ML17345A212
| ML17345A212 | |
| Person / Time | |
|---|---|
| Site: | Turkey Point  |
| Issue date: | 05/31/1988 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML17345A211 | List: |
| References | |
| NUDOCS 8806070215 | |
| Download: ML17345A212 (8) | |
Text
t t
UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555 ENCLOSURE SAFETY EVALUATION RELATED TO REA E
S L
D I
0 A
UKE P IN UN S
AN 4
D C
NOS.
0-2 251 INTRODUCTION The licensee indicates that although the Charpy upper-shelf energy (USE) for the Turkey Point Units 3 and 4) (TP) reactor vessel limiting beltline material will be below 50 ft-lb, their fracture mechanics analysis indicates that the material will meet the safety margins of Appendix G of the ASME Code for at least 40 effective full power years (EFPY).
- However, our review of the analy-sis indicates that further analysis and data acquisition are necessary.
The need for additional analysis and data acquisition is discussed in this Safety Evaluation.
Section IV.A.1 of Appendix G, 10 CFR 50 requires',
in part, that the Charpy upper-shelf energy (USE) for all reactor vessel beltline materials be above 50 ft-lb throughout the life of the vessel, unless it is demonstrated in a manner approved by the Director, Office of Nuclear Reactor Regulation, that lower values of USE will provide margins of safety against fracture equivalent to those required by Appendix G of the ASME Code.
Section V.C.3 of'ppendix G, 10 CFR 50 requires that the licensee perform analyses to demonstrate the existence of equivalent margins of safety when the Charpy USE is predicted to be less than 50 ft-lb.
In letters to the Office of Nuclear Reactor Regulation, USNRC dated May 3, 1984 and March 25, 1986, the licensee provided analyses, which are intended to d~gonstr~te that at 40 EFPY, which corresponds to a neutron fluence of 2.88 x 10 n/cm (E>1MeV) at the vessel's inside surface, the fracture toughness of each of the reactor vessels meets the safety margins of Appendix G of the ASME Code.
Appendix G of the ASME Code presents a procedure for calculating the allowable pressure for pressure vessels.
The procedure is based on the principles of linear elastic fracture mechanics (LEFM).
This ASME Code procedure postulates that the Turkey Point pressure vessels have a sharp surface defect normal to the direction of maximum stress that has a depth of one-fourth of the section thickness (1/4 T) and a length six times its depth.
For Levels A and 8 service conditions, the safety margin on the allowable pressure is required to be a
factor of 2.
Appendix G does not contain fracture toughness limits for Levels C and 0 service conditions.
In NUREG-0744, Rev.
1 dated July 1982, the staff provided guidance for perform-ing the analyses required by Section V.C.3 of Appendix G, 10 CFR 50.
The recommended procedure to be followed is based on the J-Integral Elastic Plastic Fracture Mechanics (EPFM) method.
In this method the material fracture resistance is measured using the parameters J, the intensity of the plastic stress-strain field surrounding the crack tip, and T, the tearing modulus.
These parameters must be determined by testing of neutron irradiated material, which is equivalent to the material in the reactor vessel beltline.
The test 8806070215 880531 PDR ADQCK 05000250
I P
limits on these parameters depend upon the amount of J-controlled crack growth.
The maximum load-carrying capability of the irradiated reactor vessel occurs when the calculated values J
and T
for the reactor vessel with the postulated flaw equals the J and T
for the irradiated material.
When J
exceeds J
, the postulated flaN $s considered to be unstable.
The NUREG i% lcates that late value of the ratio of J/T for surface cracks is dependent upon the material's flow stress, the postulated crack size, a geometry correc-tion factor, and a stress correction factor.
The NUREG indicates that the margins for Levels A and B service conditions should be no less than that now required by the ASME Code, Appendix G.
The NUREG does not specify the margins required for Levels C and D service condi-tions.
In a letter from R.
E.
Johnson to L. T. Chockie dated April 20, 1982, the NRC requested that the ASME Code Committee develop safety margins for all service conditions.
DISCUSSION The margins of safety against fractures were determined by comparing the pre-dicted value of J at instability to values of J due to normal operating (Levels A and 8) P$ f'esses acting on the ASME Code )8ktulated flaw.
J values for the TP limiting beltline welds were extrapolated from a Heavy SecPQ'n Steel Technology (HSST) welds, which was fabricated using the same heat of weld wire and flux as used in the limiting TP welds.
However, the HSST data was irradiated in a test reactor, which has a much higher neutron flux than a
commercial reactor.
The J-T curves used to determine the material elastic plastic fracture resistance were developed from 1.6 T compact toughness (CT) specimens.
As a result of specimen size limitations the amount of J-controlled crack growth is limited to approximately 5
mm.
NUREG-0744 describes a method for extrapolating beyond the J-controlled growth limits when small specimens are used to determine the material's fracture resistance.
This method was not followed in the licensee's analyses.
Extrapolation of data beyond the J-controlled growth limits is being studied by the Office of Nuclear Regulatory Research and the ASME Code Section XI Committee, "Working Group on Flaw Evaluation."
To determine the material fracture resistance curve (J
,T t) as a function of neutron fluence, the licensee extrapolated HSST dat5 3siE) a relationship observed between J t and T an empirically derived relationship between J
and Charpy
- USE, and th89elationship between Charpy USE and neutron fluence r8j33rted in Regulatory Guide 1.99, Rev.
1.
The relationships observed and derived in the analysis provide values for J and T beyond the J-controlled growth limits.
The licensee has not provided material test data to demonstrate that these relationships apply beyond the J-controlled growth limits.
The licensee must provide supplemental fracture toughness data from a commercial reactor surveillance program that demonstrates their analysis, which used HSST
- data, applies to material irradiated in a commercial reactor.
In accordance with Section III.B of Appendix G, 10 CFR 50, the test methods used to provide the supplemental data must be submitted to and approved by the Director, Office of Nuclear Reactor Regulation, prior to testing.
I
-3" The staff recommends that the relationship between neutron fluence and Charpy USE for the TP reactor vessel beltline materials be predicted using the methodology recommended in Proposed Regulatory Guide (R.G. ) 1. 99, Rev.
2.
This guide recommends that the calculation be performed using a line drawn parallel to the existing trend curves and bounding all the data when credible surveillance data is available.
This method, although conservative, is necessary when plant-specific data are sparse and scattered.
To date, only three capsules that contain weld metal specimens have been withdrawn and tested.
The licensee should use all weld metal surveillance data from these three capsules to determine the relationship between Charpy USE and neutron fluence.
In a letter dated April 22, 1985, the staff approved an Integrated Surveillance Program for Turkey Point (TP) Unit Nos.
3 and 4.
The test results from material irradiated in surveillance capsules in these vessels are to be used to determine the vessel's fracture toughness resulting from neutron irradiation.
Our evaluation of the fracture toughness data derived from the last capsule withdrawn from TP-3 is contained in a letter from D.G.
McDonald to C.O Woody, dated October 30, 1987.
The Safety Evaluation contained in that letter indicates that the formula in R.G. 1.99, Rev.
2 conservatively predicts the effect of neutron irradiation on the limiting weld metal in the TP-3 and TP-4 reactor vessels.
When the licensee used the empirically derived relationship between J
and Charpy USE to determine the Turkey Point material fracture resistance,
)he licensee assumed that the J tvalues from the HSST data corresponded to the Charpy USE values from R.G. k.99, Rev.
2.
This assumption is incorrect and results in a non-conservative value for J at instability.
The licensee should have used actual Charpy USE data from the HSST program to determine the relationship between Charpy USE and J t for the Turkey Point beltline materials.
To determine the value of J at instability the flow stress must be known.
In the licensee's analysis the flow stress for the Turkey Point material was derived from the HSST data.
Based on the TP surveillance program test results, the value of the flow stress at the end of the plant's license was underesti-mated.
- However, lower values of flow stress produce conservative values for J at instability.
The J at instability was determined for a neutron fluence of 1. 73 x 10 n/cm (E>1MeV).
This was calculated to be the neutron fluence at the tip of the postulate/94 T depth flaw when the neutron fluence at the inside surface is 2.88 x 10 n/cm (E>lMeV) and the TP reactor vessels reach 40 EFPY.
The attenuation of neutron fluence from the inside surface to 4 T depth was per-formed using a non-conservative method.
To determine the effect of neutron irradiation on the TP beltline materials, the neutron fluence through the vessel wall should be attenuated using the formula for displacements per atom in R.G. 1.99, Rev.
2 or SECY 82-465, "Pressurized Thermal Shock."
I
The licensee's calculation of J at the tip of the
~4 T postulated flaw included an elastic component,
%5 did not include a plastic component.
The stress calculation includes values for the membrane stress from internal
- pressure, the pressure on the crack surface, the temperature changes during heatup and cooldown and residual weld stress.
When these values are summed the author indicates that the value is low enough to permit the use of only the elastic component for calculating J
- However, when the allowable pressure is doubled, in accordance with 5e safety margins required by Appendix G, the applied stress is near the irradiated material s yield stress.
When the applied stress is near the materials yield stress the plastic compo-nent of J can be large and should be considered in the analysis.
- Hence, to demons55te that the postulated
~4 T flaw meets the safety margin require-ments of Appendix G during Levels A and B service conditions the plastic com-ponent of J must be added to the elastic component.
In addition to Levels A and B service conditions, the reactor vessel's design must consider Levels C and 0 service conditions.
The licensee's analysis does not consider these service conditions.
The safety margins for fracture resistance during all service conditions are currently under discussion in the ASIDE Code Section XI Committee, "Working Group on Flaw Evaluation."
When the Committee provides reactor vessel fracture resistance safety margins for all service conditions and when they have been approved by NRC, the licensee should determine whether TP can meet these safety margins.
CONCLUSION In our Safety Evaluation, we indicate that additional analysis and material test data are needed to confirm that the TP reactor vessels will meet the safety margins of Appendix G of the ASME Code and 10 CFR 50 for 40 EFPY.
Until this information is supplied, we can not complete our review of the licensee's submittals.
Dated:
t!ay 31 ( 1988 Princi al Contributor:
B. Elliot
hf
~
pA it
/
f