A111; ML21179A112; ML21179A113; ML21179A114; ML21179A115; ML21179A116; ML21179A117; ML21179A118; ML21179A119; ML21179A121; ML21179A122; ML21179A123; ML21179A124; ML21179A125; ND-21-0486, 0 to Updated Final Safety Analysis Report, Chapter 4, Reactor;
References
	ND-21-0486
	Download: ML21179A125 (110)
	v • d • e

VEGP 3&4 - UFSAR Appendix 3G Nuclear Island Seismic Analyses 3G.1 Introduction This appendix summarizes the seismic analyses of the nuclear island building structures performed to support the AP1000 design certification extension from just hard rock sites, to sites ranging from soft soils to hard rock. The seismic Category I building structures consist of the containment building (the steel containment vessel [SCV] and the containment internal structures [CIS]), the shield building, and the auxiliary building. These structures are founded on a common basemat and are collectively known as the nuclear island or nuclear island structures. Key dimensions of the seismic Category I building structures, such as thickness of the basemat, floor slabs, roofs and walls, are shown in Figures 3.7.1-14 and 3.7.2-12.

Analyses were performed in accordance with the criteria and methods described in Section 3.7.

Section 3G.2 describes the development of the finite element models. Section 3G.3 describes the soil structure interaction analyses of a range of site parameters and the selection of the parameters used in the design analyses. Section 3G.4 describes the fixed base and soil structure interaction dynamic analyses and provides typical results from these dynamic analyses. References 3 and 6 provide a summary of dynamic and seismic analysis results (i.e., modal model properties, accelerations, displacements, response spectra) and the nuclear island liftoff analyses. The seismic analyses of the nuclear island are summarized in a seismic analysis summary report. Deviations from the design due to as-procured or as-built conditions are acceptable based on an evaluation consistent with the methods and procedures of Sections 3.7 and 3.8 provided the following acceptance criteria are met:

The structural design meets the acceptance criteria specified in Section 3.8.

The seismic floor response spectra (FRS) meet the acceptance criteria specified in Subsection 3.7.5.4.

Depending on the extent of the deviations, the evaluation may range from documentation of an engineering judgment to performance of a revised analysis and design. The results of the evaluation will be documented in an as-built summary report by the Combined License applicant.

Table 3G.1-1 and Figure 3G.1-1 summarize the types of models and analysis methods that are used in the seismic analyses of the nuclear island, as well as the type of results that are obtained and where they are used in the design. Table 3G.1-2 summarizes the dynamic analyses performed and the methods used for combination of modal responses and directional input.

3G.2 Nuclear Island Finite Element Models The AP1000 nuclear island consists of three distinct seismic Category I structures founded on a common basemat. The three building structures that make up the nuclear island are the coupled auxiliary and shield building (ASB), the SCV, and the CIS. The shield building and the auxiliary building are monolithically constructed with reinforced concrete and, therefore, considered one structure. The nuclear island is embedded approximately 40 feet with the bottom of basemat at elevation 60-6 and plant grade located at elevation 100-0. The SCV is described in Subsection 3.8.2, the CIS in Subsection 3.8.3, the ASB in Subsection 3.8.4, and the nuclear island basemat in Subsection 3.8.5.

Seismic systems are defined, according to SRP 3.7.2 (Reference 1),Section II.3.a, as the seismic Category I structures that are considered in conjunction with their foundation and supporting media to form a soil-structure interaction model. Fixed base seismic analyses are performed for the nuclear island at a rock site. Soil-structure interaction analyses are performed for soil sites. The analyses generate a set of in-structure responses (design member forces, nodal accelerations, nodal 3G-1 Revision 4

VEGP 3&4 - UFSAR displacements, and floor response spectra), which are used in the design and analysis of seismic Category I structures, components, and seismic subsystems. Concrete structures are modeled with linear elastic uncracked properties. However, the modulus of elasticity is reduced to 80% of the ACI code value to reduce stiffness to simulate cracking.

A seismic response spectrum analysis is performed to develop the seismic design loads for the design of the auxiliary building, shield building, and containment internal structure, and the loads generated include the amplified load due to flexibility and the distribution of this load to the surrounding structures. Equivalent static analyses are used to design the shield building roof and radial roof beams, tension ring, air inlet structure, and PCS tank.

3G.2.1 Individual Building and Equipment Models 3G.2.1.1 Coupled Auxiliary and Shield Building The finite element shell dynamic model of the coupled ASB is a finite element model using primarily shell elements. The portion of the model up to the elevation of the auxiliary building roof is developed using the solid model features of ANSYS, which allow definition of the geometry and structural properties. The nominal element size in the auxiliary building model is about 9 feet so that each wall has two elements for the wall height of about 18 feet between floors. This mesh size, which is the same as that of the solid model, has sufficient refinement for global seismic behavior. It is combined with a finite element model of the shield building roof and cylinder above the elevation of the auxiliary building roof. This model is shown in Figure 3G.2-1. This finite element shell dynamic model is part of the NI10 model.

Since the water in the passive containment cooling system tank responds at a very low frequency (sloshing) and does not affect building response, the passive containment cooling system tank water mass is reduced to exclude the low frequency water sloshing mass. The wall thickness of the bottom portion of the shield building (elevation 63.5 to 81.5) is modeled as one half (1.5) since the CIS model is connected to this portion and extends out to the mid-radius of the shield building cylindrical wall. Local portions of the ASB floors and walls are modeled with sufficient detail to give the response of the flexible areas.

3G.2.1.2 Containment Internal Structures The finite element shell model of the containment internal structures is a finite element model using primarily shell elements for the walls and floors and solid elements for the mass concrete. It is developed using the solid model features of ANSYS, which allow definition of the geometry and structural properties. This model is used in both static and dynamic analyses. It models the inner and outer mass concrete basemats embedding the lower portion of the containment vessel, and the concrete structures above the mass concrete inside the containment vessel. The walls and basemat inside containment for this model are shown in Figure 3G.2-2. The basemat (dish) outside the containment vessel is shown in Figure 3G.2-3. This finite element shell dynamic model is part of the NI10 model. Static analyses are also performed on the model to obtain member forces in the walls.

This model is also used in the 3D finite element basemat model (see Subsection 3.8.5.4.1).

3G.2.1.3 Containment Vessel The SCV is a freestanding, cylindrical, steel shell structure with ellipsoidal upper and lower steel domes. The finite element model of the containment vessel is an axisymmetric model fixed at elevation 100. Static analyses are performed with this model to obtain shell stresses as described in Subsection 3.8.2.4.1.1. The model is also used to develop modal properties (frequencies and mode shapes). The three-dimensional, lumped-mass stick model of the SCV is developed based on the 3G-2 Revision 4

VEGP 3&4 - UFSAR axisymmetric shell model. Figure 3G.2-4 presents the SCV stick model. In the stick model, the properties are calculated as follows:

Members representing the cylindrical portion are based on the properties of the actual circular cross section of the containment vessel.

Members representing the bottom head are based on equivalent stiffnesses calculated from the shell of revolution analyses for static 1.0g in vertical and horizontal directions.

Shear, bending and torsional properties for members representing the top head are based on the average of the properties at the successive nodes, using the actual circular cross section.

These are the properties that affect the horizontal modes. Axial properties, which affect the vertical modes, are based on equivalent stiffnesses calculated from the shell of revolution analyses for static 1.0g in the vertical direction.

The equivalent static acceleration analyses of the containment vessel use a finite element shell model with a refined mesh in the area adjacent to the large penetrations. Comparison of this with a time history analysis for the regions immediately surrounding the large penetrations verifies that the loads from equivalent static analysis are conservative to time history using a representative study.

The stick model is combined with the polar crane stick model as shown in Figure 3G.2-4. Modal properties of the containment vessel with and without the polar crane are shown in Table 3G.2-1. It is connected to nodes on the dish model. NI10 node numbers are shown in red and NI20 node numbers are shown in black.

The method used to construct a stick model from the axisymmetric shell model of the containment vessel is verified by comparison of the natural frequencies determined from the stick model and the shell of revolution model as shown in Table 3G.2-2. The shell of revolution vertical model (n = 0 harmonic) has a series of local shell modes of the top head above elevation 265 between 23 and 30 hertz. These modes are predominantly in a direction normal to the shell surface and cannot be represented by a stick model. These local modes have small contribution to the total response to a vertical earthquake as they are at a high frequency where seismic excitation is small. The only seismic Category I components attached to this portion of the top head are the water distribution weirs of the passive containment cooling system. These weirs are designed such that their fundamental frequencies are outside the 23 to 30 hertz range of the local shell modes.

An evaluation was made of the connection of the bottom of the steel containment vessel stick model to the CIS finite element model. Comparisons were made between the unconstrained fully symmetric, radially constrained fully symmetric, and original asymmetric connectivity models. The response spectra at the elevation of the polar crane girder for the first two models are almost identical, and the third model had only minor differences. Based on this comparison, the unconstrained fully symmetric connectivity model is used.

3G.2.1.4 Polar Crane The polar crane is supported on a ring girder, which is an integral part of the SCV at elevation 228-0, as shown in Figure 3.8.2-1. It is modeled as a multi-degree of freedom system attached to the steel containment shell at elevation 224 (midpoint of ring girder) as shown in Figure 3G.2-4. The polar crane is modeled using a simplified and detailed model. The simplified model has five masses at the mid-height of the bridge at elevation 236-6 and one mass for the trolley, as shown in Figure 3G.2-5A. The polar crane model includes the flexibility of the crane bridge girders and truck assembly. The containment shells local flexibility is considered when combined with the global model. When fixed at the center of containment, the model shows fundamental frequencies of 4.6 hertz transverse to the bridge, 6.9 hertz vertically, and 9.4 hertz along the bridge. The Detailed 3G-3 Revision 4

VEGP 3&4 - UFSAR Model of the polar crane consists of 76 nodes is defined having 96 dynamic degrees of freedom. It is used to verify the accuracy of the simplified model. This model is shown in Figure 3G.2-5B.

Nodes 425 to 428 are connected to 48 additional nodes. All 52 of these nodes are used to model the truck stiffnesses. The local SCV stiffness is considered when the detailed model of the polar crane is combined with the SCV model.

1. The elements connecting nodes 433, 431/477, 435, 436, 432/478, and 434 represent the trolley.

2. Nodes 437 to 454 are located on the polar crane bridge girders and cross beams. The end nodes (437, 443, 444, and 450) are used to connect the cross beams to the girders; these nodes are also attached to the trucks (nodes 425 to 428) by rigid links.

3. When the Detailed Model of the polar crane is combined with the SCV model, nodes 508 to 511, 528 to 531, 548 to 551, and 568 to 571 are merged to SCV nodes on the polar crane ring girder.

3G.2.1.5 Major Equipment and Structures Using Stick Models The major equipment supported by the CIS is represented by stick models connected to the CIS.

These stick models are the reactor coolant loop model shown in Figure 3G.2-6, the pressurizer model shown in Figure 3G.2-7, and the core makeup tank model shown in Figure 3G.2-8. The core makeup tank model is used only in the nuclear island fine (NI10) model; the core makeup tank is represented by mass in the nuclear island coarse model (NI20).

3G.2.2 Nuclear Island Dynamic Models Finite element shell models (3D) of the nuclear island concrete structures are used for the time history seismic analyses. Stick models are coupled to the shell models of the concrete structures for the containment vessel, polar crane, the reactor coolant loop and pressurizer. Two models are used.

The fine (NI10) model is used to define the seismic response for the hard rock site. The coarse (NI20) model is used for the soil structure interaction (SSI) analyses. It is similar to the NI10 model with the exception that the mesh size for the ASB and CIS is approximately 20 feet instead of 10 feet.

This model is set up in both ANSYS and SASSI. The NI05 model is used to develop amplified seismic response for the envelope of soil profiles presented in Subsection 3.7.1.4 for flexible regions not captured by the coarser NI20 model. The NI05 model is also used in response spectrum analysis of the nuclear island to develop design seismic member forces and moments. The NI10, NI20, and NI05 models are described in the subsections below.

3G.2.2.1 NI10 Model The large solid-shell finite element model of the AP1000 nuclear island shown in Figure 3G.2-9 combines the ASB solid-shell model described in Subsection 3G.2.1.1, and the CIS solid-shell model described in Subsection 3G.2.1.2. The containment vessel and major equipment that are supported by the CIS are represented by stick models and are connected to the CIS. These stick models are the SCV and the polar crane models, the reactor coolant loop model, core makeup tank models, and the pressurizer model. The stick models are described in Subsections 3G.2.1.3 and 3G.2.1.4. The CIS and attached sticks are shown in Figure 3G.2-10. This AP1000 nuclear island model is referred to as the NI10 or fine model. The ASB portion of this model has a mesh size of approximately 10 feet.

The SCV is connected to the CIS model using constraint equations. The SCV at the bottom of the stick at elevation 100 (node 130401) is connected to CIS nodes at the same elevation. Figure 3G.2-4 shows the SCV stick model with the constraint equation nodes. The nodes are defined using a 3G-4 Revision 4

VEGP 3&4 - UFSAR cylindrical coordinate system whose origin coincides with the center of containment (node 130401).

The CIS vertical displacement is tied rigidly (constrained) to the vertical displacement and RX and RY rotations of node 130401. The CIS tangential displacement is tied rigidly (constrained) to the horizontal displacement and RZ rotation of node 130401.

3G.2.2.2 NI20 Model The NI20 coarse model has fewer nodes and elements than the NI10 model. It captures the essential features of the nuclear island configuration. The nominal shell and solid element dimension is about 20 feet. It is used in the soil-structure interaction analyses of the nuclear island performed using the program SASSI. The stick models are the same as used for the NI10 model except that the core makeup tank is not included. This model is shown in Figures 3G.2-11 and 3G.2-12. Results of fixed base analyses of the NI20 model were compared to those of the NI10 model to confirm the adequacy of the NI20 model for use in the soil-structure-interaction analyses.

3G.2.2.3 Nuclear Island Stick Model The nuclear island lumped-mass stick model consists of the stick models of the individual buildings interconnected by rigid links. Each individual stick model is developed to match the modal properties of the finite element models described in Subsections 3G.2.1.1 and 3G.2.1.2 above. Modal analyses and seismic time history analyses were performed using this model for the hard rock design certification.

The nuclear island lumped-mass stick model has been replaced in the design analyses described in this appendix by the NI10 and NI20 finite element shell dynamic models of the nuclear island described in Subsections 3G.2.2.1 and 3G.2.2.2 above. A 2D stick model is used in the soil sensitivity analyses described in Section 3G.3.

3G.2.2.4 NI05 Model The NI05 solid-shell finite element model of the AP1000 nuclear island is shown in Figures 3G.2-13 to 3G.2-15. The NI05 model is used for response spectrum analysis of the nuclear island auxiliary and shield building structures. The NI05 model is also used for the mode superposition time history analysis of the nuclear island for the amplified response at flexible floors. The NI05 model is used for the static analysis of the nuclear island for the basemat design. The NI05 model is a refined version of the NI10 model where the auxiliary and shield building mesh size is reduced from approximately 10 feet by 10 feet tetrahedral mesh to approximately 5 feet by 5 feet. The major equipment stick models supported by the CIS are the same as used for the NI10 model. The steel containment vessel stick model and connections are also the same as the NI10 model. The only difference between the NI05 CIS and NI10 CIS is the basemat (bowl) and dish region as shown in Figure 3G.2-15. The model is validated by a comparison of the mass participation by frequency of the fundamental modes to those of the NI10 model.

3G.2.2.5 Seismic Stability Model The sliding stability of the nuclear island basemat is evaluated using a non-linear 2D East-West (EW) stick model of the nuclear island structures using the ANSYS program. Three concentric sticks represent ASB, CIS, and SCV, respectively. The reactor coolant loop is included as mass only. The basemat is modeled as a rigid beam, which is free in translation along the EW and vertical directions.

The nuclear island combined sticks are attached to the rigid basemat at the nuclear island mass center.

Each node of the rigid basemat is connected with two spring elements in the horizontal and vertical directions, respectively. The spring elements only model the foundation media (rock or soil) damping, 3G-5 Revision 4

VEGP 3&4 - UFSAR not stiffness. A layer of contact elements is added along the rigid basemat bottom to simulate the friction forces between basemat bottom and foundation media as well as foundation media stiffnesses. The friction coefficient between the basemat bottom and the soil media is set at 0.55.

Figure 3G.2-19 shows the schematic of this non-linear 2D EW nuclear island stick model. The contact elements are free to uplift when the upward force (normal force) is larger than the associated dead load component. When the tangential force is larger than the friction force, sliding occurs.

3G.2.3 Static Models Member forces in the ASB are obtained from analyses of a model that is more refined than the finite element model described in Subsection 3G.2.1.1. This model is developed by meshing one area of the solid model with four finite elements. The nominal element size in this auxiliary building model is about 4.5 feet so that each wall has four elements for the wall height of about 18 feet between floors.

This finite element shell model is referred to as the NI05 model. This refinement is used to calculate the design member forces and moments using response spectra analysis of the nuclear island models with seismic input enveloping all soil conditions. The finite element shell model of the containment internal structures described in Subsection 3G.2.1.2, which includes the basemat within the shield building and the containment vessel stick model, is also included.

Finite element solid/shell models were used for the equivalent static seismic analysis. For the detailed design of the shield building roof, a finite element model of one quadrant of the roof is used as described in Subsection 3G.2.3.1. For the detailed design of the steel containment vessel, a shell mesh finite element model with a much finer mesh in the areas surrounding the major penetrations is used as described in Subsection 3G.2.3.2. For the static analysis of the containment vessel, an axisymmetric model is used as described in Subsection 3G.2.3.3. The nuclear island basemat is evaluated using the NI05 finite element model described in Subsection 3G.2.2.4.

3G.2.3.1 Quadrant Model of Shield Building Roof The one quadrant model of the shield building roof is shown in Figure 3G.2-16. The model is constructed with solid and shell elements and contains structures from the exposed shield wall through the top of the shield building roof. The quadrant model is used for the equivalent static analysis of the shield building roof. The results from the more detailed analysis are used in the design of the shield building roof and radial roof beams, tension ring, air inlet structure, and PCS tank.

3G.2.3.2 Containment Vessel 3D Finite Element Model The 3D finite element model of the steel containment vessel is shown in Figure 3G.2-17. The finite element model for the steel containment vessel is used for the stress analysis of the large penetrations (personnel locks and equipment hatches) of the containment vessel.

3G.2.3.3 Containment Vessel Axisymmetric Model The axisymmetric finite element model of the steel containment vessel is shown in Figure 3G.2-18.

The axisymmetric model is a two-dimensional model with added mass for the stiffeners, crane girder, equipment hatches, and air locks.

3G.3 2D SASSI Analyses This section describes the soil structure interaction analyses performed using 2D models in SASSI to select the design soil cases for the AP1000. The AP1000 footprint, or interface to the soil medium, is identical to the AP600. The AP1000 containment and shield building are 25 6 and 20 6 (Reference 4) respectively taller than AP600. Results and conclusions from the AP600 soil studies (Reference 2) are considered in establishing the design soil profiles for the AP1000.

3G-6 Revision 4

VEGP 3&4 - UFSAR Analyses were performed using 2D stick models of the AP1000 for horizontal seismic input with and without adjacent structures. The soil profiles included a hard rock site, a firm rock site, a soft rock site, a soft-to-medium soil site, an upper bound soft-to-medium site, and a soft soil site. Analyses were also performed without adjacent structures for a hard rock site, a firm rock site, a soft rock site, a soft-to-medium soil site, an upper bound soft-to-medium site, and a soft soil site. The soil damping and degradation curves are described in Subsection 3.7.1.4. The soil profiles selected for the AP1000 use the same parameters on depth to bedrock, depth to water table, and variation of shear wave velocity with depth as those used in the AP600 design analyses. The Poissons ratio is 0.25 for rock sites (hard and firm rock) and 0.35 for soil sites (soft-to-medium soil, and upper bound soft-to-medium soil). For all the soil profiles defined, the base rock has been taken to be at 120 feet below grade level. The soil profiles are shown in Figure 3G.3-1. The shear wave velocity profiles and related governing parameters are as follows:

For the hard rock site, an upper bound case for rock sites using a shear wave velocity of 8000 feet per second.

For the firm rock site, a shear wave velocity of 3500 feet per second to a depth of 120 feet, and base rock at the depth of 120 feet.

For the soft rock site, a shear wave velocity of 2400 feet per second at the ground surface, increasing linearly to 3200 feet per second at a depth of 240 feet, and base rock at the depth of 120 feet.

For the upper bound soft-to-medium soil site, a shear wave velocity of 1414 feet per second at ground surface, increasing parabolically to 3394 feet per second at 240 feet, base rock at the depth of 120 feet, and ground water at grade level. The initial soil shear modulus profile is twice that of the soft-to-medium soil site.

For the soft-to-medium soil site, a shear wave velocity of 1000 feet per second at ground surface, increasing parabolically to 2400 feet per second at 240 feet, base rock at the depth of 120 feet, and ground water is assumed at grade level.

For the soft soil site, a shear wave velocity of 1000 feet per second at ground surface, increasing linearly to 1200 feet per second at 240 feet, base rock at the depth of 120 feet, and ground water is assumed at grade level.

The analyses with and without adjacent structures demonstrated that the effect of adjacent buildings on the nuclear island response is small. Based on this, the 3D SASSI analyses of the AP1000 nuclear island can be performed without adjacent buildings similar to those performed for the AP600.

The maximum acceleration values obtained from the AP1000 analyses without adjacent structures are given in Table 3G.3-1. The soil cases giving the maximum response are shown in bold. Floor response spectra associated with nodes 41, 120, 310, 411, and 535 for the six AP1000 soil cases are shown in Figures 3G.3-2 to 3G.3-11.

Based on review of the above results, five soil conditions were selected for 3D SASSI analyses in addition to the hard rock condition evaluated in the existing AP1000 Design Certification. Thus, the following five soil and rock cases identified in Subsection 3.7.1.4 are considered: hard rock, firm rock, soft rock, soft-to-medium soil, upper bound soft-to-medium, and soft soil.

3G-7 Revision 4

VEGP 3&4 - UFSAR 3G.4 Nuclear Island Dynamic Analyses 3G.4.1 ANSYS Fixed Base Analysis The NI10 model described in Subsection 3G.2.2.1 was analyzed by time history modal superposition.

To perform the time history analysis of this large model, the ANSYS superelement (substructuring) techniques were applied. Substructuring is a procedure that condenses a group of finite elements into one element represented as a matrix. The reasons for substructuring are to reduce computer time of subsequent evaluations. Two sets of analyses were performed. To obtain the time history response of the ASB, the ASB finite element model was merged with the superelement of the CIS and its major components. To obtain the time history response of the CIS, the CIS finite element model was merged with the superelement of the ASB.

Deflection time history responses were obtained at selected representative locations. These locations included major wall and floor intersections and nodes at the cardinal orientations at key elevations of the shield building. Nodes were also selected at mid-span on flexible walls and floors.

Typical locations are shown for the ASB at elevation 135 on Figures 3G.4-1 and 3G.4-2.

Figure 3G.4-1 shows the rigid locations, and Figure 3G.4-2 shows the flexible locations.

ANSYS is used to calculate the maximum relative deflection to the nuclear island for the envelope case that considers all of the soil and hard rock site cases. Synthesized displacement time histories are developed using the envelope seismic response spectra from the six site conditions (hard rock, firm rock, soft rock, upper-bound-soft-to-medium, soft-to-medium, and soft soil). Seismic response spectra at nine locations are used (four edge locations, one center location, and four corner locations). It is not necessary to adjust for drift since relative deflections to the basemat are calculated and the drift would be subtracted from the results.

3G.4.2 3D SASSI Analyses The computer program SASSI 2000 is used to perform Soil-Structure Interaction analysis with the NI20 Coarse Finite Element Model. The SASSI Soil-Structure Interaction analyses are performed for the five soil conditions established from the AP1000 2D SASSI analyses. These soil conditions are firm rock, soft rock, soft-to-medium soil, upper bound soft-to-medium, and soft soil. The model includes a surrounding layer of excavated soil and the existing soil media as shown in Figures 3G.4-3 and 3G.4-4. Acceleration time histories and floor response spectra are obtained. Adjacent structures have a negligible effect on the nuclear island structures and, thus, are not considered in the 3D SASSI analyses.

Westinghouse has adopted the approach that calculates displacements internally within the ACS SASSI program based on an analytical complex frequency domain approach that uses inverse Fast-Fourier Transforms (FFT) to compute relative displacement histories instead of double numerical integration in the time domain that computes absolute displacement time histories from absolute acceleration time histories.

The relative displacement time history is calculated using ACS SASSI RELDISP module. The complex acceleration transfer functions (TF) are computed for reference and all selected output nodes. The relative acceleration transfer function is calculated by subtracting the reference node TF from the output node TF. The relative displacement transfer function is obtained by dividing the circular frequency square (²) for each frequency data point. The relative displacement time history is obtained by taking the inverse FFT.

Relative displacements are calculated between adjacent buildings and the nuclear island using soft springs between the buildings. The spring stiffness is very small so that it does not affect the dynamic 3G-8 Revision 4

VEGP 3&4 - UFSAR response. These calculations are performed using 2D models and SASSI 2000. The relative deflection is calculated using the maximum compressive spring force and the stiffness value.

In these analyses, the three components of ground motions (N-S, E-W, and vertical direction) are input separately. Each design acceleration time history (N-S, E-W, and vertical) is applied separately, and the time history responses are calculated at the required nodes. The resulting co-linear time history responses at a node due to the three earthquake components are then combined algebraically.

3G.4.3 Seismic Analysis 3G.4.3.1 Response Spectrum Analysis The response spectrum methodology used in the AP1000 design employs the Complete Quadratic Combination (CQC, Section 1.1 of Reference 5) grouping method for closely spaced modes with the Der Kiureghian Correlation Coefficient (Section 1.1.3 of Reference 5) used for correlation between modes. The Lindley-Yow (Section 1.3.2, Reference 5) spectra analysis methodology is employed for modes with both periodic and rigid response components. The modal analysis performed to develop composite modal participation is used to develop input for the response spectrum analysis. Modes ranging from 0 to 33 Hz or higher are considered. For modes above the cutoff frequency, the Lindley-Yow is used. The Static ZPA Method (Section 1.4.2, Reference 5) is employed for the residual rigid response component for each mode as outlined in NRC Regulatory Guide 1.92 (Reference 5). The complete solution is developed via Combination Method B (Section 1.5.2, Reference 5). The combined effects, considering three spatial components of an earthquake (N-S, E-W, and Vertical),

are combined by square root sum of the squares method (Section 2.1, Reference 5).

In Subsection 3.7.2.6, Three Components of Earthquake Motion, the combination of three components of earthquake motion is discussed.

3G.4.3.2 Absolute Accelerations The seismic analyses results, which include the new shield building configuration described in Section 3.8, are given in Reference 3.

3G.4.3.3 Seismic Response Spectra The AP1000 plant floor response spectra for the six key locations are provided in Figure 3G.4-5X to 3G.4-10Z. The key locations are defined in Table 3G.4-1. The design seismic response spectra are conservatively adjusted in the low frequency range in anticipation of future sites having a slightly higher response at the lower frequency.

The in-structure response spectra at six key locations, as defined below, are used if a site-specific 3D dynamic analysis evaluation as outlined in Subsection 2.5.2 is required. The site is acceptable if the floor response spectra from the site-specific evaluation do not exceed the AP1000 spectra for each of the locations identified below or the exceedances are justified.

[FRS Location Figure Numbers Containment internal structures at elevation of Figure 3G.4-5X to 3G.4-5Z reactor vessel support Containment operating floor Figure 3G.4-6X to 3G.4-6Z Auxiliary building NE corner at elevation 116'-6" Figure 3G.4-7X to 3G.4-7Z Shield building at fuel building roof Figure 3G.4-8X to 3G.4-8Z

In accordance with the departure evaluation process specified in License Condition 2.D.(13), NRC Staff approval may be required prior to implementing a change in this information.

3G-9 Revision 4

VEGP 3&4 - UFSAR Shield building roof Figure 3G.4-9X to 3G.4-9Z Steel containment vessel at polar crane support Figure 3G.4-10X to 3G.4-10Z]*

Note:

See Table 3G.4-1 for locations of six key locations.

3G.4.3.4 Bearing Pressure Demand Bearing pressure demand was calculated using both 2D and 3D analyses. Both linear and non-linear analyses are performed with the 2D nuclear island model. The maximum bearing pressures calculated include the effect of dead, live, and seismic loading.

The 2D model was used to evaluate the effect of liftoff on the bearing pressure. Since the largest bearing pressure will result from the east-west seismic excitation because of the smaller width of the basemat in this direction, liftoff was evaluated using an east-west stick model of the nuclear island structures, supported on a rigid basemat with non-linear springs. Direct integration time history analyses were performed. The bearing pressures calculated from these analyses are summarized in Table 3G.4-2. The pressures are at the edge of the basemat. Results are given for the three cases that result in the highest bearing pressure (hard rock [HR], upper bound soft to medium [UBSM] soil, and soft to medium [SM] soil). The linear results show maximum bearing pressures on the west side of 31 to 33 ksf. Liftoff increases the subgrade pressure close to the west edge by 4 percent to 6 percent with insignificant effect beneath most of the basemat.

The SASSI soil-structure interaction analyses are performed based on the nuclear island 3D SASSI model for the hard rock and five soil conditions established from the AP1000 2D SASSI analyses.

The SASSI model of the nuclear island is based on the NI20 finite element model. The bearing pressures from the 3D SASSI analyses have been obtained by combining the time history results from the north-south, east-west, and vertical earthquakes. The maximum soil-bearing pressure demand is obtained from the hard rock (HR) case equal to 35 ksf. It is noted that a maximum localized peak is obtained on the west edge of 38 ksf; a limit of 35 ksf for maximum bearing seismic demand is obtained by averaging the soil pressure over 335 ft2 of the west edge of the shield building where the maximum stress occurs.

3G.5 References

1. NUREG-800, Review of Safety Analysis Reports for Nuclear Power Plants, Section 3.7.2, Seismic System Analysis, Revision 2.

2. GW-GL-700, AP600 Design Control Document, Appendices 2A and 2B, Revision 4.

3. APP-GW-S2R-010, Extension of Nuclear Island Seismic Analyses to Soil Sites, Westinghouse Electric Company LLC.

4. APP-GW-GLN-112, Structural Verification for Enhanced Shield Building, Westinghouse Electric Company LLC.

5. U.S. NRC Regulatory 1.92, Revision 2, Combining Modal Responses and Spatial Components in Seismic Analysis.

6. APP-GW-GLR-044, Nuclear Island Basemat and Foundation, Westinghouse Electric Company LLC.

In accordance with the departure evaluation process specified in License Condition 2.D.(13), NRC Staff approval may be required prior to implementing a change in this information.

3G-10 Revision 4

VEGP 3&4 - UFSAR Table 3G.1-1 (Sheet 1 of 4)

Summary of Models and Analysis Methods Analysis Type of Dynamic Model Method Program Response/Purpose 3D (ASB) solid-shell - ANSYS Creates the finite element mesh for the ASB finite model element model.

3D (CIS) solid-shell - ANSYS Creates the finite element mesh for the CIS finite model element model.

3D finite element model - ANSYS ASB portion of NI10.

including shield building roof (ASB10) 3D finite element model Response spectrum ANSYS CIS portion of NI10.

including dish below analysis containment vessel 3D finite element shell Mode superposition ANSYS Performed for hard rock profile for ASB with CIS model of nuclear island time history analysis as superelement and for CIS with ASB as

[NI10] (coupled superelement.

auxiliary and shield building shell model, To develop time histories for generating plant containment internal design floor response spectra for nuclear island structures, steel structures.

containment vessel, polar crane, RCL, To obtain maximum absolute nodal accelerations pressurizer, and CMTs) (ZPA) to be used in equivalent static analyses.

To obtain maximum displacements relative to basemat.

3D finite element Mode superposition ANSYS Performed for hard rock profile for comparisons coarse shell model of time history analysis against more detailed NI10 model.

auxiliary and shield building and containment internal structures [NI20]

(including steel containment vessel, polar crane, RCL, and pressurizer) 3G-11 Revision 4

VEGP 3&4 - UFSAR Table 3G.1-1 (Sheet 2 of 4)

Summary of Models and Analysis Methods Analysis Type of Dynamic Model Method Program Response/Purpose Finite element Time history analysis SASSI Performed 2D parametric soil studies to help lumped-mass stick establish the bounding generic soil conditions and model of nuclear island to develop adjustment factors to reflect all generic site conditions for seismic stability evaluation.

Finite element Direct integration time ANSYS Performed 2D linear and non-linear seismic lumped-mass stick history analysis analyses to evaluate effect of liftoff on Floor model of nuclear island Response Spectra and bearing.

3D finite element Time history analysis SASSI Performed for the five soil profiles of firm rock, soft coarse shell model of Complex frequency rock, upper bound soft-to-medium soil, soft-to-auxiliary and shield response analysis medium soil, and soft soil.

building and containment internal To develop time histories for generating plant structures [NI20] design floor response spectra for nuclear island (including steel structures.

containment vessel, polar crane, RCL, and To obtain maximum absolute nodal accelerations pressurizer) (ZPA) to be used in equivalent static analyses.

To obtain maximum displacements relative to basemat.

To obtain SSE bearing pressures for all generic soil cases.

To obtain maximum member forces and moments in selected elements for comparison to equivalent static results.

3D shell model of Mode superposition ANSYS Performed to develop loads for seismic stability auxiliary and shield time history analysis evaluation.

building and containment internal structures [NI20]

(including steel containment vessel) 3G-12 Revision 4

VEGP 3&4 - UFSAR Table 3G.1-1 (Sheet 3 of 4)

Summary of Models and Analysis Methods Analysis Type of Dynamic Model Method Program Response/Purpose 3D shell of revolution Modal analysis; ANSYS To obtain dynamic properties.

model of steel equivalent static containment vessel analysis using To obtain SSE stresses for the containment accelerations from vessel.

time history analyses 3D lumped-mass stick - ANSYS Used in the NI10 and NI20 models.

model of the SCV 3D lumped-mass stick - ANSYS Used in the NI10 and NI20 models.

model of the RCL 3D lumped-mass stick - ANSYS Used in the NI10 and NI20 models.

model of the pressurizer 3D lumped-mass stick - ANSYS Used in the NI10 model.

model of the CMT 3D lumped mass Modal analysis ANSYS To obtain dynamic properties.

detailed model of the polar crane Used with 3D finite element shell model of the containment vessel.

3D lumped mass - ANSYS Used in the NI10 and NI20 models.

simplified (single beam) model of the polar crane 3D finite element shell Mode superposition ANSYS Used with detailed polar crane model to obtain model of containment time history analysis acceleration response of equipment hatch and vessel airlocks.

Equivalent static To obtain shell stresses in vicinity of the large analysis penetrations of the containment vessel.

3G-13 Revision 4

VEGP 3&4 - UFSAR Table 3G.1-1 (Sheet 4 of 4)

Summary of Models and Analysis Methods Analysis Type of Dynamic Model Method Program Response/Purpose 3D finite element refined Equivalent static non-linear ANSYS To obtain SSE member shell model of nuclear analysis using accelerations forces for the nuclear island island (NI05) from time history analyses basemat.

Mode superposition time To obtain floor and wall history analysis for the wall flexibility response and floor flexibility using characteristics.

synthetic time histories developed to match spectral To obtain maximum envelopes applied at the displacements relative to base basemat.

Response spectrum To obtain SSE member analysis with seismic input forces for the auxiliary and enveloping all soils cases shield building and the containment internal structures.

3D finite element coarse Mode superposition time ANSYS To obtain total basemat shell model of auxiliary and history analysis with seismic reactions for comparison to shield building and input enveloping all soil reactions in equivalent static containment internal cases linear analyses using NI05 structures [NI20] (including model.

steel containment vessel, polar crane, RCL, and pressurizer)

Quadrant model of shield Equivalent static analysis ANSYS To obtain member forces for building roof (See shield building roof and subsection 3.8.4.4.1 for The PCS tank is designed radial roof beams, air inlet information on use of the using the maximum structure, tension ring, and quadrant model.) accelerations at the PCS tank.

applicable elevation resulting from time history dynamic analyses of the nuclear island.

The tension ring and air inlet use maximum accelerations that are increased based on results of response spectrum analysis.

3G-14 Revision 4

VEGP 3&4 - UFSAR Table 3G.1-2 Summary of Dynamic Analyses and Combination Techniques Three Analysis Components Modal Model Method Program Combination Combination 3D finite element, fixed Mode superposition time ANSYS Algebraic Sum n/a base models, coupled history analysis auxiliary and shield building shell model, with superelement of containment internal structures (NI10 and NI20) 3D finite element nuclear Complex frequency SASSI Algebraic Sum n/a island model (NI20) response analysis 3D finite element, fixed Response spectrum ANSYS SRSS or Lindley-Yow base models, coupled analysis 100%, 40%, 40%

auxiliary and shield building and containment internal structures including shield building roof (NI05) 3D finite element model of Equivalent static analysis ANSYS 100%, 40%, 40% n/a the nuclear island basemat using nodal accelerations (NI05) from shell model 3D shell of revolution Equivalent static analysis ANSYS SRSS or n/a model of steel containment using nodal accelerations 100%, 40%, 40%

vessel from 3D stick model PCS valve room and Response spectrum ANSYS SRSS or Grouping or miscellaneous steel frame analysis 100%, 40%, 40% Lindley-Yow structures, miscellaneous flexible walls, and floors 2D stick model analyses Direct integration time ANSYS Algebraic Sum n/a with liftoff history 3G-15 Revision 4

VEGP 3&4 - UFSAR Table 3G.2-1 (Sheet 1 of 2)

Steel Containment Vessel Lumped-Mass Stick Model (Without Polar Crane) Modal Properties Effective Mass Mode Frequency X Direction Y Direction Z Direction 1 6.28 3.51 159.81 0.01 2 6.28 160.01 3.51 0.00 3 12.87 0.03 0.00 0.00 4 16.89 0.00 0.01 173.42 5 18.87 0.29 40.91 0.00 6 18.89 40.77 0.29 0.00 7 28.13 0.00 0.00 28.45 8 31.67 0.09 2.81 0.00 9 31.80 3.05 0.09 0.00 10 37.66 1.28 0.02 0.00 11 38.34 0.04 4.98 0.02 12 38.73 3.41 0.02 0.00 13 47.37 0.00 0.00 5.26 14 53.48 4.76 0.66 0.00 15 53.53 0.64 4.84 0.00 16 59.55 0.00 0.03 3.78 17 62.60 0.15 0.00 0.05 18 63.10 0.00 0.05 6.98 19 63.23 0.00 0.00 0.04 20 65.80 0.03 0.66 0.05 Sum of Effective Masses 218.06 218.69 218.05 Notes:

1. Fixed at Elevation 100.

2. The total mass of the containment vessel is 229.16 kip-sec2/ft.

3G-16 Revision 4

VEGP 3&4 - UFSAR Table 3G.2-1 (Sheet 2 of 2)

Steel Containment Vessel Lumped-Mass Stick Model (With Polar Crane)

Modal Properties Effective Mass Mode Frequency X Direction Y Direction Z Direction 1 4.20 0.00 78.15 0.00 2 4.98 149.97 0.00 0.17 3 6.44 6.80 0.01 23.59 4 6.47 0.00 112.05 0.00 5 8.00 35.09 0.00 1.19 6 8.61 0.00 4.06 0.00 7 12.58 0.03 0.18 0.00 8 16.00 0.06 0.08 0.22 9 16.09 5.16 0.00 146.06 10 17.21 24.58 0.00 36.27 11 18.86 0.00 40.88 0.00 12 20.64 13.30 0.00 0.71 13 23.74 0.00 0.40 0.00 14 27.04 0.13 0.00 15.64 15 29.62 2.88 0.00 3.61 16 30.21 0.00 0.07 0.00 17 31.62 0.00 2.92 0.01 18 32.85 0.58 0.00 0.38 19 34.02 0.28 0.01 4.87 20 36.97 0.20 1.61 0.00 Sum of Effective Masses 239.04 240.42 232.72 Notes:

1. Fixed at Elevation 100.

2. The total mass of the containment vessel with the polar crane is 261.02 kip-sec2/ft.

3G-17 Revision 4

VEGP 3&4 - UFSAR Table 3G.2-2 Comparison of Frequencies for Containment Vessel Seismic Model Vertical Model Horizontal Model Shell of Revolution Shell of Revolution Mode No. Model Stick Model Model Stick Model 1 16.51 hertz 16.97 hertz 6.20 hertz 6.31 hertz 2 23.26 hertz 28.20 hertz 18.58 hertz 18.96 hertz Note:

1. Fixed at elevation 100.

3G-18 Revision 4

VEGP 3&4 - UFSAR Table 3G.3-1 AP1000 ZPA for 2D SASSI Cases North-South Hard Firm Soft Soft Rock Rock Rock UBSM SM Soil Node El. feet ZPA [g] ZPA [g] ZPA [g] ZPA [g] ZPA [g] ZPA [g]

ASB 21 81.5 0.326 0.326 0.345 0.358 0.306 0.249 41 99 0.348 0.327 0.347 0.361 0.308 0.227 120 179.6 0.571 0.501 0.469 0.498 0.529 0.247 150 242.5 0.803 0.795 0.816 0.819 0.787 0.29 310 333.1 1.449 1.561 1.567 1.524 1.226 0.453 SCV 407 138.6 0.405 0.424 0.408 0.387 0.407 0.232 411 200 0.82 0.916 0.672 0.541 0.484 0.263 417 281.9 1.396 1.465 1.031 0.723 0.598 0.372 CIS 535 134.3 0.548 0.45 0.347 0.368 0.355 0.229 538 169 1.517 0.874 0.45 0.441 0.397 0.317 East-West Hard Firm Soft Soft Rock Rock Rock UBSM SM Soil Node El. feet ZPA [g] ZPA [g] ZPA [g] ZPA [g] ZPA [g] ZPA [g]

ASB 21 81.5 0.309 0.318 0.359 0.376 0.311 0.235 41 99 0.318 0.336 0.367 0.385 0.317 0.237 120 179.6 0.607 0.561 0.546 0.549 0.605 0.295 150 242.5 0.84 0.823 0.854 0.912 0.962 0.557 310 333.1 1.449 1.536 1.624 1.74 1.506 0.891 SCV 407 138.6 0.528 0.529 0.535 0.513 0.38 0.247 411 200 0.817 0.95 0.816 0.741 0.515 0.429 417 281.9 1.251 1.503 1.136 0.985 0.716 0.675 CIS 535 134.3 0.52 0.404 0.391 0.404 0.365 0.259 538 169 1.679 1.052 0.755 0.553 0.526 0.441 3G-19 Revision 4

VEGP 3&4 - UFSAR Table 3G.4-1 Key Nodes at Location Location General Area Elevation (feet)

CIS at Reactor Vessel Support Elevation SCV Center 100.00 CIS at Operating Deck SG West Compartment, NE 134.25 ASB NE Corner at Control Room Floor NE Corner 116.50 ASB Corner of Fuel Building Roof at Shield Building NW Corner of Fuel Bldg 179.19 ASB Shield Building Roof Area South Side of Shield Bldg 327.41 SCV Near Polar Crane SCV Stick Model 224.00 3G-20 Revision 4

VEGP 3&4 - UFSAR Table 3G.4-2 Maximum Bearing Pressure from 2D Time History Analyses East Edge West Edge Soil Case Analysis (ksf) (ksf)

Hard Rock Linear 17.18 32.77 Liftoff 17.38 34.85 Upper-bound Linear 19.46 31.69 Soft to Medium Liftoff 18.42 33.51 Soft to Medium Linear 15.84 30.82 Liftoff 17.06 32.18 3G-21 Revision 4

VEGP 3&4 - UFSAR 6ROLG0RGHORI$X[LOLDU\%XLOGLQJ

6KLHOG%XLOGLQJ PRGHOHGWR 6ROLG0RGHORI&RQWDLQPHQW

$X[%OGJURRIHOHYDWLRQ ,QWHUQDO6WUXFWXUH $[LV\PPHWULF)(0RGHORI

&RQWDLQPHQW9HVVHO &9

)(0RGHORI6KLHOG%XLOGLQJ )(6KHOO0RGHORI

5RRI &\OLQGHUDERYH$X[LOLDU\ )(6KHOO'\QDPLF0RGHORI &RQWDLQPHQW,QWHUQDO6WUXFWXUH

%XLOGLQJ5RRI(OHYDWLRQ &RXSOHG$X[LOLDU\ 6KLHOG &,6 6WDWLF$QDO\VHVWR

)UHTXHQF\DQG

%XLOGLQJ $6% GHILQH/XPSHG0DVV

0RGH6KDSHV 6WLFN0RGHO3URSHUWLHV

)UHTXHQF\DQG

)UHTXHQF\DQG 0RGH6KDSHV 0RGH6KDSHV &9/XPSHG

0DVV6WLFN

0RGHO

)(6KHOO'\QDPLF0RGHORI

&RXSOHG1,%XLOGLQJ6WUXFWXUHV

DQG0DMRU(TXLSPHQW 6HLVPLF7LPH

+LVWRU\$QDO\VHV

3ODQW'HVLJQ (TXLYDOHQW6WDWLF

5HVSRQVH6SHFWUD 6HLVPLF$FFHOHUDWLRQV

Figure 3G.1-1 Nuclear Island Seismic Analysis Models 3G-22 Revision 4

VEGP 3&4 - UFSAR Figure 3G.2-1 3D Finite Element Model of Coupled Shield and Auxiliary Building 3G-23 Revision 4

VEGP 3&4 - UFSAR Figure 3G.2-2 3D Finite Element Model of Containment Internal Structures 3G-24 Revision 4

VEGP 3&4 - UFSAR Figure 3G.2-3 3D Finite Element Model of Containment Outer Basemat (Dish) 3G-25 Revision 4

VEGP 3&4 - UFSAR Figure 3G.2-4 Steel Containment Vessel and Polar Crane Models 3G-26 Revision 4

ML21179A125

Text

Navigation menu

Search