ML11364A045
| ML11364A045 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 12/28/2011 |
| From: | St.Onge R Southern California Edison Co |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| TAC ME6179, TAC ME6180 | |
| Download: ML11364A045 (22) | |
Text
JSOUTHERN CALIFORNIA EDISON An EDISON INTERNATIONALv Company Richard 1. St. Onge Director, Nuclear Regulatory Affairs and Emergency Planning December 28, 2011 U. S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, DC 20555-0001
Subject:
Docket Nos. 50-361 and 50-362 Response to Request for Additional Information Regarding Use of American Concrete Institute (ACI) Reports for Restoration of Unit 3 Containment (TAC Nos. ME6179 and ME6180)
San Onofre Nuclear Generating Station, Units 2 and 3
Reference:
Letter from J. R. Hall (NRC) to P. T. Dietrich (SCE) dated October 14, 2011;
Subject:
San Onofre Nuclear Generating Station, Units 2 and 3 -
Request for Additional Information Regarding Use of American Concrete Institute (ACI) Reports for Restoration of Unit 3 Containment (TAC Nos.
ME6179 and ME6180)
Dear Sir or Madam:
By letter dated October 14, 2011, the Nuclear Regulatory Commission (NRC) issued a Request for Additional Information (RAI) (Reference) regarding use of American Concrete Institute (ACI) Reports for restoration of Unit 3 containment. The enclosure provides Southern California Edison's (SCE's) response.
The RAI letter requested a response within 30 days of receipt of the letter. NRC staff agreed by phone on December 14, 2011, that SCE may submit the response by December 30, 2011.
There are no new regulatory commitments contained in this letter. If you have any questions or require additional information, please contact Ms. Linda T. Conklin at (949) 368-9443.
Sincerely, P.O. Box 128 San Clemente, CA 92672 A-bol 00-K-
Document Control Desk December 28, 2011
Enclosure:
Response to Request for Additional Information Regarding Use of American Concrete Institute (ACI) Reports for Restoration of Unit 3 Containment cc:
E. E. Collins, Regional Administrator, NRC Region IV R. Hall, NRC Project Manager, San Onofre Units 2 and 3 G. G. Warnick, NRC Senior Resident Inspector, San Onofre Units 2 and 3
Enclosure Response to Request for Additional Information Regarding Use of American Concrete Institute (ACI) Reports for Restoration of Unit 3 Containment RAI I In Section 5.1 of Calculation C-257-01.04.05 (Reference 1), the licensee stated:
The new concrete mix for restoration of the containment opening will be tested to determine the compressive strength, the modulus of elasticity and the creep characteristics. However, the results will not be available at the time for the EOL [end of life] finite element analysis.
As such, the methods described in ACI 209R-92 and ACI 318-05 are used to estimate the relevant concrete properties. The moduli of existing and new concrete as well as creep and shrinkage will be used in the containment analysis to investigate the stress distribution around the opening after restoration.
The licensee's position on the use of ACI 209R-92 is further summarized as follows:
The ACI 209R Report is a widely recognized guidance document that provides a simple, yet reasonably accurate methodology for estimating creep and shrinkage design values. For the SONGS containment structure, the use of such estimated values has been further justified and validated by comparison to long term creep and shrinkage test results performed on the actual concrete mix used to restore the temporary construction opening.
Please provide the above stated comparison of the concrete properties (creep, shrinkage, elastic modulus) obtained from tests of the actual concrete mix used for the restoration of the steam generator replacement construction opening to those used in the analysis based on estimates using methods in ACI 209R-92, "Prediction of Creep, Shrinkage, and Temperature Effects in Concrete Structures," that would demonstrate that the properties used in the SONGS containment analysis are comparable or conservative relative to those obtained from the tests.
In establishing values of creep and shrinkage, please indicate how any important differences in the environment between the test samples and the actual concrete in the structure, if any, were considered.
Response to RAI 1 The creep and shrinkage study report (Ref. 8) for the new concrete, prepared by the CTL Group of Chicago, provides tested data that were developed for loading at both 7 and 28 days. The dates of the concrete pours and the dates when the vertical and hoop tendons were tensioned are provided for Units 2 and 3 in Attachment 1 to this RAI response.
Page I of 18
Creep:
In Ref. 1, estimates of creep were made assuming loading 7 days after the concrete placement. to this RAI shows that the average number of days was actually 14 days for Unit 2 and 9 days for Unit 3. Considering this, the comparison of creep coefficients obtained from tests of the actual concrete mix to that used in the analysis can be established as follows:
Creep coefficient from tests One year with loading at 7 days (VIy) 7 = 1.235x10"6 (Ref. 8)
One year with loading at 28 days (vly*)28 = 0.93x 10-6 (Ref. 8)
One year with loading at 9 days (vlyr)9 = 1.180x10-6 End-of-life with loading at 9 days VEOL = 1.52x106 Creep coefficient in Ref. 1 VEOL = 1.20x 10-6 The increase in the creep coefficient will increase the creep strains by the ratio 1.52/1.2 = 1.27. The corresponding increase in the creep losses given in Section 8.3.2.3 of Ref. 1 will be about 0.1 ksi, which is negligible compared with the average tendon stresses of about 170 ksi, per Tables 16 and 17 of Ref. 1.
Shrinkage:
The shrinkage loss that takes place prior to placement of the load on the concrete is irrelevant to prestress losses. Since the loading was applied in about 9 days after concrete placement (Unit 3), the comparison between the shrinkage strain from tests and that used in the analysis can be summarized as follows:
Shrinkage strain from tests Ultimate shrinkage strain for a specimen loaded at 7 days:
(Es)
= 288x10-6 in/in (Ref. 8)
Shrinkage after 2 days for specimen loaded at 7 days:
(esh)7 = 72x 10-6 in/in (Ref. 8)
Ultimate shrinkage for a member loaded at about 9 days:
(s~h)- = 216x 10-6 in/in Shrinkage strain in Ref. 1 (sh),= 117 x10-6 in/in Page 2 of 18
This is an increase of about 85% over the value used in Ref. 1. The prestress losses due to shrinkage were less than 1 ksi (Tables 16 and 17 of Ref. 1). The net increase in the prestress losses due to increase in shrinkage will be less than 0.25 ksi. Therefore, the actual prestressing loss due to shrinkage as determined by the tests has negligible impact on the design.
Elastic modulus:
The concrete modulus of elasticity at 28 days can be calculated as follows:
(Ev) 28 = ace,/ = (2100 psi)/(463 x 106 --2x 10-6) = 4555 ksi (from Ref. 8)
This is an increase of about 3% over the value used in Ref. 1, which has negligible effect on analysis results.
The creep and shrinkage test report for the original containment construction (Ref. 3) indicates that a seal was provided to the concrete cylinders as shown below:
- 2. 0 TEST PROGRAM (Excerpt from Ref. 3)
The test program comprises the evaluation of the following properties on two concrete mixes, one with 3/4 in., and other with 1 1/2-in, maximum size aggregate.
Both of these mixes are designated for fc = 6000psi @ 90 days.
2.1Compressive strength to be determined on three 6-in. by 12-in. sealed concrete specimens, stored at 730 F. at ages of 7, 28, 90, 180 and 365 days.
2.2Modulus of Elasticity and Poisson's Ratio to be determined on three 6-in.
by 12-in. sealed concrete specimens, stored at 73 0 F, at ages of 28, 180 and 365 days.
2.3Coefficient of Thermal Expansion to be determined on two 6-in. by 16-in.
sealed concrete specimens, stored at 730 F, at ages of 28, 90, 180 and 365 days.
2.4Diffusivity to be determined on two (total of four) 8-1/2-in. by 17-in.
sealed concrete specimens, stored at 730 F, at age of 90 days.
2.5Creep Characteristics of sealed concrete specimens to be determined at a sustained stress of 2100 psi initially applied at ages of 28, 180, and 365 days. The autogenous strains changes for specimens stressed at ages of 28 and 180 days shall be determined from sealed creep specimen that will be stressed at age one year. Changes in autogenous strains are small after the age of one year; therefore, no corrections of autogenous strains will be applied to creep specimens stressed at one year. The creep tests shall be carried out at 73 0 F.
Each creep test shall be conducted on a set of two 6-in. by 16-in. sealed concrete specimens.
The CTL report (Ref. 8) for creep and shrinkage tests applicable to concrete in the restored containment opening follows similar requirements for sealing concrete cylinder specimens from ASTM C 512 (Ref. 4) to prevent loss of moisture throughout the period of storage and testing.
As the thickness of the containment wall is large and the existence of the liner completely prevents any moisture loss from the inside face, the restored concrete in the structure will have insignificant moisture loss. Therefore, the environmental difference between the test samples and the actual concrete in the structure is minimal.
Page 3 of 18
RAI 2
Please justify why it is acceptable to apply the methodology in Section 4-1 of the ACI 224.2R-92 report, concerning the axial stiffness of one-dimensional members due to cracking in reinforced concrete caused by direct tension, to account for cracking in: (a) prestressed concrete, and (b) more complex systems such as post-tensioned containments, for the end-of-life evaluation of the restored SONGS containments in Calculation No. C-257-01.04.06 (Reference 2).
Response to RAI 2 The methodology in Section 4-1 of the ACI 224.2R-92 report, concerning the axial stiffness of one-dimensional members due to cracking in reinforced concrete caused by direct tension, is acceptable to account for cracking in prestressed concrete provided that prestress forces and tendons are properly considered in the design of members. In other words, a prestressed concrete member can be treated as a reinforced concrete member if prestress forces are modeled as another load. (See Section 4 of Ref. 13 for general discussion on the subject.) Accordingly, in Calculation No. C-257-01.04.06 (Ref. 2) the prestress forces were considered as an external force, F, and the existence of tendons was included by accounting their tributary area in the calculation of the stiffness of a cracked member, as further shown in Response to RAI 3.
Using the ACI 224.2R-92 methodology to account for cracking in complex systems, such as post-tensioned containments, is justified by the observed behavior of containments during pressurization, which is the governing condition in this analysis. During pressurization, such as the integrated leak rate testing performed at SONGS after restoration of containment (Ref. 9 through Ref. 12), portions of containment structures subject to cracking typically develop hairline cracks that are primarily oriented in the hoop and vertical directions. This behavior confirms the one-directional response of the structure.
The complexity of the structure (including removal of a number of tendons, cutting an opening while under partial prestress, repairing the opening, post-tensioning the replaced tendons, etc.) indicates that a more complex analytical model may provide improved results for the end-of-life evaluation.
However, such a complex approach was not deemed necessary to obtain a solution to the problem since (a) reasonable assumptions were made in modeling and application of loads, (b) conservative approximations were applied to maximize the critical design forces, and (c) checks were made at intermediate steps to validate the approach.
Also note that the original design calculation of the SONGS containment employed the same one-dimensional modeling methods. Where preliminary analysis indicates potential for cracking, the Updated Final Safety Analysis Report (UFSAR) method of evaluation for containment analysis, which is contained in UFSAR Subsection 3.8.1.4 (Ref. 14) and further detailed in Bechtel Topical Report BC-TOP-5, Rev. 1 (Ref. 5), requires the potential for load redistribution due to concrete cracking to be considered by adjusting the analytical model. Specifically, the method of evaluation of reducing the concrete modulus of elasticity in areas subject to cracking is consistent with the original calculations. The resulting forces are then combined in accordance with the UFSAR load combinations, and the design is carried out using one-dimensional elements in the hoop and vertical directions. This is the approved method for the original SONGS 2 & 3 containment analysis and design.
In summary, the repair design incorporating the methodology of ACI 224.2R-92 is appropriate, consistent with the approved method of evaluation for SONGS 2 & 3, and was independently reviewed by industry consultants experienced in containment analyses and similar SGR projects. All the results were Page 4 of 18
subjected to reasonableness, completeness, adequacy and appropriateness tests that are an integral part of the nuclear safety-related structural design practice.
RAI 3
Appendix H of Calculation No. C-257-01.04.06 (Reference 2) describes the methodology and criteria used, based on Equations 4.12 and 4.13 of the ACI 224.2R-92 report, to estimate a reduced concrete sectional stiffness to account for cracking in the restored containment opening area, in the ANSYS shell-element-based linear elastic finite element model of the SONGS containments.
With regard to the application of Equations 4.12 and 4.13 of ACI 224.2R to calculate the effective cross-sectional area, Ae, of a cracked member in the above calculation, please provide the following information:
(a)
For both the hoop and vertical directions, was the cross-sectional area of prestressing tendon steel included in the calculation of Ag, A, and Acr? If not, please provide a supporting justification.
Response to RAI 3(a)
Yes, the cross-sectional area of tendons was included in the calculation of Ag, A. and A,,.
(b)
For both hoop and vertical directions, was the gross cross-sectional area, Ag, replaced with the transformed area, At=Ag + (n-I)As, to include the contribution of bonded reinforcing steel and unbonded prestressing steel in the post-tensioned containment? If not, please provide a supporting justification.
Response to RAI 3(b)
Yes, the gross cross-sectional area, Ag, was replaced with the transformed area, A, = Ag+(n-1)As, to include the contribution of both bonded reinforcing steel and unbonded prestressing steel in the post-tensioned containment.
(c)
How was the cracking load, Pmr, calculated for the hoop and vertical directions?
Please identify what values of Por were used for the hoop and vertical directions.
Please indicate the material property threshold (such as tensile strength) that was used to determine the cracking load.
Page 5 of 18
Response to RAI 3(c)
Pr is calculated for the hoop and vertical directions using Equation (2.1) of ACI 224.2R (Ref. 7),
which is shown below:
P. = (1 - p + np)Agft' in which p is the reinforcing ratio, A/Ag; A8 is the area of reinforcing steel plus tendons; Ag is the gross cross-sectional area; and n is the ratio of modulus of elasticity of the steel to that of concrete. The direct tensile concrete strength, f,', is used in this expression, which can be calculated as 4qIf,' (, 0.334(150fc'))
(per Equation (3.2) of ACI 224.2R).
The calculation details and material properties that were used to determine the cracking load, Per, in each direction are shown below:
Vertical direction Hoop direction A, = 5.22in2/ft + 3.12in2/ft = 8.34in2/ft A, = 4.80in2/ft + 5.42in2/ft = 10.22in 2/ft (see Attachment C.2 and Appendix I of Ref. 2)
(see Attachment C.2 and Appendix I of Ref. 2)
Ag = 624in2/ft Ag = 624in2/ft p = 0.01 34 p = 0.0164 n = 6.57 n = 6.57 ft' = 4qfo' = 310 psi ft'= 4/f'= 310 psi P, = 208kip/ft P;, = 21 lkip/ft Note that the comparison with the cracking load without including reinforcing steel and tendons, that is Pr = Agft' = 193kip/ft, with values presented in the table above suggests that the contribution of reinforcing steel and tendons to the cracking load is small because the strain level at the cracking load is also small.
(d)
Please provide a numerical example of all steps (with all inputs used) of a typical calculation (e.g., for the ratio, (EA)ANSYs/EAg = 0.4 or 0.6) that was performed to develop a data point (one in hoop direction and one in vertical direction) in Figure H.2 of Reference 2.
Response to RAI 3(d)
The data point in Figure H.2 of Ref. 2 is obtained using the following procedure:
(1) Perform ANSYS analysis with an assumed effective stiffness, Ec.
(2) Measure the axial strain for the selected load combination along the desired direction, CANSYS.
(3) Find the effective stiffness corresponding to the axial strain obtained from ANSYS, E,.
The numerical example for (EAe)*Asys/EeAg = 0.4 is shown below:
Page 6 of 18
- 1. Vertical direction ANSYS vertical strain, vransys, of (D+F+1.5P) at restored area: E, = 0. 4EC EC:= 57.*
psi *i = 4415.ksi Ec.ansys := 04EC Sv.anys := 0.00022 Calculation of cracking load
.2 Ag = 624w
.2 As := 8.34 m ft A8 p:=
0.0134 Ag Es := 2900Cksi n:= -! = 6.568 Ec
- 4.
i.pi 31.psi psi P 1 := (1 - p + n.p).Ag-. = 208.k-Lp ft Uncracked section stiffness EA.t
- = EB.Ag = 2755085k'-
ft Tensile strain at cracking Pe cr := --
= 0.000075 EAtm Vertical tensile strains corresponding to E, = 0.4E0 from ANSYS Eq (3.2) of ACI 224 forw. = 150 lb/ft3 Eq (2.1) of ACI 224 Page 7 of 18
ACt 224 Calculations to find EAe corresponding to the concrete strain of ev.ansys Assume an axial load P:=250-L'P
> Pcr= 208.-
Cracked section ft fr Ar : n*A, Definition given in ACl 224 Ae "=
+.L.
= 381.m EA,:= Ec.A, = 16 8 36 3 0.kip ft EAe B-Ae
-~n= 0.611 EA,
- -Ec Eq (4.13) of ACI 224 P
= 0.00015 Eq (4.12) of ACI 224 Fe Therefore, the effective stiffness of 0.611EcA can be related to the concrete strain of Fv = 0.00015.
By repeating the calculation for different axial loads, P, the relationship between the concrete strain and the corresponding effective stiffness can be obtained. The following figure shows the resulting relationship.
1.2 1
0.8 EAe/EAun 0.6 0.4 0.2 224) 0 0
0.001 0.002 0.003 0.004 Concrete strain Fig. 1 Relationship between concrete strain and effective stiffness: Vertical direction Page 8 of 18
- 2. Hoop direction ANSYS hoop strain, 8 h.ansy., of (DiF+1.5P) at restored area: Ee = 0. 4Er Sh.asys:= 0.00033 Hoop tensile strains corresponding to Ee = 0.4E0 from ANSYS Calculation of cracking load
.2 A.:= 10.22--ft P
= 0.0164 A8 Pcr:= (1 - p + n-p)*Ag 211. ft Eq (2.1) of ACI 224 ACI 224 Calculations to find EAe corresponding to the concrete strain of Eh.,nsys Assume an axial load P:= 250-IP
> per= 211.LIP ft ft Acr:= n.As Cracked section Definition given in ACI 224 Ae :=Ag. Pe
+ Ncr[
EAe := Ec.Ae = 1773981. -kp ft
.2
= 402-- ft Eq (4.13) of ACI 224 EAe
. = 0.644 EAM0 Ect:
EA.~ -E P
E V:= -
= 0.00014 EAe By repeating the calculation for different axial loads, P, the relationship between the concrete strain and the corresponding effective stiffness can be obtained. The following figure shows the resulting relationship.
1.2 1
0.8 EAe/EAun 0.6 0.4 0.2 0
224) 0 0.001 0.002 0.003 0.004 Concrete strain Fig. 2 Relationship between concrete strain and effective stiffness: Hoop direction Page 9 of 18
From Figures 1 and 2, the ratio of effective stiffness in ACI 224.2R to that of ANSYS can be calculated as follows:
Vertical direction: (Ec)AC224/(E)ANsys = 0.44/0.40 = 1.10 Hoop direction: (Ec)ACJ24/(Eo)ANsys = 0.34/0.40 = 0.85 Average difference in both direction: (L.10+0.85)/2 = 0.98 These points are reported in Figure H.2 of Calculation C-257-01.04.06 (Ref. 2).
RAI 4
From Section 8.1.2.2, "Cracked Conditions," and Appendix H of Calculation C-257-01.04.06 (Reference 2), it appears that the same value of reduced concrete section stiffness (AeEc) of 0.4EcA% was used in the model for each of the load combinations III, IV, and VI.
Please confirm if this is true. If so, please provide the justification for using the same value for all the load combinations, considering the fact that the axial strains and the extent of concrete cracking, and therefore the sectional stiffness, is a function of the magnitude of the forces due to the applied loads.
Response to RAI 4 A single value of 0.4EcAg was used in the model for load combinations EI, IV and VI. The reduced section stiffness of O.4EcAg was derived based on load combination IH (D+F+1.5P), for which the most severe cracking condition is expected as explained below.
The containment structure is under bi-axial compression under operating conditions and therefore, full stiffness is expected throughout containment wall and dome due to continuity of the prestressing system. Since the construction opening is away from any discontinuities, the membrane forces will be the dominant factor in the behavior of the wall in this area. In case of a LOCA accident, internal pressure decreases the membrane compression in the structure. The maximum decrease in membrane compression will occur under the loading combination that includes 1.5P. Since the prestressing tendons are continuous, the membrane forces are expected to remain constant throughout the wall and dome.
However, it is conceivable that membrane compression may be reduced in the construction opening area, resulting in membrane tension, thus leading to small cracks. This in turn may lead to re-distribution of the internal forces. For this reason, the analysis was performed to determine the maximum possible reduction in stiffness and the consequent re-distribution of internal forces.
The phenomenon described above is best understood from a review of the figures in Appendix G of Ref. 2 which illustrate the state of stress in the repair and surrounding areas as the work progresses.
Figure G. 1 shows the construction sequence, from (a) to (f), Figures G.2 and G.3 show the corresponding state of stress in the vertical and horizontal directions, respectively (these figures are not to scale, i.e., the Page 10 of 18
forces shown are qualitative only; however, these figures are based on ANSYS analyses). The following may be observed from a review of Figure G.2:
" Fig. G.2 (b) shows that membrane compression in the vertical direction is zero under dead load once the opening is cut (Fig. G. 1, stages (c) through (f)). Membrane compression due to dead load is never restored in the repair area.
Fig. G.2 (c) shows that vertical membrane force is reduced when partial tendon de-tensioning is achieved (Fig. G. 1 (b)), and it is zero when the opening is cut (Fig. G. I (c)).
Fig. G.2 (d) shows that as creep and shrinkage take place, membrane compression is reduced in the repair area and increased in the adjacent areas (i.e., transfer of membrane forces).
Fig. G.2 (e) shows the final state of stress in the area corresponding to Fig. G. 1 (e). As shown in this figure, the membrane compression in the repair area is significantly less than the adjacent wall segments due to the reduced stiffness. If the stiffness in the repair area were assumed to be zero, the membrane compression would be zero.
" Fig. G.2 (e) indicates that membrane compression adjacent to the repair area is higher than it was before the opening was made. Also, it shows that, after the completion of the prestressing of the replacement tendons, the membrane compression will not reach to the level of the undisturbed containment wall. If the internal pressure is applied in the condition of Fig. G.2 (e), it is easy to visualize that the membrane compression in the opening area will become tension and therefore, it is prudent to consider the possibility of cracking in this area.
As the above summary of the analysis procedure implies, one goal of the analysis methodology was to determine maximum possible increase in the membrane forces and moments in areas surrounding the opening so that adequacy of the existing design can be demonstrated. Since load combination Ed generates the maximum membrane tension due to internal pressure, it is expected to provide an upper-bound design condition in the surrounding area, in conjunction with a large decrease in axial stiffness (from 1.0 down to 0.4). Therefore, the calculation of the reduced stiffness based on the load combination III is adequate.
Page 11 of 18
RAI 5
The methodology used in the parametric study in Appendix H of Calculation C-257-01.04.06 (page 84 of Reference 2) and the ANSYS containment analyses accounting for cracking is based on the assumption that the reduced effective axial stiffness (AeEc) for the hoop and vertical directions are equal. Please justify this assumption considering the fact that the degree of cracking is likely to not be the same in the two directions.
Response to RAI 5 The SONGS UFSAR description of the method of evaluation for the containment structure is contained in UFSAR 3.8.1.4 (Ref. 14) with detailed instructions contained in BC-TOP-5, Rev. 1 (Ref. 5).
The analysis "consists of two parts, the overall analysis... and the local analysis." The overall analysis employs an axisyrnmetric model of the containment structure that takes advantage of the basic radial symmetry of the structure about the vertical axis to reduce the model size. The overall analysis, however, does not account for non-symmetric features such as buttresses, penetrations, brackets, and liner plate anchors. These features are considered in the local analysis, which employs a variety of evaluation techniques, depending on the specific feature, including testing (for tendon anchorages), computer programs (for large penetration openings, such as the equipment hatch), and manual calculation methods (for small penetration openings and anchors).
The temporary SGR opening meets the UFSAR definition of a large penetration: "having an inside diameter equal to or greater than 2.5 times the nominal shell thickness." Even though the concrete and tendons are ultimately restored, the temporary SGR opening introduces a permanent, non-symmetric redistribution of prestress loads similar to, although not as pronounced as, a large penetration opening.
Calculation C-257-01.04.06 (Ref. 2) was performed to account for the permanent effects of the restored temporary SGR opening. This re-analysis of the containment was performed using the methods for evaluating large penetration openings contained in UFSAR 3.8.1.4 and BC-TOP-5, Rev. 1.
Calculation C-257-01.04.06 (Ref. 2) does not supersede the overall analysis or various local analyses of the original containment analysis. Instead, this calculation provides a supplemental local analysis that accounts for redistribution of stresses in the areas within and surrounding the restored temporary SGR opening. The goal of this calculation was to evaluate the restored temporary SGR opening following the UFSAR-described methods of evaluation and applying the same structural acceptance criteria used in the original containment design. In that case, it would not have been appropriate to apply methods and acceptance criteria different from SONGS original containment design.
The analysis was done by considering a severe condition in an approximate, yet conservative manner. The restored area will find the equilibrium conditions since the more it cracks, the more the forces will be redistributed to the surrounding area. As such, the critical area for checking the structural integrity of the containment is the surrounding area of the temporary opening. The analysis approach taken in Calculation C-257-01.04.06 (Ref. 2) was to use a reduced stiffness value which will provide Page 12 of 18
reasonable upper-bound estimate for both hoop and vertical directions while providing conservative results.
It is also important to note that this concept of reducing the concrete stiffness for both hoop and vertical directions simultaneously by reducing concrete modulus of elasticity is consistent with methodology provided in Section 7.2.1.2 and 7.2.1.4 of BC-TOP-5, Rev. 1 (Ref. 5).
RAI 6
(a)
The ANSYS parametric analyses in Appendix H of Reference 2 used the same effective axial stiffness [(EA)ANSYS] values for the hoop and vertical directions (see assumption described in RAI 5). However, the criterion used in Appendix H (page 84 of Reference 2) to determine the convergence of the effective sectional stiffness values between the parametric ANSYS analyses and the ACI 224.2R-estimated values [for the two directions] does not seek to satisfy nor does it satisfy the assumption that the effective stiffness in the two directions are considered equal. Instead, it averages the ACI 224.2R-estimated vertical and horizontal effective stiffness (see Figure H.2 in Appendix H of Reference 2). The average curve so obtained intersects the line representing the ratio (EcAe)AC1224/(EAe)ANSYS =1 at two points corresponding to the ratio, (EOAe)ANSYS/EAg, of 0.4 and 0.7. It can be noted from Figure H.2 that the ACI 224.2R-estimated effective stiffness are not equal in the two directions for both of these values. The smaller of the two values (with no explanation provided), 0.4EA, was selected as the reduced effective stiffness and was used for the containment opening area in the concrete cracking analysis, even though the larger value would occur earlier when cracking occurs.
Please explain the basis for the criterion used to determine the effective stiffness value with regard to the SONGS containment analysis.
Response to RAI 6(a)
Calculation C-257-01.04.06 (Ref. 2) provides a supplemental local analysis that accounts for redistribution of stresses in the areas within and surrounding the restored temporary SGR opening. The critical area for the structural integrity is the surrounding area. As such, the use of smaller reduced stiffness provides the more conservative estimate for design requirements.
Page 13 of 18
(b)
Assuming that the data and assumptions in Figure H.2 are correct, it appears that the appropriate criterion to be used to determine the converged value of the effective axial stiffness between the ANSYS parametric study and the ACI 224.2R-estimated values should be the value of (EcAe)ANSYS for which:
[(EdA)AC1224/(E:Ae)ANSYS]Hoop = [(EcAe)AC1224/(E.Ae)ANSYSvertica
=
1 This criterion also satisfies the assumption that the effective stiffness in the two directions are equal. These ratios for the two directions are not expected to converge exactly to 1 because of the approximations in the 1-dimensional ACI 224.2R method relative to the 3-dimensional ANSYS parametric analyses, but would likely be roughly close to 1.
Accordingly, from Figure H.2 on page 84 of Reference 2, the converged value of the reduced effective axial stiffness to be used in the SONGS containment analysis would be the value corresponding to the intersection of the vertical and hoop curves, which is 0.55EA, with the ratio (EcAe)AC1224/(EcAe)ANsys being approximately 0.9 (close to 1).
Regarding this approach, please address the impact of the noted difference in the effective stiffness value on the SONGS end-of-life containment analyses, while also considering the questions raised in all of the other RAls. Alternatively, please justify why the value of effective stiffness used (0.4E,%) by the licensee is appropriate, considering the issues raised in paragraph (a) above and in all of the other RAls, as applicable.
Response to RAI 6(b)
Please refer to the responses given for RAI 4 and 5. The correct value of effective stiffness may vary depending on the directions and the load combination used. As stated in response to previous questions, the critical design goal here is to maximize the redistribution of the forces to the surrounding area to assure design adequacy in case the stiffness is reduced in the repair area. The greater the reduction in effective stiffness value, the greater the re-distribution. The effective stiffness factor of 0.4 is a reasonable lower-bound value. Further reduction in stiffness would lead to unrealistic design requirements. A higher value of the reduced stiffness would result in lower forces in the surrounding area.
Therefore, the use of the lowest possible effective stiffness (0.4EcAg) will maximize the redistribution of the forces to the surrounding area to assure design adequacy.
Page 14 of 18
RAI.7 Assuming that the forces and moments at the concrete sections expected to be cracked, obtained on the basis of the uncracked ANSYS analysis, are reacted entirely by the combination of unbonded prestressing tendons and bonded reinforcing steel, please provide the following information for each of the hoop and vertical directions for the critical load combinations in the SONGS containment EOL analysis:
(a) the maximum tensile stress in the prestressing tendons, (b) the maximum tensile stress in the reinforcement for the primary forces in the load combination, (c) the maximum tensile stress and the maximum strain in the bonded reinforcement for the combined primary and secondary forces in the load combination, and (d) the maximum stress and strain, as appropriate, in the liner (please indicate if tensile or compressive).
Response to RAI 7 The following table shows the requested information for each direction:
Vertical direction Hoop direction (a) maximum tensile stress in the prestressing tendons 183 ksi 177 ksi (b) maximum tensile stress in the reinforcement 47 ksi 52 ksi for the primary forces (c) maximum tensile stress and the maximum strain 51 ksi 48 ksi in the bonded reinforcement for the combined primary (1745x10 6 in./in.)
(1655x10 6 in./in.)
and secondary forces (d)
-22 ksi 24 ksi (= fy,limer) maximum stress and strain in the liner
(-775x106 in./in.)
(2207x 10-6 in./in.)
Note:
(+) Tension
(-) Compression fyxmer = Yield strength of the liner
RAI 8
Concrete cracking could also result in reduction in flexural stiffness and shear stiffness that could contribute to redistribution of moments and forces, which have not been considered in the SONGS analyses accounting for concrete cracking in Reference 2. Therefore, please provide Page 15 of 18
the justification as to why the end-of-life evaluation of the SONGS containment following steam generator replacement in Reference 2 selectively considered only reduction in axial tensile stiffness, and resulting redistribution of tensile membrane forces, due to concrete cracking.
Response to RAI 8 The stiffness study described in Ref. 2 and the resulting "effective stiffness" was based on axial behavior of a one-dimensional concrete element. The effective stiffness was modeled simply using a reduced modulus of elasticity.
Using this method, the axial stiffness will be proportional to:
cR.
oc ExA and bending and shear stiffnesses will also be proportional to:
Rbending CC E~xI Rehear oc Er/2(1+v)xA where E, is the effective modulus, A is the section area and v is the Poisson's ratio. Thus, the flexural stiffness and shear stiffness are reduced by the same "effective modulus" ratio. The analyses results obtained by using the 0.4 effective stiffness values were included in Ref. 2. In the following, the effects of reduced stiffness in the repair area will be examined for both flexure and shear.
Flexure As shown in Figures 9 and 10 of Ref. 2, the bending moments in both horizontal and vertical directions are reduced compared with the original design values. As expected, the bending moments in the surrounding areas are increased, as shown in Figures 11 and 1 A. Tables 12 and 12A indicate the same trend in terms of element stresses. Therefore, the goal of maximizing the axial forces and bending moments in the surrounding area is realized.
It is recognized that the reduction in bending stiffness may be greater than implied by the above approximation. However, the primary parameter affecting the repair area behavior is the axial force and not the bending moment. Therefore, the potential uncertainty introduced by the above approximation is insignificant. It is also important to note that the conservative approximations made in the analyses to maximize the cracking predicted by the analysis in the opening area. Review of Figures 9 and 10 in Ref. 2 indicates that the resulting design forces under all load combinations are entirely within the allowable interaction diagram. Similarly, review of Tables 12 and 12A in Ref. 2 shows that the maximum stresses, which represent the stresses at a point in the most critical element (without any averaging), are far less than the allowable values.
Therefore, the simplified stiffness modeling used in the repair area analysis is adequate and conservative.
The adequacy of the repair area was further justified by the results of the crack mapping programs for both Units 2 and 3 after the repairs, during the Integrated Leak Rate Test (ILRT). The results of the crack mapping are included inRef. 9 through Ref. 12. Review of the ILRT test data lead to the following conclusions:
In all cases, the maximum measured crack width in the repair area was less than 0.013" or 0.33 mm, Page 16 of 18
Any cracks that appeared when the structure was under the ILRT pressure were less than 0.2 mm,
" Cracks that appeared during the pressurization closed upon completion of the test,
- Nothing unusual or significant was observed during these tests.
The above observations confirm the statements that the containment analyses for SONGS SGR were performed with assumptions that resulted in a conservative design in the repair area.
Shear In the case of shear, the in-plane and out-of-plane results should be considered separately:
Tangential (In-Plane) Shear:
Since the containment is an axisymmetric structure and the repair area is away from any discontinuities, the only source of significant in-plane shear is the earthquake load. This is the "tangential shear" and has been evaluated in Section 8.1.2.3 of Ref. 2. The maximum design forces as calculated in that reference are summarized below:
Reinforcement Reinforcement Demand/Capacity Direction Load Combination Demand, in2/ft Provided, in2/ft Ratio Vertical IV 3.9 5.2 0.75 Hoop VI 2.6 4.8 0.54 If it is assumed that the wall in the repair area is ineffective in resisting any tangential shear, the demand capacity ratio in the remainder of the wall can be approximated by:
(As)dend = L/(L - L.) = (nt x 77ft) / (7c x 77ft - 32ft) = 1.15 where L is the one-half containment perimeter and Lo is the width of the opening area.
Thus, the demand in the remaining segment of the wall would be increased by about 15%.
D/C = 0.75 x 1.15 = 0.86 The above simple check shows that the containment wall has sufficient capacity to resist the tangential shear that may be imposed on it by the design basis earthquake.
" Out-of-Plane Shear:
In the case of the out-of-plane shear, the shear stresses are very small in the repair and surrounding areas due to continuity of the containment structure. Stiffness reduction in the repair area will have negligible effects on out-of-plane shear stresses in the restored and surrounding areas.
In addition, radial ties (#8 @ 1') were provided in the repair area, following the original rebar drawing of the containment wall. As such, radial ties are placed at repair and surrounding areas to resist any radial tension.
Page 17 of 18
References:
- 1. Calculation No. C-257-01.04.05, ECP No. 061200409-6, RO, Evaluation of Restored Containment - Concrete Modulus Ratio and Tendon Retensioning Forces, SONGS, Unit 2 and Unit 3.
- 2. Calculation No. C-257-01.04.06, ECN/Prelim CCN No. D0020134, Evaluation of Restored Containment End-of-Life Analysis, SONGS, Unit 2 and Unit 3.
- 3. UCB/SESM-1979/05, "Studies of Concrete for San Onofre Nuclear Power Plant Containment Structures, Units 2 & 3," Structural Engineering Laboratory, University of California, Berkeley, California, 54 pp.
- 4. ASTM C 512-02, "Standard Test Method for Creep of Concrete in Compression."
- 5. BC-TOP-5, Revision 1, "Prestressed Concrete Nuclear Reactor Containment Structures," Dec.
1972.
- 7. ACI 224.2R-92, "Cracking of Concrete Members in Direct Tension."
- 8. CTL Group, 25221-000-HC4-SYO0-0002, "Final Report for ASTM C 512 Creep and Shrinkage, San Onofre Nuclear Generating Station Units 2 and 3, Steam Generator Replacement Project," 5 PP.
- 9. Inspection Report S023-XXIV-3.8.3, Rev.0, "In Process Visual Examination of the Temporary Containment Opening - SONGS Unit 2," 14 pp.
- 10. Examination Report S023-XXIV-3.8.1, "Visual Examination of the Containment Construction Opening Concrete Surface prior to ILRT - SONGS Unit 3," 16 pp.
- 11. Examination Report S023-XXIV-3.8.1, "Visual Examination of the Containment Construction Opening Concrete Surface at ILRT Test Pressure-SONGS Unit 3," 8 pp.
- 12. Examination Report S023-XXIV-3.8.1, "Visual Examination of the Containment Construction Opening Concrete Surface After the ILRT - SONGS Unit 3," 6 pp.
- 13. Collins, M.P. and Michell, D., "Prestressed Concrete Structures," Response Publications, Canada, 1997, 766 pp.
- 14. San Onofre Units 2 and 3 Updated Final Safety Analysis Report (UFSAR), Rev.22.
Attachment(s):
- 1. Summary Sheets of Tendon Retensioning Activity for Units 2 and 3, 2 pp.
Page 18 of 18
- 1. Date for concete pouring:
- 2. Tendon retenslaning 12/1912009 (Ref. doc: 1)
Vertical Tendon ID Date 48-104 NA 49-103 NA 50 -102 12/31/2009 59-3 1/3/2010 60-2 11312010 61-1 1/312010 62-180 1/112010 63-179 1/1)2010 64-178 1/112010 65-177 1/4J2010 66-176 1/4/2010 67-175 1/4J2010 68-174 1/j/2010 69-173 1/1/2010 70-172 1/2/2010 71 - 171 1/4/2010 72-170 1/4/2010 73-169 1/4/2010 74 - 168 1/2/2010 75 - 167 1/2/2010 7 -
168 1/2/2010 77-166 1/4/2010 78 - 164 1/4/2010 70 - 163 1/4/2010 80 - W62 1/1/2010 81 - 181 1/1/2010 82-160 1/1/2010 83-159 1/3/2010 84-158 13/2010 85-157 1/3/2010 86-156 12/31/2009 87 - 155 12/31/2009 88-154 12/31/2009 99-53 1/2/2010 100-52 1/2/2010 101 -51 1/2/2010 Ref. doc 2
Hoop Tendon Group 1 ID Date 14 1/2/2010 17 1/2/2010 20 1/2/2010 23 1/4/2010 26 1/312010 29 1/3/2010 32 1/3/2010 35 1/3/2010 38 1/3/2010 41 1/3/2010 44 1/3/2010 47 I/Z/2010 50 1/2/2010 53 1/1/2010 56 1(1/2010 59 1/1/2010 62 1/1/2010 65 1/1/2010 68 111/2010 71 1/1/201O 74 12/31/2009 77 12/31/2009 80 12/31/2009 Ref. doc 3
Hoop Tendon Group 2 1D Date 15 12/31/2009 18 12/31/2009 21 12/31/2009 24 12/31/2009 27 1/2/2010 30 1/1/2010 33 1/1/2010 38 1/1/2010 39 1/2/2010 42 1/2/2010 45 1/2/2010 48 1/4/2010 E1 1/4/2010 54 1/4/2010 57 1/3/2010 60 1/3/2010 83 1/3/2010 68 1/3/2010 69 1/3/2010 72 1/3/2010 75 1/312010 78 1/3/2010 81 1/2/2010 Ref. doc 4
References:
- 1. Turnover Package No. 25221.002-COT-3054-D0127 Dated 419/10, Constuctlon Opening Formwork and Concrete Placement
- 2. Turnover Package No. 25221-002-COT-3051-00119 Dated 2/8/1O, Vertical Tendon Removal and Installation 3, Turnover Package No. 25221-002-COT-3051-00121 Dated 1/21/10, Unit 2 Removal/Reinstallation of Horizontal Tendons Buttress LA & 31
- 4. Turnover Package No. 25221-002-COT-3051-00123 Dated 2/9110, Removal/Relnstatlatlon of Horizontal Tendons between Buttress 3A & 2B for Unit 2
- 1. Date for concete pouring:
- 2. Tendon retensioning 12/16/2010 (Ref. doc : 1)
Vertical Tendon
.ID Date 48-104 12/22/2010 49-103 12/23/2010 50-102 12/23/2010 59-3 12/26/2010 60-2 12/26/2010 61 - 1 12/26/2010 62-180 12/24/2010 63-179 12/24/2010 64 - 178 12/24/2010 66-177 12/27/2010 66-176 12/27/2010 67-175 12/27/2010 68-174 12/26/2010 69-173 12/26/2010 70-172 12/26/2010 71 - 171 12/28/2010 72-170 12/28/2010 73-169 12/28/2010 74-16B 12/26/2010 75-167 12/26/2010 78-166 12/26/2010 77-165 12/27/2010 78 -164 12/27/2010 79-163 12/27/2010 81-162 12/25/2010 81 - 161 12/25/2010 52-160 12/24/2010 83-159 12/27/2010 84-158 12/27/2010 85-157 12/2612010 88-156 12/23/2010 87 - 165 12/23/2010 88-154 12/23/2010 99 - 53 12/25/2010 100- 52 12/25/2010 101 -651 12/25/2010 Ref. doc 2
Hoop Tendon Group 1 ID Date 14 12/25/2010 17 12/25/2010 20 12125/2010 23 12/25/2010 26 12/26/2010 29 12/26/2010 32 12/26/2010 35 12/26/2010 36 12f27/2010 41 12/28/2010 44 12/28/2010 47 12/28/2010 50 12/27/2010 63 12/27/2010 56 12/2.7/2010 59 12/27/2010 62 12/27/2010 65 12/27/2010 68 12/24/2010 71 12/23/2010 T4 12123/2010 77 12123/2010 80 12/23/2010 Ref. doc 3
Hoop Tendon Group 2 ID Date 15 12t23/2010 18 12/23/2010 21 12/23/2010 24 12/23/2010 27 12/24/2010 30 12/27/2010 33 12/27/2010 36 12/27/2010 39 12/27/2010 42 12/28/2010 45 12/28/2010 48 12/28/2010 51 12/2812010 54 12/27/2010 57 12/26/2010 60 12/26/2010 63 12/26/2010 6e 12/26/2010 69 12/25/2010 72 12/25/2010 75 12/25/2010 7B 12/25/2010 81 12/25/2010 Ref. doc 1 4
References:
- 1. Turnover Package No. 25221-003-COT-3054-00127 Dated 219/11, Containment Constuction Opening. Formwork and Concrete Placement 2, Turnover Package No. 2522.1-003-COT.3051.00119 Dated 1/24/11, Remove and Replace Vertical Tendons for Unit 3 SGR
- 3. Turnover Package No. 25221.003-COT-3051-00121 Dated 1/25/11, Remoy/Replace and degrease Horizontal Tendons Between Suttress IA-I3
- 4. Turnover Package No. 25221-003-COT-3051.00123 Dated 1/25/11, Remov/Replace and degrease Horizontal Tendons Between Buttresý.2B-3A